Dtic Ada326694
Dtic Ada326694
Dtic Ada326694
Proceedings of the
19970623 324
18-21 November 1996
Naval Postgraduate School, Monterey, California
pgBTWBimOW CTATEMEHT
I Afipzoved ic3t gtthlie teleoMI
V. Diimbxmnc LbJxnuted i
Welcome
to the
Naval Postgraduate School
Few post-Cold War challenges possess the urgency of the Mine Problem
both in military and humanitarian terms. We at NPS are dedicated to the
exploration of technical approaches to the solution of the Mine Problem. Your
generous participation in this Symposium series underscores the community-wide
appreciation of the urgency of this problem. Together, I feel certain we shall
meet our objective of “changing the world.”
The Faculty, Staff and Students of NPS stand ready to acquaint you further
with this very special place. I sincerely hope that, while you are here, you will
avail yourselves of the opportunities to get to know us, to see the possibilities in
the Technology Transfer Program, and to forge professional networks to deal
with the multi-faceted dimensions of the Mine Problem.
Marsha J. Evans
Rear Admiral, United States Navy
Superintendent
Proceedings of the
Edited by
Volume I ofII
1
ACKNOWLEDGEMENTS
An enterprise of the magnitude of the Symposium on Technology and the Mine Problem is
a coordinated team effort amongst sponsors, speakers, schedulers. Session Chairs, symposium staff,
and the attendees. The planning horizon for such a symposium is over one year for staff and
presenters alike. On behalf the Mine Warfare Association, I extend thanks and appreciation to the
following:
Navy Sponsors:
Office of Naval Research : RADM Paul G. Gaffney II, USN, Chief of Naval Research; and
Dr. Fred Saalfeld, Deputy Chief of Naval Research and Technical Director,
Office of Naval Research
Naval Postgraduate School: RADM Marsha J. Evans, USN, Superintendent; and
Professor James N, Eagle, Chair, Undersea Warfare Academic Group
Corporate Sponsors (for the Session on the Littoral Environment at the Monterey Bay Aquarium):
Office of Naval Research
Raytheon
Mercury Computers
A&T Inc.
TRACOR
Lockheed-Martin, Inc.
OmniTech Robotics
ObjectTime Pacific LLC
International Submarine Engineering (ISE), Inc.
Symposium Staff:
Ms. Barbara Honegger, M.S., Program Coordinator and Proceedings Editor,
ROLANDS & Associates
Ms. Suzanne Wyatt and Ms. Melody Burgess, Destination Monterey/Carmel
Dr. Joseph J. Molitoris, Secretary-Treasurer, Mine Warfare Association
CDR John Peterson, USN (Ret)
Dr. Robert F. Rowntree
The Symposium Announcement and Registration Issues of MINE LINES were sent to an
expanded mailing list. This was made possible by planning funds from the Office of Naval Research.
The PROCEEDINGS of this Symposium will be mailed in February 1997 to each registrant as part
of the Registration Fee.
MINE LINES is the Newsletter of the Mine Warfare Association. Through MINE LINES
we seek to stimulate professional exchange and to announce the periodic workshops and meetings
that MINWARA will sponsor and co-host. These events are in addition to the sesquiannual
Symposium on Technology and the Mine Problem.
A Membership Application for the Mine Warfare Association is available in this Proceedings.
Further information, and information on Corporate Membership classes and benefits, can be obtained
from the MINWARA Secretary-Treasurer, Dr. Joseph Molitoris, at (703) 339-7244.
IV
TABLE OF CONTENTS
PREFACE
Volume I of H
Acknowledgements. .iii
.V
Table of Contents.
Introduction. ...xix
Prof. Albert M. Bottoms
Welcoming Remarks. xxvii
CAPT James M. Burin, USN
Sponsors’ Remarks. xxix
RADM Paul G. Gaf&iey II, USN; Dr. Fred E. Saalfeld; Dr. David Skinner
Introductory Remarks. ,xxxi
Prof. Albert M. Bottoms
Keynote Address
GEN John J. Sheehan, USMC
Introduction.
Introduction.3-1
vi
CHAPTER 4: LANDMINES AND HUMANITARIAN DEMINING
Introduction.4-1
The History and Future of U.S. Policy on a Universal Anti-Personnel Landmine Ban.4-5
The Hon. H. Allen Holmes, Assistant Secretary of Defense, SOLIC
Bringing New Technology to Bear on Landmine Detection: The Role of NGOs as Catalysts
and Liaisons Between Technology Providers and Mine-Affected Countries.4-35
Richard M. Walden
Minefield Proofing and Route Clearing in Bosnia Using Unmanned Ground Vehicles
and the Standardized Teleoperation System.4-67
David W. Parish, LTC Jon Moneyhun
vii
Multi Sensor Vehicular Mine Detection Testbed for Humanitarian Demining.4-73
Douglas Brown, Joseph Bendahan, Giancarlo Borgonovi, Delmar Haddock, Jason Regnier
Introduction.5-1
The Basic UXO Gathering System (BUGS) Program for Unexploded Ordnance
Clearance and Minefield Countermeasures; An Overview and Update.5-3
Christopher DeBolt, Christopher O’Donnell
Vlll
GPS and Mine Warfare.
James R. Clynch
Mission Definition for AUVs for War Gas Ammunition Deposit Assessment.5-181
Marek Narewski, Leszek Matuszewski
ix
CHAPTER 6: COUNTERING MINES ON LAND
Introduction.6-1
Introduction.
The Underwater Influence Fields of Target Ships: Some Mine Sensor System
Considerations and the Strengths and Weaknesses of Influence Mine Sweeping.7-3
Dennis R. Hiscock, Royal Navy, U.K. (Ret)
XI
Rapid Response Minesweeping 7-123
Carl Fisher
xii
Mine Burial Experiments in Carbonate Sediments.7-253
Daniel F. Lott, Kevin L. Williams, Darrell R. Jackson
Introduction.
Is the Application and Utilization of All the Required Skills Being Applied
Within the Mechanical Mine Clearance Arena?.8-57
William N. Baker
Application of Seismic Vibration Concepts for Rapid Milie Clearance and Detection....8-65
Geoffrey C. Davis, Steve Ballinger
Deployment Modeling.8-113
Keimard Watson, Thai Nguyen
XIV
CHAPTER 9: THE UNDERLYING SCIENCE OF MINE WARFARE SYSTEMS
Introduction.
The Use of Ocean Optical Data to Predict the Performance of Mine Detecting
Ocean Lidar Systems.9-49
R. Norris Keeler
Introduction.19-1
XV
A Network-Centric Information Systems Architecture...10-47
Rex Buddenberg, LT Steve Graves, USN
Introduction..Ij.l
Introduction.12-1
Identifying and Sizing the Mine Warfare Market: Summary Overview of Roundtable 12-5
Ric Trotta, with Comments by Ron Blue
xvi
APPENDIX A: A TUTORIAL BRIEF: The Technological Revolution
in the Mine Problem, including Bibliographies on
Naval Mines and Landmines.A-1
Prof. Albert M. Bottoms
xvii
XVlll
INTRODUCTION
Professor Albert M. Bottoms
Naval Postgraduate School
This volume contains the PROCEEDINGS and contributed papers of the Second Symposium
on Technology and the Mine Problem, held at the Naval Postgraduate School November 18-21,
1996. The First Symposium, entitled Symposium on Autonomous Vehicles in Mine Countermeasures,
was held at the School in April 1995.
This Second Symposium was dedicated to the memory of Admiral Jeremy "Mike" Boorda,
USN, the former Chief of Naval Operations, who was a staunch supporter of efforts to harness
technology to deal with the Mine Problem.
The Honorary Chair of this Second Symposium was Rear Admiral John D. Pearson, USN,
the outgoing Commander of the U.S. Navy Mine Warfare Command.
The Organizing and General Chair of the Symposium was Albert M. Bottoms, Ellis A.
Johnson Chair of Mine Warfare at the Naval Postgraduate School and President of the Mine Warfare
Association.
VISION STATEMENT
THE CHALLENGE
XIX
GOALS FOR THE 1996 SYMPOSIUM ON TECHNOLOGY AND THE MINE PROBLEM
* Identify the technologies that can revolutionize approaches to dealing with the mine
problem;
* Emphasize those technologies which contribute to the Navy-Marine Corps Mine Warfare
Campaign Plan and its thrusts to support Operational Maneuver from the Sea and "organic" mine
countermeasures;
* Define the scope, magnitude, and future course of the national and international markets
for mine clearance-related technologies and systems, including those based on commercial
off-the-shelf (COTS) technology and products.
In consonance with the objective of establishing the Naval Postgraduate School as a focal
point for mine-related technology and analysis, it is the intent to hold a major technical Symposium
at the Naval Postgraduate School at intervals of 18 months. The next Symposium will be the week
of April 5, 1998, and will emphasize progress in the development of autonomous systems for mine
countermeasures/countermine applications, C4I, including tactical decision aids and distributed
modeling and simulation; and progress toward breaching — overcoming obstacles in the surf zone,
on the beach, and inland. Each of these major subject areas will be viewed from the standpoint of
applications to military mine warfare on land and at sea and to humanitarian demining.
Session Chairs:
XX
RECOGNITION OF SYMPOSIUM CHAIRS (continued)
XXI
RECOGNITION OF TECHNICAL CONTRIBUTED PAPERS
The Mine Warfare Association established two prize categories for contributed technical
papers, first presented at this 1996 Symposium on Technology and the Mine Problem. The CAPTAIN
SIMON PETER FULLINWIDER Awards are for the best papers submitted by serving members of
the Armed Forces. The First Prize in this category will carry an honorarium of $500 and a Life
Membership in the Mine Warfare Association. The Second and Third Prizes will, respectively, carry
honoraria of $250 and $100. Each will also be accompanied by Life Membership in the Mine Warfare
Association.
Captain Simon Peter Fullinwider (1871-1957) is deemed the Father of Mine Warfare by the
U.S. Navy. Additional information about the contributions and energy of this remarkable man can be
found in Dr. Greg Hartmann’s book Weapons That Wait. This year the award was presented by
RADM John D. Pearson, USN (Ret), Honorary Chair of the 1996 Symposium.
The Charles Rowzee Awards are for the best overall technical papers. The schedule of awards
is the same as that for the Fullinwider Awards.
Charles Rowzee is the individual who applied years of experience in mine design to, in effect,
enable the conversion of the large stocks of bombs into influence mines. This technical achievement
led to the mining campaign against North Vietnam. That campaign, in turn, led to the return of the
North Vietnamese to the negotiating table and to the subsequent release of Americans held captive
by North Vietnam. The 1996 Rowzee Awards were presented by Mr. Charles Rowzee himself
Dr. Ellis A. Johnson, Captain Simon Peter Fullinwider and Mr. Charles Rowzee are but three
of the intellectual and operational giants to whom the United States owes its distinguished
accomplishments in the fields of Mine Warfare. There are many others, both in and out of uniform.
Perhaps a long-term project for the Mine Warfare community could be the creation of a Mine
Warfare Hall of Fame.
The Award recipients for this Second Symposium on Technology and the Mine Problem are:
The 1996 Fullinwider Awards
First Prize
Col. Robert Greenwalt, Jr., USA
The Engineer Center, Ft. Leonard Wood, MO
"Systems and Techniques for Countering Mines on Land"
Col. Greenwalt's papers appear in Chapters 2 and 6
Second Prize
Lt. Col. Dennis Verzera, USMC
Coastal Systems Station, Panama City, FL
"A New Dimension in Amphibious Warfare"
Lt. Col. Verzera’s paper appears in Chapter 7
XXII
The 1996 Fullinwider Awards (continued)
Third Prize
Capt. Charles Young, USN
U.S. Navy Unmanned Undersea Vehicles Program Office
"Clandestine Mine Reconnaissance, Unmanned Undersea Vehicles"
Capt. Young’s paper is in Chapter 5
Honorable Mention
Col. Leroy Bamidge, USAF
Commander, 28th Bombardment Wing, Ellsworth AF Base
"U. S. Air Force Roles in Mine Warfare"
Col. Barnidge’s paper is in Chapter 2
Group Awards
Don Brutzman, Bryan Brauns, Paul Fleischman, Tony Lesperance,
Brian Roth and Forrest Young, Undersea Warfare Academic Group, NPS
"Evaluation of AUV Search Tactics for Rapid Minefield Traversal Using
Analytic Simulation and a Virtual World”
Their paper is in Chapter 10
Group Awards
Capt. Thomas R. Bemitt, USN, Commander, Explosive Ordnance Disposal
Group One; CWO G. Mike Johnson, USN; Senior Chief Petty Officer
Chris A. Wynn, USN; and Lt. Eric Basu, USN
"Developments in the Very Shallow Water Mine Countermeasures Test
Detachment Program”
Their paper is in Chapter 3
First Prize
Prof Carl Schneider, Ph.D
Professor of Physics, U.S. Naval Academy
"Maxwell's Equations in Magnetic Signature Analysis"
Prof Schneider’s paper is in Chapter 9
Second Prize
Major Colin King, Royal Army (Ret.)
Jane's Information Group
"Landmines and Humanitarian DeMining"
Major King’s paper is in Chapter 3
XXlll
The 1996 Rowzee Awards (continued)
Third Prize
Ms. Helen Greiner
I SR Robotics, Inc.
"Enabling Technologies for Swarm Coverage Approaches"
Ms. Greiner’s paper is in Chapter 5
Honorable Mentions
Prof J. D. Nicoud, Ph D.
Laboratoire de Micro-Informatique EPFL
Lausanne, Switzerland
“GPR and Metal Detector Portable Systems,” “Post-conflict and Sustainable
Humanitarian Demining,” and “Cooperation in Europe for Humanitarian Demining”
Prof Nicoud’s papers are in Chapters 4 and 6
XXIV
Remarks of
Mr. Charles A. Rowzee
To have an award bear my name is truly an honor and one of the highlights of my career —
especially an award in recognition of solving Mine Problems.
Before proceeding with the award presentations to the winning participants, let me say a few
words about a weapon whose development is the reason why I’m here tongiht. This weapon
development was responsible for resolving a difficult sea problem — the interdiction of roads and
inland waterways. I am refering to the Destructor Weapon. This weapon system, consisting of an
armful of components, converts the MK-80 series bomb into an underwater or land mine.
Development fi'om concept to deployment was accomplished in ten months, providing the Fleet with
a safe, effective weapon at a cost of less than a pound of hamburger per pound of weapon.
If you think that this is the complete story, don’t believe it. Now let me tell you the “rest of
the story.” Very simply, it’s the Navy Laboratories, where individuals gain knowledge and experience
to resolve challenging problems. This major weapon development -- concept to deployment in record
time “ could have only been achieved through the years of experience I gained at the Navy
Laboratoiy in White Oak, Maryland. So I say, “Thumbs Up” for the Navy and Defense Labs.
XXV
XXVI
WELCOMING REMARKS
Good morning. First of all, let me welcome you to the Naval Postgraduate School and
Monterey. You have assembled an impressive group and you are meeting on a critical topic.
We have a lot of expertise and technology research here at NPS in mine warfare and
related fields, so it’s an ideal setting for your conference. While you’re here, please feel free
to talk to NPS staff and students and see what’s going on here in these vital areas.
Since I have the opportunity, I would like to give you a brief on mine warfare. I have
dropped some mines, both on land and at sea, and have some mine warfare experience. I
have noticed that mine warfare, like nuclear weapons, used to be in war games. But no
longer; no one wants to play because these weapons are a real ‘show stopper.’ Hopefully,
you can fix that, so that mines are no longer show stoppers.
And that leads to my final point. We need to think about mine warfare as a full
spectrum problem, like we did about regimental backfire raids on Navy battle groups. Cruise
missiles were another tough problem, like mines. So we tried to kill the archer, not the
arrow. Then we went beyond air-to-air warfare, where we didn’t just kill the arrow and the
archer, but the quiver — we used strike warfare. That should become a part of mine warfare,
too. Get our strike warriors to find mines and kill them on the beach. We need to use our full
spectrum warfare capabilities and technology to attack this difficult problem.
Again, welcome to the Naval Postgraduate School. I hope and trust you will have a
valuable and productive conference.
xxvii
XXVlll
SPONSOR’S REMARKS
5 November 1996
Dear Colleague:
XXIX
SPONSOR’S REMARKS
Dr. David Skinner
Executive Director
Coastal Systems Station, Dahlgren Division
Over the course of the past year we have made great strides in the improvement of Mine Warfare
(MIW). The NSIA, ADPA, and now the Mine Warfare Association Conference have succeeded
in educating and involving industry and academia to a high degree.
The Campaign Plan has provided a rallying point for the future direction of MIW and the Navy is
taking notice. Readiness has been improved through the forward basing of MCM-1 ships and the
development of contingency systems like RMS and Magic Lantern. Development programs have
been streamlined and integration improved from 6.1 through 6.5. Management coordination and
interaction have been improved throughout the MIW community through forums like the Flag
OfLsites, the Acquisition Coordination Team, and MIW Technology Team.
The near-term goals have been accomplished. The mid-term is closing in and we are on track, but
the MIW problem is far from solved. Our vision for the far-term must now be crystallized.
The road ahead is sure to be as full of changes as the recent past. The Navy and the DOD are still
evolving roles, missions, and functions. As we strive to keep pace with this evolution, several
things are clear:
How, then, do we solidify this far-term vision? We have already started down the path. We are
establishing a common language for analytical discussion of MIW, and we are quantitatively
baselining our near and mid-term capabilities with sophisticated modeling and simulation
capabilities and Fleet exercises. We must then:
Assess our expected capabilities against these needs to determine if we have shortfalls;
Program, restructure, and adjust as required to provide the Naval MIW capabilities.
XXX
INTRODUCTORY REMARKS
Professor Albert M. Bottoms
Symposium General Chair and
President, Mine Warfare Association
General Sheehan, General Howell, General Gill, Admiral Conley, Admiral Pearson,
Admiral Gaffney, Dr. Saalfeld, distinguished guests and attendees:
It gives me great pleasure to open this second in the planned series of major technical
Symposia at the Naval Postgraduate School on Technology and the Mine Problem. We plan
to hold these every 18 months. The next one is scheduled for April 1998.
The seriousness and urgency of the mine problem can scarcely be overstated. Each
person here has interest in and responsibility for some facet of the problem of mines —
operational, technical, programmatic, or policy. These concerns apply to sea mines, land
mines, and to humanitarian demining. We note that technologies that relate to mines and
mine countermeasures also apply to the efforts to remediate areas contaminated with UXOs
or hazardous materials. Mine technology and countermine processes may also be applicable
in counter-terrorism.
This week Monterey is the mine capital of the world. We at the Naval Postgraduate
School have a vision as to how the emergent technologies about which you will hear
eventually will be combined into affordable, autonomous systems that can deal effectively
and in a timely manner with mines, booby traps and obstacles. This is precisely what we
mean when we say that our objective is “to change the world.”
Systems people, along with programmatic sponsors, also think in terms of milestones
and time lines. You will hear about the ACTD candidate technologies. The ACTD field
exercises in FY’97 and FY ‘98 are the next official milestone events. But much of what you
will hear falls on either side of these ACTD milestones.
Some ideas, such as bulldozers and rakes, may be described as “low tech,” but, as Mr.
Bill Baker points out, also “high technique.” Other ideas involve computational power and
xxxi
flexibility only now coming within grasp. We urge you to help us identify these post-ACTD
milestones. Help us to define for the mine warfare System of Systems the initiatives that
correspond to “planning wedges” and “block upgrades” for platform acquisitions.
Now, I ask that we pause in our anticipation of the program over the next four days
to honor the memory of Admiral “Mike” Boorda, USN, former Chief of Naval Operations,
who personally encouraged our efforts and our vision for mine countermeasures systems at
NPS. He wrote that he concurred that the vision is within grasp.
I now call upon the NPS Command Chaplain, Chaplain John Wright, to give the
invocation for the Symposium.
XXXll
CHAPTER 1: THE CHALLENGE
KEYNOTE ADDRESS
Key Points:
The General’s remarks underscored the current inadequacy of U.S. military mine
countermeasures, both at sea and ashore, and forecast the likely continuation of the
mine problem if money and resources do not match rhetoric.
Presented herein with General Sheehan's express permission are the visuals that
he used in his presentation. He spoke without notes.
1-1
Confronting the Mine
Warfare Challenge of
U
a
u
0)
0)
a
‘■*3
C<3
1-2
•Nii
a
l.
a
18 November 1996
Outline
Asymmetrical Warfare
1-3
U
H
Q
o
g
a
a>
0^
1/5
01
Allied Command Atlantic
Ihiited^ates Atlantic Command
21st Century Military Forces
a
c«
QJ
/3
O
aU .ti ^
u
QJ O'
1/3
An Asymmetrical Battlefield—
c
Cm
■33.2f.2f .2f
ffl !l! Z
Q
a
/3
a> fl
o 1/3
o c/3
•PN
-M
o fl
a b ®
OX) o
.2 ^ u O *r*j
c -a G
•M © a
•Sy sS 5s a a
United^ates Atlantic Command
O
a pa 1/3
a; ©
u 'a ©
^ ^ s 3 H a pa a^
Allied Command Atlantic
OJ a
o .2f5
^ ^ C/5 N c/i
^ pO •M
I a u
© ©
e
...
R " *3d a <
“
Q U c/5 an Q P
®
.5s
O
♦
1-4
Asymmetrical Warfare
—The Mine Threat—
N
U
QJ
u
o
a
o
U
S>-
<3J
c/i
cs
I/}
a>
U
1-5
©
©
©
!/3
>>
©
C/J
a>
1/5
diD
WD
hS
S
0^
CQ C3
c/^
uu
Allied Command Atlantic
llmie^rSlates Atlantic Command
© ‘C
^ £
IS U
W I
^ QJ
I
c/5 5m
a Sk -2
«.s ^
tzi
•m
^
-BS
© ^
© c« ns O
2 I
s©
fl flj
lyj i
o _rt
& ®
_© .5
C3 ^
United ^ates Atlantic Command
I^ "d
Allied Command Atlantic
2 c »—I o
.2
II
1-6
i
•A ^
*1 I
VI ©JD © ^
!3 ^
CNO’s Assessment
^ «s
©JD C©
§ S § 5
o ©
^
^
c/3
© I
s
•N
^ «t-l
s: ©
© ©
V X*
© ^ ©
e
o ^ g
S
-Si
03
'S I
c -Si
r. ©
I
©X) TJ © ^
^ ©
Uaited^ates Atlantic Command
.s ^ ^ ©ID
fl
© ^
©
Allied Command Atlantic
©
& 4>
© ;-i
X! © © ©
H o X
1-7
Since Desert Storm
Oh
O
O
’S
©
<
H
O
.2 *2
O
X
©
©
©
©
••
fl
u
"3
#C
"O
ifi
m
0^
til
o
u
c/3
CQ
d
■
Q
c/3
o
a>
'cs
1-9
<
u
0^
o
O
u
0^
O
• •
C/5
I
a
a
1-10
c
a;
••
yi
O
u
CJ
ffl
1^
c/3
c/5
D
a>
C/5
u
u
1-11
Allied Command Atlantic
llmted~Slates Atlantic Command
JCM ACTD Filtering Process
Testing Approval
cj
Cd
fl
0^
i/J .aj o c/5
c/5
-« la • pN
'aS W cd
u
® "S a> O
aj a> fi H
c .S o Cd
fl
CZ)
S So fl
©
• pN
-<-»
O o
T3 U
0
TS
u
'o
•^•d
iS
c« c« P"N 3
S (DiD
m
a> 0^ ©
S ©
cd
fl
• P-N
O)
c/5
o ©
©
W)
u
© c3 fi
• pH
&4 cd
o u
0^ ;_4 fl
cd
u s
"S
©
1 ^ 3
•pH
CQ
H
llnitedlSfates Atlantic Command
©
S
>►
o u C/)
©
HH Cd
©4
(DJD
c
3
• pH
a> u
a
©
'4'^
U s© 3 • paH Td
©
Allied Command Atlantic
T3 U • PM<
© o
CQ O 1
u
o
U
a I I I
1-13
v>
Joint Countermine ACTD
O I
Ck CC
On a ON
O U
P r 1
<
On
u H
/■ >
o tTi
>►
a>
w
tzi
©
A
•rj Sn c/3
ON
©
ON A c/3 .S^
(1^
^ Vl-N
United States Atlantic Command
O)
o H ’& «
s
Allied Command Atlantic
O
< Bi
Q Q
1-14
Q
H
O) u
©
©
©
U ©
c/5
© ©
o
Joint Countermine ACTD
c« p
A ^
-Where we need to go?—
c/5
&? c^
Cm
©
© *s ^
© .
D s O
fl
s© s©
Bo O
©
!h PQ 'S?
© ^
•p-<
^ or'
o
W) ?>
93 a> Sn
o 5
;h
5
'o O fl 13
u c«
c« 5 a
U
.. o
©
u
*5/5
2
©
c^ o d
“ rv^
U
p ©
u
©
pp 2</5
S T5 P
^ cj cd Pi o © ©
©
©
£ 2 g © © ©
^ H *2 P ©
llMited~^ates Atlantic Command
• • ^ •P-N WD
H ^ ^ © p
W O T3 "©
Allied Command Atlantic
"d ©
p © o ©
© ©
•Pii^ © © ©
^ 11
c/5 o s
© O ns
Q ©
c«
5-
U ^ P
H S Pi
1-15
NATO’s MCM Resources
ifi
<
U
H
H
u
JS
*S
a
1-16
Allied Command Atlantic
l^ee^qtes Atlantic Command
i
I
Cooperative Telos
-Corpus Christi, TX,18-29 Sept 97-
1-17
Allied Command Atlantic
Unite^rSiates Atlantic Command
Challenges
MCM Technology Remains Suspect
a
S!
u
C/5
c/5
a>
C/5
W)
t/5
1-18
O
Allied Command Atlantic
United ^ates Atlantic Command
mpediments to Progress
1-19
0
ns
a>
U
a
W)
a
&
p
a
u 13
u
'p 0^
g
P 1 O #
C/2 ••
QJ
•• c/3
CZl
a> fl
United States Atlantic Command
o
cu
13 c/3
Allied Command Atlantic
a> 0^
• •
1-20
DUMB MINE
What is Undersea Warfare?
1-22
C^*
1-23
Allied Command Atlantic
llnitedlSfates Atlantic Command
MINE WARFARE
1-25
Allied Command Atlantic
UjiUedrSlates Atlantic Command
MINE WARFARE
1-26
Allied Command Atlantic
llmtedlSfates Atlantic Command
BACK-UP SLIDES
rs Atlantic Command
1-28
Allied Command Atlantic
Ihuted^ates Atlantic Command
1
INTEGRATED MINE EIELD
1-29
Allied Command Atlantic
Umted States Atlantic Command
Mines are available to forces that
1-30
Allied Command Atlantic
United ^ates Atlantic Command
.a <» «a
S a>
o ■
M <S£
I93
C«
T5
U r* ;-i
n .a a>
5 aIV CJ &
fl
S — a»
QJ
G
s QJ
in
s U
^ t« T3
a S a
lljiited~^ates Atlantic Command
Ti
o
tc .2 o DX) OX)
^ '■&
c G
Allied Command Atlantic
53 *3 53
•
• pN
*-ia> "S ^
a^ D G
"SG
^ a a< a A G G
& ffi 0 3 0 h-9
• • • • • •
1-31
&JD
CJ
U
a
cr
!/5
Mine Warfare
(Iraq)
B
c
4)
Qi
VI
o ^
C«
Xi
® S
T5
©
©
0^
iS
©
©
"E
1-33
""
u
©
©
© ©
03
CA5
© P
©
Ph ©
©
P
©
W)
•
a
a
QJ
c/5
o^
'sO
•
Tl
2
©
1-34
Allied Command Atlantic
United ^ates Atlantic Command
Technology, the Budget
and Politics
You have heard from a variety of speakers: GEN Sheehan, RADM Pearson, RADM Denny
Conley and others who have recounted the history and menace of mine warfare from colonial days
through our latest full-blown conflict in the Persian Gulf
One of the major lessons of the Iran-Iraq War has to do with the continuing menace of even
relatively low-technology seamines. When we were providing tanker escort through the Gulf, the
Supertanker Bridgeton hit a 1908-designed mine laid by the Iraqis the night before, tearing a 30 foot
by 40 foot hole in her 2-inch steel hull. (Bill Mathis, the escort commander who is in the audience,
proved that surface warriors are not very smart — he went through the minefield twice!) The Roberts,
the Tripoli and the Princeton are all examples that have been cited during the conference of how
deadly and inexpensive mines are, and how a militarily inferior country can use them to their strategic
advantage to effect a political or military outcome. The cost to taxpayers to repair the damage to
these ships totalled about $21.6 million. The cost of the two mines has been estimated at about
$15,000 — a great return on investment.
Generals Sheehan and Howell have described the strategic and economic SLOCs and outlined
their concerns regarding our capacity as the only superpower in town to effect the ability to maintain
freedom of the seas and rights of passage in the face of a real or perceived mine threats.
I would like to share some perceptions and conclusions I came to regarding mine warfare and
its effect on the battlespace and overall campaign planning during my tenure as C7F and as the Naval
Component Commander during Desert Shield/Desert Storm.
As early as 1990, it became apparent to us that the Iraqis had begun a massive defensive
mining operation in the northern Persian Gulf. We were seeing that the minelayers were going to sea
every night and coming back every day. And we knew they were popping in somewhere between 40
and 80 mines each night. Before we took action, Iraq had laid 2,500 more mines. This proves once
again what RADM John Pearson has emphasized — that the best countermine operation is to destroy
the inventory at the source.
This Symposium has tackled the problem of how technology fits into the mine warfare
problem and how it can bring innovative and rapid solutions to a very, very difficult warfare area.
* .ADM Arthur was introduced by ADM Thomas B. Hayward, USN (Ret), former Chief of Naval
Operations and Honorary Chair of the Mine Warfare Association. ADM Hayward had been
introduced by RADM Charles (“Chuck”) Horne III, USN (Ret.).
1-35
We have made a lot of progress since the Iran/Iraq War and Desert Storm. The Navy’s recent
focus on mine warfare suggests it has gotten the message and is now placing a significant level of
effort into improving mine countermeasures (MCM) capabilities.
We now have 14 oceangoing mine countermeasures ships. These 1300-ton wooden vessels
are equipped with the most sophisticated combat weapons system in the world. The Avenger class
MCM is a fully equipped MCM ship capable of crossing the oceans on its own power and of
operating for up to 30 days without replenishment. These vessels were designed to counter the
modem mine threat. With capabilities to conduct mine hunting and minesweeping, both mechanical
and influence, these modem ships provide us with a far better capability than the MSO ships we had
in the Persian Gulf
The introduction of 12 coastal minehunters, the Osprey class MHCs, into the fleet is well
underway. The MHC is an 800-plus-ton vessel constmcted of glass-reinforced plastic (GRP). The
program is an example of what you can do with existing technology, in this case GRP technology
transferred from Italy. Please note that, in this case, the transfer of technology was positive, as
opposed to another transfer of technology from Italy to a not-so-ffiendly country, Iraq.
You heard RADM Conley tell you that he intends to deploy three MCMs and one of these
coastal minehunters to Denmark to participate in Exercise Blue Harrier — the largest mine warfare
exercise in the world.
During Desert Storm, we diverted an LPH (the U.S.S. Tripoli) from its primary amphibious
mission to serve as a support ship for the MH-53Es. This took away a valuable Marine Corps lift.
To provide command and control functions and a platform for airborne mine countermeasures
helicopters and to support mine warfare operations, the Navy has converted the U.S.S. Inchon into
a Mine Warfare Command ship. Inchon will carry an MCM Group Commander and his staff and will
provide support to surface, airborne and underwater MCM operations without degrading Marine
Corps amphibious lift capabilities.
To respond to the integration of mine warfare forces into fleet exercises and deployments, the
Navy has an aggressive fleet exercise program underway. Mine warfare is playing a prominent role
in Joint/Allied exercises. Our MCM crews are showing up and performing well. Our crews have
participated in Blue Harrier, Kernel Blitz, JWID ‘95, Foal Eagle, and other exercises.
As RADM John Pearson pointed out, the CINCs want to integrate mine warfare and mine
warfare forces into every work-up, with the MPSRONS and Marine ARG/MEU’s. In fact, the
demand for more mine warfare assets to participate in exercises and work-ups is far greater than
RADM Conley can yet provide.
The focus is now shifting to developing and providing an organic MCM capability in our
deployable forces. This capability is needed to find minefields at forward deployed areas with organic
systems. We can no longer wait days, weeks and even months to execute plans while MCM forces
transit to operating areas. A Remote Mine Hunting (RMS) System is now under contract to provide
the fleet with a mine reconnaissance capability. A prototype developed during FY ‘94 and
1-36
successfully demonstrated during Kernel Blitz ‘95 is now being deployed on board U.S.S. Cushing
with the U.S.S. Kitty Hawk battle group.
In order to improve response time to a fast-breaking crisis, four minehunters are now pre¬
positioned at overseas locations. Two are deployed in the Middle East and two more in the Far East.
But let’s move on to more advanced technological systems. A laser-based system will soon
be fielded to provide the Navy with a capability to quickly survey an area suspected of mines. The
two systems, ATD-111 and Magic Lantern, will have a fly-off competition in April 1997 to determine
the Navy buy. Both systems show promise of detecting underwater and partly buried mines,
particularly in the surf zone. This will provide us with a long needed rapid reconnaissance tool.
A program is underway to replace legacy systems and electronics with Navy standard
computers and work stations. The Integrated Combat Weapons Systems (ICWS) will reduce
infi-astructure costs while improving capabilities. The Program Executive Office for Mine Warfare
(PEO-MIW), RADM Williams, has placed the highest priority on reducing life cycle costs while
simultaneously seeking ways to improve mine warfare mission effectiveness. For instance, borrowing
techniques from commercial oflF-the-shelf equipment allows us to reduce the number of printed circuit
boards to 44 fi-om 744. And this is just the beginning. This approach will provide the much needed
improvements in reliability and maintainability. The estimates on the life cycle cost avoidance are
approximately $400 million for just the first two phases of the program. This program is targeted for
both MCMs and MHCs.
Over the years, computers and workstations have become more powerful at significantly
lower costs. The Integrated Combat Weapons Systems will take advantage of the latest available
computer and display technology to meet the emerging requirements of mine warfare ships.
Advances and improvements in other areas, such as shallow water MCM, minesweeping and
lane clearing, are underway both in industry and in Government labs. These systems will provide a
much needed capability to find minefields at forward deployed areas without the immediate need for
MCM ships. MCM ships will be needed to follow up with hunting and clearing operations.
We are far away from solving the majority of mine warfare problems that exist from deep
water, through the surf zone and up the beach, and through the entire battleground. There probably
vwll never be that “Silver Bullet” you’ve heard about to shoot the threat of mine warfare in the heart
and kill it. But I believe the powerful combination of our Defense labs and industry R&D can provide
us the technology to overcome the “show stopper” effect that mines can produce.
1-37
The harder part of coating that bullet with silver is maintaining the interest of Congress and
the budget masters that Mine Warfare is as important in the food chain as the SSN, CVX, new attack
fighters, and other more glamorous and ‘sexy’ programs. It is the task of the attendees at this
Symposium, whether active duty or industry, to get this message out. Because, if the greatest
fighting force the world has every known can’t put troops ashore during a conflict, it becomes a
blockade force with limited ability to carry out national policy or exert force in the name of freedom.
We have the capability. Go out and convince others that we need the proper funding to bring
the end products to fruition.
I know this has been a Navy water-oriented brief I don’t envy the ground pounders’
problem, with probably ten times the amount of mines to encounter with the addition of anti¬
personnel, trip wires and booby trap features, that don’t exist in our sea mines. But remember that
only one sea mine can kill 600 sailors in an instant — not a happy incident for a grandmother in Peoria
or a Commander-in-Chief looking for re-election. So, I am prejudiced, but I think I understand both
sides of the equation.
In closing, let me leave you with a simple way to remember how I view the mine problem.
It’s by the acronym PIMSA. P is for Prevention — an ounce of prevention is worth a pound of cure,
or, in this case, weeks of searching and sweeping. I is for Intelligence — Essential Elements of
Information (EEI). We need to know numbers, types, intentions, storage locations, methods of
deployment, etc. M is for Mapping. We must be able to precisely locate not only the fields but also
the disposition of various types of mines within those fields. S is for Swiftly Sweep/Neutralize only
what is essential to accomplish the task. And A is for: And when all else fails, remember the words
of a wise and masterful leader in our earlier encounters with mines — “Damn the torpedoes (mines)!
Full speed ahead!”
Again, I want to thank the Mine Warfare Association and the Naval Postgraduate School for
the invitation to speak tonight, and Admiral Thomas Hayward for his kind words of introduction.
Full speed ahead — and good night.
1-38
Technology and the Mine Problem:
An Evolutionary Revolution
I. INTRODUCTION
With the end of the Cold War and the lessening of the threat
of instantaneous annihilation from nuclear attack, new sets of
military and terrorist problems re-emerge. Prominent among these is
the problem of mines, a problem - as was demonstrated in Kuwait -
that required no time to come to the forefront. The fact is that
mines have always been a weapon of choice for technologically and
materially inferior groups. The 125-175 million land mines
currently in place around the world attest to their popularity with
rogue groups and insurgent forces as well as with conventional
military establishments.
1-40
OPERATION STARVATION, coupled with the anti-shipping campaign
conducted by American submarines brought the import of strategic
materials virtually to a halt before Hiroshima.
Land Mines. Much of what has been said about sea mines applies
also to land mines. The main differences are the orders of
magnitude differences in numbers of land mines and their relative
ease of emplacement. The sources of doctrine for the use of mines
are Anglo-American, German, and Russian (former Soviet Union).
1-42
dsterrtiined groups ~ include barbed and concertina wire (often
sprinkled with anti-personnel mines and grenades), hedgehogs and
welded iron tetrahedrons, primitive Jersey Barriers (weighing up to
4000 pounds), and cement blocks that may weigh up to 10,000 pounds.
The practice is to intersperse such barriers with anti-personnel
mines so that engineer sappers cannot emplace destructive charges
or use wire-cutting devices. The organization of the beach defenses
at Kuwait by the Iraqi show how quickly extensive and formidable
defensive positions can be organized.
What must be done, where the activities take place, and the
1-43
military or civilian context define mission, environmental, and
contextual niches or classifications for the principal uses of MCM
hardware. These are the bins of the MCM "Tool Box". Those niches
subdivide the time domains of present, near term, mid-term, and far
term. Introduction of the time domains immediately suggests that
the system-of-systems approach to MCM is an evolutionary one -
albeit with some revolutionary elements conferred by emerging
technologies.
a. Contexts.
1. Military combat. The requirement is to bring the risks to
an "acceptable" level in a timely manner. The terms acceptable and
timely are related and are military situation dependent. This
contexts is often associated with amphibious assault or minefield
breaching in the land-warfare context.
2. Military administrative. Clearance of relatively limited
areas for logistics support or administrative use but in an overall
1-44
hostile environment. Assurance of minimal risk is the desired
criterion.
3. Civilian administrative (Humanitarian Demining). This
differs from Military administrative in that a high degree of
Assuranc© that an Area is MINE FREE is reguired over potentially
huge areas. Furthermore the measures of effectiveness include
economic terms as well as suitability for employment by minimally
trained indigenous personnel. And the measures are also in terms of
human suffering, the killing and maiming of innocent civilians,
often women and children. Anti-personnel mines are designed to maim
and maim they do.
1-45
F
Naval Domains:
1. Deep Ocean (<100 Fathoms)
2. Shallow Water (40 - 600 feet)
3. Very Shallow Water (10 - 40 feet)
4. Surf Zone
5. Estuarine
6. Riverine Including lalces)
Because of the physics that govern sensor performance, the above
must also be categorized with respect to presence of man-made or
natural mine-li)ce objects, salinity, turbidity, electrical
conductivity, and possibly other characteristics such as the
presence/absence of biological and botanical organisms and plant
growth.
Land Domains^ (Subdivided by Presence/Absence of Vegetative
Cover (trees, bushes, of grasses))
7. Rocky
8. Sedimentary
9. Sand (Dry Desert)
10. Sand (Moist/Beach)
11. ice and snow cover
As with the naval domains, the land domains must be grouped by
similarities in the physical properties that determine sensor
performance.
The above defines the bins for the tools of mine countermeasures.
Next, a motivation for having a taxonomy is provided.
1-46
Potential Utility of an MCM Taxonomy in System Development and in
MCM Operations
1-47
importantly, enabling the enemy to establish "killing zones" is
supposedly weak or unmined areas (traps). In this scenario, early
defense suppression by strike air, missiles, and naval gunfire are
very important precursors to any attempt to circumvent of breach
the minefield. (The vital role in mine countermeasures operations
that is played by elements capable of offensive power projection is
rarely explicitly called out. It should be as it was during the
Normandy Invasion). The composition of the striking forces and the
weapons load-out for these forces must also reflect the defense
suppression and objective areas preparation requirements.
There are roles and tasks for specifically configured and dedicated
MCM assets in this reconnaissance phase. Here lie some of the
opportunities for MCM technologies to augment human resources and
minimize the risks to those resources. This aspect will be covered
more fully in a later section of this paper.
1-48
IV. Emerging Technologies Supporting the Mine Warfare Paradigm
Shifts and "Evolutionary Revolution"
Emergent technologies are enabling the paradigm shifts in Mine
Warfare. These technologies contribute to the intensification of
the threats from mines through the explosive proliferation and
availability of sophisticated mine mechanisms. Technology also
offers the potential for biological, chemical, and radiological
warheads. Already available to rogue groups are land and sea
versions of mines capable of effective damage at stand-off ranges -
the so-called "off road" mines that can be thought of as land
versions of our CAPTOR mine.
The approach for the rest of this Section of the Paper and for
the following section is to progress from the general to the
specific while attempting to minimize redundancy.
1-49
"separateness" is the widespread combination of ignorance of and
distrust in the mine countermeasures forces.
1-50
military operations and activities where discipline and absence of
unplanned events are the rule. However the technologies of
communications, of systems control, and of vehicular design come
together to promise generations of mine countermeasures systems
that demonstrate a spectrum of degrees of supervised autonomy.
1-51
C. Dealing with the Mines and Obstacles. In operational
maneuver from the sea as well as in breaching operations in land
warfare the objective is to is to pass through the mined areas as
quickly and safely as possible. The term "in stride" is used to
signify that an objective is to breach the minefields and/or land
the landing force without slowing the troop and equipment-carrying
vehicles. Clearly, it is most desirable to "go where the mines
aren't" - thus the emphasis on surveillance and reconnaissance. If
that is not possible then the mines and obstacles must be
neutralized or removed.
The U.S. Army and the U.S. Marine Corps have or are developing
families of rakes and plows that can be attached to a variety of
armored vehicles and engineer equipment. One, the JT^C, is
advertised to have some capability against light obstacles but has
not been tested against the same stresses that the CLAUSEN POWER
BLADE has (October, 1995, at Camp Pendleton, CA).
1-52
scope in emergent modern Mine Warfare.
1-53
cellular telephones are finding their way into such emergent mine
countermeasures surveillance and reconnaissance systems as the
Autonomous Ocean Network. The C4I systems that permit the
establishing of virtual environments are outgrowths on internet
technologies.
1=54
countermeasures autonomous brigade.
Tactical Decision Aids and C4I - New Tools for the Force
Commander. The stereotype or mine countermeasures as repeated
passes of minesweepers through an area is as outmoded as the
"grease pencil" surface plot. Soon the commander will "see" the
objective areas through the fusion of data and information from
both technological and human sensing nodes. Mine countermeasures
C4I has already demonstrated the ability to mirror events off the
West Coast in the Gulf of Mexico. Such training and evaluation
activities are by-products of the major efforts in distributed
modeling and simulation that have been supported by the Navy and
the Department of Defense.
1-55
The Autonomous MCM Brigade consists of 3 regiments; a land warfare
regiment, a naval warfare regiment, and a land civilian-
humanitarian regiment. Each regiment has organic air/space
squadrons attached. Each type of regiment has appropriate human
operated and staffed logistics, maintenance, and operations support
personnel. Focus on the hardware organization supported by people
rather than the conventional way of describing military
organizations in terms of their personnel is intentional as we wish
to emphasize the potential operational roles of the autonomous
hardware components.
Readers will note that most of the current land warfare mine
countermeasures equipment either is or could be configured for
combinations of TV scanning and radio control. So could much of the
equipment that could be applied to the humanitarian demining
mission. The problem there is that cost factors force demining
operations into using large numbers of unskilled personnel to
conduct mine neutralization and area sanitization operations by
hand. Things are less well developed in the naval environments. The
Mine Neutralization System is a tethered multi-sensor system
operated from the major mine countermeasures platforms. Surface
units and aircraft can be teleoperated and radio controlled.
Underwater vehicles resembling torpedoes can (and are) programmed
to run pre-determined courses and are useful in oceanographic data
collection. At present these vehicles do not have hover
capabilities.
Whether for land or sea use, the companies (or battalions) of the
robotic regiments might be organized to fill specific environmental
niches. Greater operational flexibility will be conferred if the
sensor/mission packages can be modular so that each vehicle can be
efficiently outfitted to perform in the environment at hand. There
are competing design approaches. There is a trade off between cost
and multiple-capability in a single vehicle. The other extreme is
to have a hierarchy of vehicles with each level having greater
sensor or mission package capability. It was this latter concept
that was envisioned in the introduction to the Autonomous MCM
Brigade at the Symposium.
Today, robotic vehicles run the gamut in size from those the size
of a cigar box to giant walking machines such as DANTE II. To fix
ideas, most of the members of the conceptual Autonomous MCM
regiments will be sized between a Standard Gauge Model Train car
and a small self-propelled lawnmower. A design principle is to have
a total system that degrades gracefully with operational losses
rather than catastrophic systems failure that can occur when an
irreplaceable unit is lost, costs run from millions of dollars at
the high end to as low as 2-5000 dollars for single purpose
vehicles (the low end of the hierarchy).
1-56
SUGGESTED READING AND REFERENCE
1-57
1-58
CHAPTER 2: OPERATIONAL
REQUIREMENTS AND PERSPECTIVES
The Invited Papers in this Chapter are by Senior Military Commanders from the Army, Navy,
Marine Corps and Air Force. These papers complement and amplify the Keynote Address by General
John J. Sheehan, USMC. Taken together, the papers in Chapters 1 and 2 summarize the needs for
operational capabilities.
The ensuing Chapters provide the technical responses to those stated needs.
2-1
U.S. Army Initiatives in Mine Warfare
2-3
INTRODUCTION
what we had many years ago-- “You mean after all the money
we’ve spent, all we’ve really got are probes and coin detectors?” I
read your vision statement for this symposium. I’m gratified and
probes and coin detectors. What I want you ladies and gentlemen
2-4
container filled with explosives in the earth. I’m thoroughly tired
the landmine will wither away. I know that finding this “vaccine”
today, but I’m convinced that it can. It’s my job to frame the
requirements for the Army’s needs and to work closely with our
the academic community has let us down! Over the years, we have
2-5
which means achieving victory and saving lives. You must
and I have not seen anything, in my time, that would indicate real
frustrated.
countermine lead is the Navy. The U.S. Navy does a fine job of
training EOD specialists from all the services, but the Navy’s got a
land countermine than the services who stand and fight upon the
operational aspects of land and naval mine warfare and to look for
2-6
a single set of solutions. This distinction is acknowledged in the
lane and will only highlight some of the differences as they apply
to the land environment, and, indulge me with just a bit of the surf
2-7
know about sea mines is that they are large and a single mine
States and the death of many sailors. A single sea mine can have a
When the Navy gets the mission to open a sea lane, that is
“missing one.” The consequences of a mine strike for the Navy can
every mine, every time. For the Army, the loss is usually a
remove the bulk of the danger -- which may actually leave some
2-8
mines. Unlike naval mines, land mines are usually covered by fire
and are only one component of a complex obstacle set. Since the
500 UXOs for a single fire mission! These are unstable UXOs that
2-9
Considering just the land mines, there are two basic types;
anti-tank mines are similar; the mines are larger, easier to detect
obstacle - and they do a great job. They are small, well-hidden and
countermine tasks, which you will hear about during the week.
2-10
detection operation under combat conditions. Usually it is night
up, someone may shoot him, and, of course, he must carry the
and 105 pounds-- the old Army joke is that it’s one hundred pounds
this operational scenario, consider the fact that that the detector
does not find all the mines and has a high false alarm rate that can
hear all signals. Sometimes when the detector does find a mine,
2-11
may start to miss signals. We’re asking a lot of this youngster! If
this trade.
crew, and they search for larger mines that generate more
HUMANITARIAN DEMINING
sea. Professional land forces account for their mines and are bound
2-12
hazard they caused; losing forces, however, often fail to do so.
for this symposium was to: match technologies and systems with
2-13
COUNTERMINE TECHNOLOGICAL SOLUTIONS
since 30% of the mines are not detected. A suggestion is to put the
2-14
Medal of Honor winner will acknowledge fear, but a caring
false alarm rate, but how does that convert to a rate of advance?
For this to work we have to get on the same team. Not just
talk about it, but really do it. You need to have a fundamental,
2-15
For example, probability of detection is really meaningless. I don’t
need to know how many mines you found, I need to know how
many you left. I live in a world that quantifies and deals with
language.
so must you. You know your equipment better than anyone and
2-16
After the performance parameters are translated
30% of the mines and must stop every 7.5m to investigate a signal.
Soldiers’ problem?
CHALLENGES
2-17
standard should be your own willingness to take your equipment
who has to walk through the minefield. Toward this end, I have
identified four specific areas where you can focus your attentions:
Employment Concepts:
way. You know, it is a lot safer and more effective than any of the
2-18
detectors in the inventory. I am a strong advocate for the
the size of the item it’s detecting. You know the most about your
Soldiers.
Sensor Fusion:
has intriguing possibilities, but I have not seen much in the form of
complicated the job of the Soldier. They have relied on the neural
2-19
decipher two differing tones and monitor a heads-up display. The
In the very worst cases, the “fusion” has been a simple link that
alarm rate.
there, you need to consider what is the best way to deliver that
visual signals are more definitive, but they require the soldier to
take his eyes off of the ground. There is a lot of work to be done
in this area, but it has a lot of promise. Sensors that are actually
Humanitarian Demining
2-20
humanitarian demining, you need to know that US policy
success.
2-21
The fourth area, test beds, is a very disturbing issue for me.
Summary:
2-22
understand that materiel is only one piece of the puzzle.
2-23
I would like to take a minute to address a related issue. All
aimed at reducing the residual risk left by land mines and in the
2-24
munitions that have organic self-destruct or self-neutralizing
Perhaps you’ve never really thought of this, but the fact that
2-25
large numbers: unfortunately the dumb mines create a residual
I’ve spoken for about thirty minutes and I’m still frustrated.
I know this has been a blunt speech - it has been tough to deliver.
anxious for a product. I cannot ignore the fact that our Soldiers,
Marines, Airmen and Sailors are still putting their lives on the line
2-26
problem, there is no corresponding “mine solution.” There may be
that and forge a strong team, dedicated and resourced in the same
2-27
2-28
An Entirely New Approach
to the Countermine Mission
Countermine has become the challenge of the day for the US military and for
the world. For too long, too little attention has been paid to this critical battlefield
function. Now we are seeing the effects of that neglect, not only in the materiel field, but
in everything from a basic problem definition, to force structure implications to a
discipline way to approach the challenge, to the terms used to measure success. This
briefing will address some of the cutting edge ideas we are exploring to fill these gaps. I
bop^ you will embrace this novel approach and use it as your frame of reference for the
remainder of the conference.
With the end of the Cold War, the Army has come to the very real conclusion
that countermine has very distinct components. There two primary types of missions.
Combat countermine is well understood, as are the terms associated with it. The lexicon
really falls short for defining the “other” operation. Terms range from Operations other
than War, to Security and Stability Operations. For the purposes of this presentation, I
will call them Contingency Operation. These refer to the missions that do not have the
same lethal environments as combat and therefore lack the stringent speed requirements.
The solutions that meet one mission do not automatically fit the other. For example the
requirements for an Ml based breaching vehicle lead us to the Grizzly and the ESMB (a
rocket propelled explosive breacher). They are clearly designed for a very narrow
mission that does not carry over to the CONOPS arena. This has left operational voids in
the missions we are finding the most prevalent. The POM has started to recognize this
reality and I find it quite interesting that it does not focus on high intensity conflicts
anymore.
Within the Army, the Training and Doctrine Command is responsible for
defining and formalizing user requirements. The Army has instituted a new process for
generating operational requirements. TRADOC is now defining requirements by
mission areas. The branch Commandants form Integrated Concept Teams which look at
requirements from a holistic perspective - across all of the \Army mission areas. The
TRADOC Commander has legitimized this ICT concept by stating that all requirements
must be generated or validated by an ICT before he will accept them. As the
Commandant of the Engineer School, I have stood up several ICTs, one of them is a
countermine ICT.
2-29
We have gone to great pains to define a base level countermine capability - what
It IS and what it should be. Clearly every unit must have some organic CM capability, we
feel that extends beyond self-protection. The entire military has responsibility for the
correct implementation of CM doctrine and use of equipment. When a unit encounters a
mine threat, their training, doctrine, and equipment must be employed in response. If
the risk is too high they must call for additional CM assets, and there needs to be logical
progression in the request protocols. Having the force equipped with sufficient, properly
equipped CM response units is proving a challenge.
The problem is bigger than just countermine, it is operations in a mined
environment. Mines are a condition of the battlefield and, as such, they influence
everything we do. This is not an Engineer problem with Engineer solutions. Every
soldier has the potential to step on a mine, every mission has the potential to be
complicated by mines, and every piece of equipment has the possibility of encountering
mines.
This briefing will highlight the challenges we face and will continue to face in
this arena. I will highlight the significant differences between the two countermine
operations, combat and CONOPs, and show that they have drastically different
requirements. I will more fully explain the components of the countermine mission and
finally I will address our vision to more effectively conduct, monitor and evaluate these
missions. If we are going to succeed, we need to find and use all of the right tools for the
jobs; each job.
The military is coming to grips with the new nature of conflict. In the
countermine arena, the differences between combat and contingency operations are
significant. I will highlight, from an operational commander’s perspective, those key
distinctions:
Combat
The combat countermine mission revolves around speed and mobility and the
basic components of this mission have not changed much since World War II. The
maneuver commander is trying to accomplish his tactical or operational mission and
mines are a hindrance. He wants to know where they are, but he doesn’t necessarily
want to encounter them!
In combat, the commander has many sources of danger to deal with, in addition
to the mines, such as lethal fires and fratricide. A commander must make choices that
minimize losses and ensure mission accomplishment - sometimes that means being
willing to accept a less than perfect breach. The commander must weigh the losses he
expects to take crossing a minefield vs. the losses he will take standing still or waiting.
Breach methods are fast, violent, destructive, less than complete, but suited to a high
paced, combat mission.
Contingency Operations
Countermine CONOPs represent a relatively new mission for the Army - at least
in these numbers. Unlike combat operations which are complicated by mines; in
CONOPs mines sometimes are the mission! The job is to find them - every single one of
them, and neutralize them. As if the problem were not complicated enough, many of the
areas are cluttered with UXOs. In a CONOPs situation, they pose a very significant
challenge since the mission is to deal with each explosive individually. The standards for
clearance are unbelievably high both quantitatively and morally. These areas are
potentially future playgrounds for children. These operations are subject to US policy
2-30
restrictions which do not allow US soldiers to actually go into the minefields. Our roles
are supervisory in nature. We track the minefields, train the trainers, collect intelligence,
conduct macro-detection operations, mark dangerous areas, conduct mine awareness
training and assist in proofing, to name a few. As you can imagine, collateral damage is
a consideration and so the typical combat explosive breach techniques are not usually
appropriate.
CONOPs are typically being conducted under the close scrutiny of the State
Department, the UN, the American public, CNN or some combination of the above. This
gives risk management an unusual qualitative spin. Military officers have been trained to
make the hard, but quantitative, judgments regarding mission accomplishment and
soldier harm. Now a humanitarian mission, that results in a single casualty can be termed
a failure due to public scrutiny and media spin. These intangibles have changed the
complexion of the CONOPs countermine mission more than anything else, because they
have changed the standards. Although not specifically stated anywhere, the unwritten
standard for CONOPs countermine has become “zero-casualties.” This poses some
extraordinary material and operational challenges.
The two scenarios define the settings in which countermine operations take
place, but countermine is not an isolated task, it has many components. The countermine
ICT has developed nine mission categories, which is a significant departure from the
tasks defined in the previous Army Countermine Modernization Plan. A few of these
represent new missions and all of the previous missions now have a broader scope.
Finally, the definition and implementation of each of these sub-tasks will be tailored to
the specific mission.
I will now briefly review the nine countermine mission areas and highlight their
differentts as they apply to Combat and CONOPs missions.
2-31
a. Combat: The basic question that the maneuver commander asks is “Where
can I maneuver most freely?” , i.e. Where are the mines not? The speed of the maneuver
force is paramount so fast, accurate detection of mined areas or minefield boundaries
drive the requirements train.
b. CONOPS: The “customer” for these markings are the non-combatants and
the deminers. The markings need to be obtuse: visible and understandable by all. One of
the interesting problems has been the selection of marking material. These poor
countries tend to scavenge anything of value and even simple wooden pickets don’t
survive.
Breaching:
b. CONOPS: Largely does not occur. Concerns regarding civilian safety and
collateral damage tend to override this. The minor exception would be self-extraction.
This would fall into this category because it is the one CONOPs countermine mission
where speed dominates the mission success criteria.
Clearing
The scope of the problem is expanded by the legal requirement to conduct Battle
Area Clearance (B AC). In accordance with international law, the emplacing force must
restore a country to normal by removing the threat from the land contaminated by mines,
submunitions, unexploded ordnance, ammunition, missile fuels, weapons and other
hazardous debris.
2-32
involves the very time consuming task of examining every inch of ground. This is made
more frustrating by the policy restrictions that forbid soldiers to enter the minefields. We
are limited by the motivation and capabilities of the indigenous population. Their
education and competency levels usually fall well below that of the average soldier.
Right now we proceed at the rate of PSS-12 and sometimes probes. Solutions that speed
up the process seem to either compromise detection rates or are too high tech to be
feasible.
Protecting
a. Combat: Protection includes the force, the unit and the individual soldier.
Our concept for protection links into intel which warns forces of mine threats. At the
unit level it incorporates mine survivability in to all vehicle, weapon systems and
building designs. Some of these changes are relatively simple, like not designing
passenger seating over tires, or extending the wheel base so that the 60 degree blast cone
doesn’t take out the driver’s legs. This also includes basis of issue items, such as placing
rollers on scout vehicles. At the individual soldier level, we are investigating the optimal
mix of protective equipment that does not encumber the mission.
Neutralization
Neutralization is not a term that clearly understood - or rather everyone has their
own definition. It is the act of making an individual mine safe. Here mission focus and
success are defined in terms of the individual as opposed to clearance where the standard
of measure is an area.
a. Combat: Pat yourselves on the back - this may be one we have mastered.
Assuming the mine has been accurately detected, located and isolated; neutralizing it is a
relatively simple task and our kit bag is full. It is, of course, a large assumption to think
that we can accurately detect, locate and isolate all mines. More often than not,
technology limits us from large area neutralization. In an effort to minimize risk, we are
looking at methods to neutralizes mines remotely.
Irainipg
2-33
prior to deployment and updated/reviewed once in theater. Beyond this, though, leaders
need to conceptualize mines as a threatening condition of the battlefield. The NBC
model is very comparative. Leaders need to learn how to ask and answer the right
questions regarding the mine threat - using maneuver language. The goal is to make the
presence of mines relevant so that leaders account for them in COA development and
selection.
-Demining:
b. CONOPS: Many of the issues that define this mission have already been
touched upon. The technical difficulties of meeting such high standards, coupled with
the cultural challenges place this one on the top of the “too hard” list. It is important to
mention that the military already performs this mission in a limited capacity. Right now,
the Special Operations Forces have the mission. The Army is exploring the best way to
expand their support of humanitarian demining. Many of the lessons we have learned
while performing CONOPs are proving valuable. International pressure and current
administration goals are great motivators to expand this mission, but the Engineers are a
limited asset. If the Army takes on this mission, there will have to be some
compensatory resourcing, primarily in the form of force structure increases.
CLOSING THOUGHTS:
- Take off your blinders. Your job is much more than making a better
mousetrap.
- The mission has much expanded horizontally too. There are nine sub-tasks -
not just detect and neutralize.
- We need to speak each others’ languages (a dual challenge). Define the value
added of your system in operational terms. Pd is dry, sterile and essentially meaningless.
2-34
UNITED STATES ARMY ENGINEER CENTER
2-35
ESSAYONS
D
H
(0
UNITED STATES ARMY ENGINEER CENTER
2-36
UNITED STATES ARMY ENGINEER CENTER
2-37
ESSAYONS
UNITED STATES ARMY ENGINEER CENTER
2-38
ESSAYONS _ "Let US Try'
COUNTERMINE MISSION AREAS
2-39
bO
bD C
• rH
cb
• ^
cn
N o 3
COUNTERMINE MISSION AREAS
cd O
(D H
O
c (D
c3
C
a
bX)
bJO
a
<0
o a B z
o
in
« s
I rH
<
CO
CO
bD O m
"o
o C!
4h '5
cd -S
td
o
C 'S
czi
CO
(D w ^
O ^ D
O ^
H CO cd
g .rH
P-io
’ I 1 ^
_S ^
cd cd
z
3
2-40
SCENARIO DIFFERENCES
Minefield C4I
H
CQ
U
O
.
H
»
2-41
PL4
o
(/)
o
o
Z
C/i
a
</>
Z3
4 -«
0)
(0
z
I
0)
(0
lU
UNITED STATES ARMY ENGINEER CENTER
u t>0 bO
bO
a
Demining
T3 O •T-!
c N
bJD bJD d
• ym^_ b/J
u 2 C VO
c:
*r^ o • 1^
(D o (D H a
rH
v>-< U4 Co C3 -t-o D • r—I
• rH 0) <D O c3
iD 't-O
S Q S U CX; H
2-42
UNITED STATES ARMY ENGINEER CENTER
2-43
ESSAYONS _ "Let jjs Try
o
<D
tdo
2g -S o
H
cn
3
COUNTERMINE MISSION AREAS
4-rf
■H s-i <D 0)
4-»
«3 O 3
+3 o
c3
;3 c/5
^ 1) on
.S G
ESSAYONS
^ O
73
o
£7, ^P o
^ T-
• S =3 c/5
ao a s O
o
(D
c/5
o a
C/5 o
fi
< CU B..a
2-44
I
"Let uS Try
C/)
ESSAYONS _
O fiM
o
CX) ©
u
"3 04
a £
u ©
C/2 a
nd xfiCQ
©
u ©
a
UNITED STATES ARMY ENGINEER CENTER
B P
o ©
0) ©K
CD 1
HI cd ©
fi
o
ed u
©
&
•
2-45
C/)
a o H
a (/)
COUNTERMINE MISSION AREAS
w <u
D O
T3 ^ 'Q
<o
^ B u
o X) <D
O o
Io a
ESSAYONS
o
cd
o
<D 2 o
‘r? ^
4->
(D
SI <D
o
• rH
a O ^ (D
c/3 X
O
a H <D
o
O ■rt ^
a.
. cd <D
c/3 <D "
3 S
C *
I-H
bX)
u
P
•PX bJ3
G
• r-1
12 G N bi)
w) bO .^ bX)
'i/ .2 "G «G
rH
Clearin
Trainin
4 > OJ 4- > ■g
3 4d 1 G
• —
r~( o Ul o 0)
5- 4 CD
Q n 0. Q
2-46
"Let US Try
Sa>
ESSAYONS _
0)
00
W ;-i
U **C5
a
I
tq c/2
;-i
Pt
fe ^
I
g
o
T5
o
Q ^ o;
UNITED STATES ARMY ENGINEER CENTER
(=U ©
O ^
y CP
c/2 a
c/2
©
a>
0^
o
H A
/2
c uy
-<55 Ph ©
£ © O (DD
u ©
©u ^
©
Z
o
;h
00 o hJ
u • u
2-47
V-
o o
> h-
0)
COUNTERMINE MISSION AREAS
3
o
cd o
O
US ^
(D
ESSAYONS
T5
cd
a
(D
a
4-> “
^ (D
CD O O
r? ^ o
H o o
2-48
"Let US Try
c/} (/)
ESSAYONS _
00 O
• pM
o
• pM
m cd
u ;-i
I
w
O
.2
o
u
o
C4H
a
^
u
tin c
' t-H
c;
0^
a>
S WD
a> u
Q a P iS
<D
ss ^
UNITED STATES ARMY ENGINEER CENTER
(/J
CQ
0^
c«
so
<
Sx
CS CS
H 05' C> o
:zi pS
'o Ph
w P3 o
u z S3 ^2
00 O o o O S
U • u
4^ (D
•1-H ^
4h ^
b-
(0
O 3
a
a ?)
(D
s
C/J T3
CD a
a *o
« 8 a
• i-H
*C
O
cd
CN
<u too
^ ’T3 CD "o (0
• t«h ■4—> z
Dh ^ • rH o
cd
<
2 S fi
cd
u CD
>
(0
CO
UJ
g o «+H
a
•
O
o O
B c *r^
o
O
T3
•tH
'g
?—(
CD c/5 > D
•^ 4->
o
UNITED STATES ARMY ENGINEER CENTER
(U > CD O
CD O
00 .5 c/:) T5
a
•^ Oh
bX) b^)
u fl S3
bJ) •»—(
mining
2 c N
o bG a bO • 1—1 bX)
-immJ
0^ 13 G
(D O o 4-^ V-
C
CD
CD
a
CD CD
o ccj
a> <D
Q CQ U CLh :z; Q
2-50
I'Let uS Try
_
00 42'' Isi
W O
1/5 S s
U ^
ESSAYONS
s 2
§ •! ^
S3 5 cs ^
g Cm *3 *=^
tin *=i O g ©
Ml cfag gS
^ O
O
is ^
4^ si
UNITED STATES ARMY ENGINEER CENTER
O 5 Cm Si M
g ^
^
'M M c2 4
© ’g g 'tt
s 2 S
a>
ts
g S3
Si
'M §
o Q*
S O
£U £ .1
IS 2-51
bb
•S
u
COUNTERMINE MISSION AREAS
: rH
(D
<D o
<D
c/3
41!
ESSAYONS
(D
.S
a • rH
T3
(1)
a
0)
OX)
xl- c
u
• pm^
ci bX bX bX
OX C
0) o c • r~~i( bX .s 13 bX
r- 4—> p s
4r~s
^ r-G « V—< '5:
Q o O *
* r-(
<D Vh (D
C
<D u cd CCj +-> • T-**^ a
+-J >—•
CD
CD CD p d CD
r—i
G u O.: H Q
2-52
"Let US Try
CiD
ed
scd
ESSAYONS _
00
u Id
u
o
S s iS
o
pH N
HH
]p
ed
Q s
UNITED STATES ARMY ENGINEER CENTER
9 I o
12
PU O
2
c/^
pH
I/)
•N
>> -e
^
§ o QJ (/5
u %
00 o 2
^
O
O
2-53
<u
O CO 00
CO
cP ^
C3 5h 7t
,P C P
^ T! hh Qh ^ o
-
fO (D
- CL) ^
2^ a.
S o c s ^ O
4=1 ■!= o O
ESSAYONS
^ ^ ‘53
00
a .5 "d
”
(D
d ^ d
CO
CO
<D . , a -r fl
y
2 2^ 2
CO 7!^ >
0 a>> ^p
« &•-
■g
>^
£ g S3
UNITED STATES ARMY ENGINEER CENTER
d ^ rH
H Ci^ cd <D 713 .S
U
'Td bS)
r—H
QJ o ♦c
r--i(
U~‘
H—i sz!
o
(D
c <D
4-J
<D
'I
CD
o P
2-54
UNITED STATES ARMY ENGINEER CENTER
2-55
ESSAYONS
c/2 >
Si.
H
CO
0) (/)
c/2
C/2 O 3
<D
cd ^
c/2 C/2 Iz; lu
cd
c
<D O O CO*'
u fi
O
cd O
c3 O O c/2
ESSAYONS
c3 > > c2 o (]J
o ^
bJO ^ a
(D
^=3 ^ CO
o a c/2 ‘rH
a
o
o O
I ^ (D o
GO I>
a
r\
'4—> (D
O (U a o
CO
4-> bX)
a c:5 • tH
o
a ;:3 'c3 >
C S ^3
UNITED STATES ARMY ENGINEER CENTER
o o
a
C
O r'l
S
u o CJ ^ §
2-56
Ml sn
(DD
cd • fl
PN
a
03
3
o
is
SNOAVSSa
1
CA
& S3
o
13 ts
tif) •S s-l
C
rH
*3 (/}
a o ;h
rH
1 o
UNITED STATES ARMY ENGINEER CENTER
B
<D C/}
'o
03 1
(/}
a •pM
HI cd
-1-A DiD
Ph CZl 3 fl
0^
o IS(OX)
•pM
o
O W -3
O
2-57
UNITED STATES ARMY ENGINEER CENTER
2-58
H
L.
UNITED STATES ARMY ENGINEER CENTER
2-59
ESSAYONS _ -Let US Try
UNITED STATES ARMY ENGINEER CENTER
2-60
<D ^3
o cr <0
D
♦-»
0)
<2 "
Vh
o
c
s
o s
C/)
<0
o Vh (Z5
z
o
o (U <
C/D CO
o (0
> UJ
cd
o
CZ) T3
o (U
C/)
(D
(D
CZ) Ph
(D cd CZD
UNITED STATES ARMY ENGINEER CENTER
ci
;h
O (N
o 4h
<u
u
o Vh
c/^
(D "o
.s bJO ••
a fi
• rH
•g
Vh
• tH
fl
u ri::) I'
C3
Vh © W
O c5 VO
o
u
<o
0) o
Ph cd
2-61
UNITED STATES ARMY ENGINEER CENTER
2-62
UXO removed 50m each side
UNITED STATES ARMY ENGINEER CENTER
2-63
ESSAYONS
Bosnia Bureaucratic Obstacles
CO
Z)
o
(0
z
I
(0
CO
u
<D
c/5
a
00
(D
Q
o
& bO
<D C
o(D • rH
00
I Ph
2-64
UNITED STATES ARMY ENGINEER CENTER
2-65 (1/2)
ESSAYONS
Bosnia Lessons
2-65 (2/2)
UNITED STATES ARMY ENGINEER CENTER
2-66
■ US ■ Canada
UNITED STATES ARMY ENGINEER CENTER
2-67
UNITED STATES ARMY ENGINEER CENTER
ESSAYONS
UNITED STATES ARMY ENGINEER CENTER
2-70
UNITED STATES ARMY ENGINEER CENTER
2-71
ESSAYONS _ "Let uS Try
UNITED STATES ARMY ENGINEER CENTER
2-72
ESSAYONS
e ob
^ tq
o ^
I ^ S tt) ^
bj g§
III gll g p^ g
1^ Kl
3
rtj fc:;
O ^ 3S g|i H
<0
rD
bj
O)
fcl
b P
b^kjct- fe,< (5<
UNITED STATES ARMY ENGINEER CENTER
2-73
This Page Intentionally Left Blank
2-74
UNITED STATES ARMY ENGINEER CENTER
2-75
z
(0
lU
<0
5cn
o
3
<0
0
> V
cn
3
4«i
<D
a
c3 o
• rH
o JO C/)
o
ao CZ5
•1-H
(0
z
o
o
o
a <
(0
(0
liJ
§ cd
X)
4->
GO
JIJ
<D
O <u
a €N
a (/5
UNITED STATES ARMY ENGINEER CENTER
r\
No “ Silver Bullet
"o
c/5
• tH
o
HH
<D d-)
CZ)
.S • rH
a (D
N
00
a.
O
o
13 O (D • 1-H
<-i 0
0 d-) 0 0 (D
u a u U GO
• • • •
2-76
UNITED STATES ARMY ENGINEER CENTER
2-77
2-78
U.S. Air Force Roles in Mine Warfare
OVERVIEW
TODAY.
SAME.
2-79
• AFTERWARDS, I’LL EXAMINE SOME OF THE ADVANTAGES AND
HISTORY
CONFLICTS?
(DST) MINES.
EQUIPMENT.
2-81
• DURING THE PERSIAN GULF WAR, THE IRAQIS LAID MORE THAN
DAMAGED.
MISSION EMPHASIS.
2-82
• NOW, THAT SINGULAR THREAT HAS CONSIDERABLY DIMINISHED
SAME:
PORTS OR WATERWAYS.
5,000 MILES
2-83
PRECISION STRIKE CAPABILITY—UTILIZING THE GLOBAL
POSITIONING SYSTEM
• LARGEPAYLOAD
HOURS
2-84
training and to familiarize each service with the OTHERS’
IMPENETRABLE BARRIER.
2-85
• SELF-DESTRUCT DEVICES MUST BE USED WHENEVER
CLEARED A MINEFIELD.
PREVIOUS CONFLICTS.
I
MINE COUNTERMEASURES/DEMINING
2-86
LARGE AMOUNTS OF ORDNANCE ON KNOWN MINEFIELDS TO DEMINE
THEM.
CLEAR A CORRIDOR.
WEAPONS.
SCALE.
2-87
OUR PRESENT PRIORITIES SHOULD BE TO IMPROVE CURRENT
ARE UNAVAILABLE.
CONCLUSION
2-88
Technology
and the Mine Problem
I think Sir John Fisher, Great Britain's First Sealord during World War I, captured the
key to success in naval warfare when he said, "The whole principle of naval fighting is to be free
to go anywhere with every dammed thing the Navy possesses."
Unfortunately, at a relatively low cost to our enemies, mine warfare can seriously
interfere with our uncontested control of the sea and fi'eedom of maneuver.
While this low cost weapon, available to nearly any nation, can challenge our uncontested
control of the sea and fi-eedom of maneuver, it is not an inpenetrable barrier. Reflecting on the
USS TRIPOLI and USS PRINCETON incidents in the Arabian Gulf, we find that both these
ships operated un-knowingly in the midst of a minefield for two days with no adverse effect.
The mines were effective, however, as no amphibious landing took place in Kuwait due in part to
the risk of troop carriers and landing craft being lost to mines. The potential physical destruction
to our assault forces from mines was simply too great.
Another example of the effectiveness of mines etched deep in the history of amphibious
warfare is our attempt to land in Wonsan during the Korean Conflict in 1952. The North
Koreans were able to delay the amphibious landing by almost a week using antiquated Soviet
mines. Even the nine days we allowed for mine countermeasures proved insufficient, despite the
assistance of eight Japanese minesweepers. After two U. S. and one South Korean minesweeper
were lost, the amphibious landing took place only after Wonsan was seized by ground forces.
Rear Admiral Smith, the commander of the amphibious task force, wrote in frustration:
"We have lost control of the seas to a nation without a navy, using
pre-World War I weapons, laid by vessels that were utilized at the time of
the birth of Christ."
2-89
Although I know a lot of effort has been put into mine warfare and mine countermeasures
since the Korean War and even more so since the Gulf War, we still do not have a satisfactory
solution to this problem. Countermine warfare continues to evolve as does amphibious doctrine.
Since mine warfare is an ongoing effort that affects many of you in this room, I would like to
share some of my concerns about mine warfare and how they affect the way Marine Forces
operate both now and in the future.
The Pacific Basin and Southwest Asia are the areas my warfighters operate in. These
areas are characterized by vast distances between islands and continents—nations and population
centers—separated by water.
As you can see from the chart, the Pacific Theater of Operations covers a large part of the
world (CHART OF PTO SHOWS SEA LANES IN PACIFIC THEN CENTCOM)
To put the Central Command Area of Operations in perspective, consider that the AOR
takes up an area the size of the United States. In that AOR, two choke points in the sea lines of
communication — the Suez Canal and the Strait of Hormuz — are vital to the health of the world's
economy.
Sea lanes link the oil resources contained in the Central Command area of responsibility
with the expanding economies of the Pacific Theater of Operations.
These lines of communication run through an area of the world with a long history of
turmoil and strife. Both Southwest Asia and the Pacific-Indian Ocean Basins are plagued by
historic animosities, population growth, weapons of mass destruction and illegal drug trafficking.
In addition, rapidly developing industrial manufacturing and export capabilities in these
countries furthers their need for Mideast sourced petroleum to fuel a growing economy.
Keeping these sea lanes open and maintaining security highlights the need for an
amphibious force strong enough and able to maneuver throughout the region. Any naval
campaign intended to keep critical SLOCs open will require a landing force. My Marines need
to be able to operate from the sea — from over the horizon up to 200 nautical miles inland.
Because we operate on both sea and land, Marines are affected by the entire spectrum of mines;
deep water, shallow water, very shallow water, and landmines. Very shallow water mines, those
in 10 to 40 feet of water and the surf zone, are probably the most serious and most difficult
challenge we face.
The primary effect sought by laying mines is to shape the battlefield to the enemy's
disadvantage. The image of a landing craft being heaved in the air by a huge blast probably
played itself out in the mind of just about anyone who has ever landed onto a hostile shore.
However, as spectacular and sensational as this image is, the physical destruction of ships and
landing craft is not the primary desired effect of mines. In amphibious warfare, mines serve to
force an advancing opponent to move in a disadvantageous manner. This may mean forcing a
landing away from a good landing beach or steering it into an area where it can be
counterattacked or attacked by pre-planned fires.
2-90
Even with an amphibious doctrine that uses speed and maximum flexibility to allow us to
pass over or around enemy minefields using the Operational Maneuver From The Sea Concept or
OMFTS, we still need to have a way to clear mines. In the two areas we are most likely to be
employed in major regional contingencies, Kuwait and Korea, suitable landing beaches, even
with air cushioned landing craft, are limited. The enemy knows where these beaches are, and has
plans to defend them either physically or through the use of denial by mines. Even if we have
tactical vehicles and utilize dispersion and precision navigation as defined by OMFTS, some
mine clearing will be required. The entire concept of OMFTS is to hit the enemy by surprise
from beyond the horizon. Mine clearing is not an evolution we can start well in advance of the
assault if we want to maintain our surprise.
The assault echelon-- the trigger pullers-can use tactics and limited mine clearing to
counter the mine threat. For the follow-on echelon , the echelon that is going to sustain the
assault echelon with fuel, ammo, and food it is a different story. Sustainment comes in bulk.
Shipping used to transport sustainment materiel isn't compartmentalized and to achieve
efficiency can't be dispersed. One or two landing craft in the assault echelon going off course or
falling prey to more sophisticated mines that escape our detection efforts may be operationally
acceptable; the battle won't be lost. On the other hand, if one or two ships carrying critical
supplies from the follow-on echelon get destroyed, we can lose the entire naval campaign. Even
though Marines can get a force ashore, if I can't sustain it, I can't employ it. The mines will have
achieved their desired effect.
Even in a relatively benign environment, mine warfare affects our ability to project
power. As many of you know we are going away from forward bases and are becoming
increasingly reliant on maritime prepositioned ships or MPS. Under the MPS concept, the
Marine Corps maintains three squadrons of four to five ships. Each squadron of ships carries
enough ammunition, logistical support and combat equipment to sustain a 17,000 Marine force
for thirty days. The ships pull into a benign port and offload their supplies and equipment while
the Marines fly into a nearby airfield. During the Gulf War, we successfully offloaded all three
squadrons of MPF ships. Flad the Iraqis mined the Strait of Hormuz instead of or in addition to
the coastline off Kuwait, we would have faced an entirely different scenario.
Whether moving our MPS squadrons to an area in conflict, moving our naval forces into
position for battle or protection of sea lanes, or launching our Marines ashore, we need to know
the water is clear. Or we need to make it clear.
Once ashore. Marines face the threat of landmines. The desired effect of landmines is no
different from those in the sea. 1 don't envision losing mass formations of armor to landmines.
What I do see is landmines delaying formations long enough so the enemy can engage with other
weapons. This could range from antitank weapons to artillery, or even chemical weapons. On
today's modern battlefield, the effects of fires are far more lethal than those of previous wars,
even more so against a stationary target. Compounding the problem, in places such as Korea, the
terrain is so restrictive that a temporarily halted force quickly becomes a lucrative target with no
place to disperse.
Even in operations that don't constitute major regional contingencies, mines pose a
challenge. In Military Operations Other Than War or MOOTW, in which we recently provided
humanitarian assistance in places such as Somalia and Bosnia-Herzagovina, the disturbing trend
2-91
is the casualties we sustain are increasingly from mines. One legacy of the conflicts we face in
the Pacific region is the proliferation of leftover mines. Millions of mines are scattered
throughout Vietnam, Cambodia, Laos, Iraq, and Korea. The historical and continual use of
mines presents us with challenges that now are receiving attention. Over the last five years,
Marines have been involved in humanitarian demining missions in Laos and Cambodia. We are
just beginning to tackle the overwhelming task of clearing mines from former areas of conflict.
But it is a concern we all share, and we are trying to do something about.
So how do we deal with this challenge? From an operational perspective, we cannot rely
on any one system. No matter what countermeasure we come up with, someone will come up
with a counter to our counter. Also, we can't just rely on technology alone. We need to integrate
technology into our tactics and techniques. If we reduce the threat of mines through technology
we lower our risk. If we develop our tactics to counter the threat, we reduce our risk some more.
The two measures combined reduce our risk to an overall acceptable level.
Probably the most important thing technology can provide for us is accurate and reliable
detection. If we can accurately detect mines, then we have options. We can detect and avoid or
we can detect and clear. I can't over-emphasize the word, "reliable". Psychologically, we need
to be confident that we know what is out there. This is no small task. In the Arabian Gulf, the
water is shallow enough to make bottom influence mines effective while the muddy bottom
makes them increasingly difficult to detect. In North Korea, a country that chooses guns over
butter every time, sophisticated rising mines, new technology, and clandestine mine sowing
methods make detection a continually changing challenge.
Once detected, mines can be avoided. With the increased availability of Global
Positioning Satellite receiver equipment, every landing craft, vehicle, and individual Marine can
conceivably navigate through narrow mine fi'ee corridors.
Technology will also play a part in the other option, detect and clear. We can no longer
rely on minesweeping technology that takes weeks to complete. The trend in modem warfare is
towards shorter not longer conflicts. Our amphibious doctrine relies on speed and surprise.
Even a couple of hours preparation time may be compromising our surprise. We need to develop
an instride capability to reliably breach even the most sophisticated minefields whether they be in
shallow water, very shallow water, the surf zone or on land.
We also need to work on clandestine means of clearing. We learned before long before
Desert Storm that we needed a night vision goggles capability to complement helicopter
minesweeping. Unfortunately, when Desert Storm came around we still had not developed this
capability. Clandestine mine detection and clearing needs to extend into the surf zone. We need
to be able to pick our beaches — not let the enemy pick them for us. Operational Maneuver From
the Sea demands that to maintain the advantages of surprise and maneuver, we must develop and
enhance our clandestine and covert reconnaissance, clandestine mine clearing, amphibious
maneuverability, and in-stride breaching capabilities.
2-92
Finally, whatever mine countermeasures we come up with, the equipment has to be
immediately available and deployable. We can't afford to give up scarce amphibious platforms
to be used as mine countermeasure platforms. Specialized equipment can't be so bulky that there
is no room for the trigger pullers. Also, we can't relegate the mine countermeasures mission to
the reserves who will never work with naval expeditionary forces until there is a crisis. Mine
countermeasure forces need to be integrated and worked into our peacetime exercises. We can't
put off mine countermeasures until we go to war again.
The lesson from Desert Storm, from Korea, from Vietnam is not that minefields are
impenetrable. The true lesson is that if we ignore the threat, we will pay for it.
Your work is important to the future operational success of your Marine Corps. Fighting
forward from the sea takes courage, tenacity and aggressiveness. Mueh of that comes in
knowing where the mines and obstacles are and eliminating them.
2-93
SENKAKU ISLANDS
2-95
2-96
The Present and Future
of Mine Warfare
* RADM Conley was introduced by RADM Herbert C. Kaler, USN, PEO Mine Warfare (P), who
was introduced by RADM Richard D. Williams III, USN, PEO Mine Warfare.
2-97
HIS TENURE IN THE ELLIS A. JOHNSON CHAIR OF MINE
OF THIS VENTURE.
2-99
AND WEDNESDAY USS KINGFISHER , THE SEVENTH OSPREY
2-100
OPPORTUNITY TO CONDUCT TRUE INTEGRATED TRAINING IN
MCM, BOTH WITHIN OUR MCM FORCES, AND WITH THE FLEET
2-101
PLAN WILL EVOLVE CONSISTENT WITH THE ACCOMMODATION
2-102
THAT COMCMRON 2 HAD ABOARD INCHON IN SOUTH TEXAS.
2-103
FORWARD DEPLOYED., CONDUCTING EXERCISES AND
2-104
CONTINUING BASIS., NOT JUST WHEN WE ARE PREPARING
2-105
THEATER, AND THEN FORWARD DEPLOYING ADDITIONAL
FLEXIBILITY.
2-106
CAPABILITY OF OUR AMCM AND EOD FORCES COMBINED
THINK IS THE BEST MCM FORCE THAT OUR NAVY HAS EVER
2-107
ORDER TO CONDUCT EXPEDITIONARY OPERATIONS WE WILL
UNDERWAY.
2-108
AROUND MINED AREAS IF POSSIBLE, (3) EXPLOIT
2-109
WHICH WILL BE A REAL FORCE MULTIPLIER IF WE CAN
2-110
WARFARE PROGRAM TO REVALIDATE SHORTFALLS IN
MID TERM IS THE POM YEARS, BEGINNING IN '98 AND THE FAR
TERM THE YEAR 2003 AND BEYOND. THE FY 96-97 MIW PLAN
2-111
MCM SYSTEMS OPERATING FROM SURFACE SHIPS AND
SUBMARINES.
2-112
EMPLOYED IN A 5TH FLEET EXERCISE IN JANUARY ALONG
2-113
DETECTION SYSTEM ONBOARD THE SH2G HELICOPTER, THE
2-114
AS I NOTED EARLIER, WE NEED TO FOCUS ON THE VERY
TO THE LITTORALS.
ZONES.
2-115
THIS IS A GOOD BEGINNING IN A VERY DIFFICULT AREA.
2-116
CLEARANCE, TO IN-FACT REDUCE THE IMPACT OF MINES TO
2-117
The Future of the Pacific Fleet
2-119
crews. When I reported to CINCPACFLT I initially was the
14% a year.
The GDP of the region caught up with and exceeded that of,
Europe in the early part of this decade, and the gap is predicted to
2-121
Our friends in the Pacific are extremely dependent on Sea
Lines of Communication and their and their dependence on
United States interests in the region are both clear and vital.
There are 3 million U.S. jobs directly tied to trade in the region,
and if you include the trickle down effect 9 million jobs are
dependent on that trade. By the end of the next decade those
numbers are expected to grow to 6 million and 18 million U.S.
jobs respectively. American pocketbooks are inextricably linked
to the economies of the Asia-Pacific rim. Clearly stability in the. _
region is critical, not only to continued Asian prosperity, but to the
well-being of the United States.
2-122
and that unfortunately, is particularly fragile in a couple of areas
smuggling.
Wherever any of us travel in the region, a common theme
among those we talk to is the much appreciated role that the
Pacific Fleet plays as part of the forward deployed force, by
sustained forward presence, in regional security and stability.
There is however another common and growing theme, and that is
a perception of U.S. lack of concern and withdrawal from Asia.
the close of the Cold War we were down to under 300 ships.
Today your Pacific Fleet is at 194 ships. On the other hand our
2-123
ships today are much larger and far more lethal than their World
War II counterparts. Of course, the cost has risen by orders of
magnitude as well. An interesting point in all of this is that, with
the exception of battleships, the types of ships that we have today
are essentially the same as they were over 50 years ago.
and integration of fully joint C4I suites in our fleet, battle group
2-128
Most of you should be familiar with the decision and
execution loops in both peace and war. In peace, we start by
drafting war plans, then we evaluate courses of action, assess their
impact, develop new courses of action, select the best and then
faster ships it will still take a long time to get anywhere in our
policy.
The trends we see in the future include increasing U.S. trade
with the region and faster economic growth in the near term with a
flattening in the long term for Asia. This implies an increasing
regional competition for markets, for access to a limited money
supply and for constrained natural resources.
- Technological change will continue to accelerate. As I
mentioned we’re down to 18 months for computing generations.
And as military technology is increasingly driven by commercial
developments we will see both the same acceleration and greater,
availability to anyone who can afford it. And more nations and
organizations will be able to afford advanced technology as
competition drives down prices and overheated economies provide
more capital.
- At the same time, U.S. defense budgets may show slight
growth, but flat or declining budgets are more likely as deficit
reduction and other competing demands rise in perceived relative
importance.
2-133
At the same time there are more than enough unknowns
about our future in the Pacific to make our planning a real
challenge.
the Spratleys?
2-134
access to forward basing? What would be the affeet on regional
Asia Pacific, Middle East and Latin America and range in each
to deter.
And we are currently developing the baseline for potential
fleet postures in the years beyond 2010 that will range from a
robust, forward deployed, well equipped Navy to a smaller, in¬
garrison force brought home by a combination of dwindling
2-136
a small museum displaying Nimitz and Pacific Fleet memorabilia.
All of you are invited to come. On one of the walls is a quote
from Admiral Nimitz, it reads: “We must make certain, now and
for the future, that peace is secure. We must remain strong.
Never again should we risk the threat which weakness invites.”
This quote spoken in 1945 is more true today than ever and must
2-137
2-138
The Joint Mine Countermeasures/Countermine
Advanced Concepts Technology Demonstration
(ACTD) Process
2-139
Joint Countermine ACTD Novel Systems
Navy Systems
• Advanced Sensors
• Magic Lantern (Adaptation) [ML(A)]
• Advanced Lightweight Influence Sweep System (ALISS) ATD
• Explosive Neutralization Advanced Technology Demonstration (ENATD)
• Near Term Mine Reconnaissance System (NMRS)
• Littoral Remote Sensing (LRS)
Marine Corns Systems
• Crastai Battlefield Reconnaissance and Analysis (COBRA)
• Joint Amphibious Mine Countermeasures (JAMC)
Joint USMC/Armv Systems
• Off-Route Smart Mine Clearance (ORSMC)
Army Systems
• Close-In Man-portable Mine Detector (CIMMD)
• Airborne Standoff Minefield Detection System (ASTAMIDS)
• Army Classified Program (ACP)
Advanced Sensors
System functions: Underwater mine detection, classification, and identification in
support of finding mmefield gaps.
u j Description: Advanced sensors will replace the AN/AQS-14 sonar in the RMS tow
body. These sen^ will expand the RMS capability. Search rates in deep water against
moored o^es will ^ual 6sq nmi hr. Sensor search rate in shallow water and very shallow
watCT wm decrease m accordant with the decrease in threat area. Sensor data fusion will
provide D/Cyi aga^t all sea mmes. System endurance will provide an 8-12 knot search
^ed fOT up hours on a single tank of fuel For more information contact Dr. W.
Chmg, ONR 321, on 703-696-0804; e-mail: chingw@onr.navy.mil
T TT. A The ML(A) ACTD system will demonstrate the capability of gated,
UDAR imagmg for detection of minefields and obstacles. The ACTD objective will be to
demonstrate a capahitiiy to rapidly detect, classify and localize minefields and obstacles in
m^urf zoneand c^ ^ding ztme. ML(A) will be qrerationally demonstrated during both
Demo I a^ Demo IL The components of ML(A) are the laser transmitter, seamier,
camms, bottom follow®, GPS and processor. The system will also employ real-time
automatic tar^t recogmtion (ATR) and a datalink to ground station for viewing target
i^ges. For Demo H, the ML(A) ACTD System will demonstrate a furth® inmroved ATR
mgotithm md an enhanced tactical decision aid (TDA) for the surf zone mis^
Fot more information conta® Dr. W. Ching, ONR 321, 703-696-0804; e-mail-
chingw@ onr.navy.mil
2-140
Advanced Lightweight Influence Sweep System (ALISS) ATD
The pinpose of the Advanced lightwdght Influence Swe^ System Advanced
Technology DraKHislration (ALISS ATD) is to denKMistrate the ^Uity to successfiilly
conduct autcHiomous influence sweeping of magnetic and acoustic influence mines targeted
against anphibious assault craft in very shallow waters. ALISS will utilize
superconducting magnet and plasma-discharge pulse power technology to provide a high¬
speed lightwei^t acoustic airi magnetic signature wnulation sweeping ct^aWlity. This
technology will aJso significantly r^uce sweep power requirements. ALISS may
eventukly be dqiloyed from a variety of platforms (helicopter, ship, LCAC, or
remote/autonomous controlled boat). During its demonstration in die Joint Countermine
ACID, it will be installed on a Rigid Hull Mlatable Boat for autonomous influence
sweeping of the intended an^hibious assault lanes. For more infimnation contact Mr.
Steve Collignon, ONR 32CM, on 703-696-3039; e-mail: colligs@onr.navy.mil
2-141
Army Systems
Close-In Man-portable Mine Detector (CIMMD)
System fimction: Detects surface and buried met^c and nonmetallic landmines.
Description: The CIMMD Program has developed a standoff IR Thermal Imager
(IRTI), and a confirming Ground Penetrating Radar (GPR) hrassboard man-portable mine
detector. These detectors, which may be employed singularly or in combination ate
suitably packaged and available for inclusion in Warfi^ting Experiments. The standoff
IRTI system has a backpack that includes an image processor and battles; a helmet with
an eyepiece display and COTS forward looking infrared. The IRTI system weight is
approximately 30 pounds with batteries divided between the backpack (25 lbs) and helmet
(5 lbs). The GPR system closely resembles the configuration of the U.S. Army AN/PSS-
12 metal detector. The wand contains an electronics package, the antenna, and an LOT
di^lay.' A backprok contains batteries. The GPR systenrweight is approximately "25 lbs
with batteries - divided between the backpack (10 lbs) and wand (15 lbs). For more
information contact CECOM RDEC/Mr. Mark Locke, on 703-704-2418.
2-142
Army Systems
Close«In Man-portable Mine Detector (CIMMD) i j •
System function: Detects surface and buried metallic and ntmmetallic landmines.
Description: Tbe CIMMD Program has developed a standoff IR Thermal miag^
(IRTI) and a confirming Ground Penetrating Radar (C3*R) hrassboard man-portable mme
detector. Thesedetectors, which may be en^loyed singularly or in combm^on are
suitably packaged and av^able for inclusion in Warfigjiting Experiments. The standoff
mil svst^ has a backpack that includes an image processor and battenes; a hel^t with
an eyepiece display and COTS forward looking infrared, llie IRU system weight is
appiorimately 30 pounds with batteries divided betwwn the bac^k (25 Ite) and htot
(5 lbs). The GPR system closely resembles the configuration of the U.S. Army AN^S-
12 mWa^ detertor. The wand contains an electronics package, the antenna, and an l^D
display.' A backpack contains batteries. Hie GPR system weight is appaoxi^ielyxS lbs
wim batteries ~ divided between the backpack (10 lbs) and wand (15 lbs). For more
information contact CECOM RDECTMr. Mark Locke, on 703-704-2418.
Army Classified Program (ACP) For more information contact CBCOM RDECVDr.
David Lee on 703-704-1063.
2-144
The Importance of Keeping
Historical Records Available
in Mine Warfare
Picture if you will, a bird’s eye view of a riverine squadron winding their way
precariously up a mined river in interior combat operations in hostile territory. The shoreline
of the narrow, twisted river is lined with dense foliage. Leading the fleet is a heavily-armed
monitor, the riverine version of a battleship, armed with mine-avoidance and protection
equipment, and capable of withstanding many different types of enemy attacks. Following
astern are minesweepers, followed by gunboats, their guns trained at an invisible enemy
ashore, their decks crammed with landing parties of three services. In support of this
squadron fly the Cavalry, beating the bushes to expose guerilla raiders, contact mine
operators, and snipers planning to waylay the invading fleet. As you picture this fleet
proceeding up river, you can easily spot the narrowest, most tortuous passes ahead, those that
There are several documents which describe such a scene in detail. Some are sketches
by on-scene observers, as well as detailed battle reports submitted to the government by the
Union Navy in describing riverine operations in the Civil War of the 1860s. Replace the
Cavalry horses with aircraft, and this scene is exactly recreated in similar detail in
photographs, books, and memoirs of the riverine war in Vietnam in the late 1960s. In fact,
putting an etching of the overhead view of a riverine assault in 1864 side-by-side with an
aviation photograph of a similar expedition over 100 years later, as some authors have done, is
the most eerily prescient reminder of an undeniable truth in mine warfare: if you assume we
2-145
will never come this way again, you will live to be proven wrong.
History does not repeat itself; but people often do. Our nation will, in future, continue
to fight wars on open oceans, interior rivers, and the narrow passageways between nations.
Sometime in future, our children and grandchildren will have to worry about defending their
own shores, landing their troops on an enemy-held beach, or flanking a formidable foe by sea.
While budget constraints restrain active planning to foreseeable options, there should be
nothing which restrains us from keeping the lessons of our past alive and viable for the future.
The words of the Vietnam veterans of riverine warfare who wrote about their wrenching and
deadly experiences and came back to study both the past and the present, should haunt us. No
longer should we ever read the words, “why didn’t we learn those lessons 100 years ago?”
For several years, I had the opportunity to freely study the history and current
professor at the Naval War College, and as a researcher funded by the Commander, Mine
Warfare Command. During that time I produced a book and some articles on the history of
mine warfare, deployed three times to document mine warfare operations in Operations
Earnest Will and Desert Storm, and conducted interviews, research and writing for a future
Along the way, I have learned some lessons about mine warfare which pertain to all
future studies, technical, tactical, and operational. The first is, of course, that we should
never assume that old requirements won’t be needed again. We have ample evidence that it is
in our national interest to assume that our future naval forces will require surface mine warfare
ships, aircraft, training and technologies we have found useful in the past, and to be certain
2-146
that the operational requirements from the actual minefigld e?cpgrienc£ of our forces are
Second, mine warfare has depended for its entire history on the efforts of individuals to
keep it a viable option. We recognize today that situation is unworkable. Current worldwide
emphasis on the landmine situation has recently brought that problem into the public
consciousness, although the effect of sea mines on ships in the 1980s and 1990s has already
been erased from public view. The solution to keeping all mine warfare areas from declining
in capabilities must be, in part, to keep it from becoming handed back to individuals within
each service to solve. If there is one “Mr. Mine Warfare” in each service, I can assure you,
from the vantage point of the study of 200 years of our history, it will revert to an individual
person’s problem. “Mine Avoidance” by ships and “Mine Warfare Avoidance” by Naval
Third, we don’t know that much about our successes. Exactly why were our
amphibious landings in the Pacific in World War II so successful? How did we successfully
avoid the painful lessons learned at Wonsan, Korea, in the following two years of mine
clearance during that war? How effective were helicopters in mine clearance at Haiphong?
How did U.S. and allied mine warfare forces keep shipping lanes clear during Earnest Will
and assist in operations resulting in a cease fire in Desert Storm? I have touched on all of
these topics in my book, “Damn the Torpedoes,” which is out of print, (but which is still
regularly plagiarized from in student papers and in naval publications) but only to point out
how valuable more detailed studies of these operations would be. We have in the bookstores
today, self-help books entitled “Don’t Know Much About History,” and “Don’t Know Much
2-147
About the Civil War. My book on mine warfare probably should have had a similar title.
What is needed now is more in-depth, technical study of those elements of mine warfare of the
past.
Fourth, so much information is available in the world press that our declassification
efforts, the product of the perceptions of the Cold War, are largely outdated. We need to
release more information on mine warfare to scholarly study than is currently being done. For
example, no one can properly understand the decisions made during Desert Shield and Desert
Storm in regard to mine warfare unless they understand the purported capabilities of the
illusive Iraqi Sigeel mine, against which most planning was predicated. My own work which
has survived declassification to date allows me to mention the Sigeel, but not to say anything
about it. If we can’t openly discuss a mine that may never have even existed, we will never
fully comprehend and disseminate the circumstances under which our forces actually operated
in Desert Storm.
Fifth, such dissemination of knowledge is the most crucial aspect required to fully
integrate mine warfare as a regular, daily part of the operations of the U.S. military. The lack
of true understanding and communication of mine warfare requirements throughout the Navy
was the biggest failure of our naval mine warfare efforts in Desert Storm, and kept our forces
from being fully utilized to the true extent of their capabilities. We cannot combat lack of
knowledge of mine warfare in all of our services without emphasizing mine warfare education
and better communication up and down the chain of command as a matter of priority.
I suggested in my earlier talk during this conference that the most effective solution to
meet all these challenges is the establishment of a centralized facility for access to the many
2-148
scholarly mine warfare studies and documents proliferating throughout the United States. By
this, I do not mean to suggest that one office take on the task of housing all known mine
warfare documentation, but that they act as the coordinator of document information flow for
studies utilizing documents of historic and operational use in mine warfare. It has become
apparent to me over the past two years through late-night calls from students at many Naval
and Military activities and Colleges, shipboard personnel. Pentagon staffers and mine warfare
officers that there must be a better way to keep everyone studying the problems of mine
warfare informed of the location of various mine warfare documents or of the feasibility of
obtaining sufficient research data on a specific topic rather than having them resort to calling
an unemployed historian. The need for better access and communications within the mine
warfare community and those studying it are crucial to ensure proper and accurate
For those of you who have forgotten or who have never had access to my book, I d
like to end this talk with a few words about the future taken from my conclusions, written in
1991. In it, I recounted the advice of several mine warriors over the years calling for a
number of changes in the way the Navy approaches Mine Warfare. I am glad to see, from the
vantage point of 5 years, that several of these specific conclusions on how to reintegrate mine
warfare back into naval warfare have been met. We now have a flag officer assigned as PEO
Mine Warfare to keep track of programs in the Pentagon. We have altered the strictly
advisory role of Commander, Mine Warfare Command, as it was in Desert Storm, to one of
considerable power to reshape mine warfare funding and forces. We have begun building a
mine warfare community of trained personnel through creation of new leadership billets and a
2-149
Center for Mine Warfare Excellence. We have focussed discussion on prevention of mining,
jointness, and flexibility. We are combining forces to ready emerging technologies and to
rethink existing ones. Most important, we have stopped making mine countermeasures look so
darned easy that bureaucrats can continue to assume that it can be accomplished without proper
There remains much to be done. I have written about many mine warfare “heroes”
whose work left a legacy upon which future mine warriors were able to build. One particular
example stands out in my estimation. Those are the unnamed people who, during the
devastating budget crises and downsizing of the Navy after World War 11, still managed to
leave behind complex and compelling documents requiring considerable study and foresight.
These were the studies of the status of mine warfare at war’s end, and the outlined
improvements which could be achieved. The best were the planning documents for the next
class of ocean-going surface MCM ships, requirements derived from combat mine
countermeasures experience. The existence of this particular document became crucial a few
years later when the disaster at Wonsan forced immediate production of such ships long after
pundits proclaimed them unnecessary for future of the Navy. This is how we got one of the
most capable and long-lasting ships ever in the naval service, the MSO, designed from the
operational experience of two wars, which served as our main surface platform for over a
generation.
I’m going to close this talk by reading the final two paragraphs of my conclusion,
“The central problem of MCM throughout history has been the difficulty of sustaining
2-150
maximum capability over time. By its very nature, MCM evolves as the result of new mine
developments and changing threats. Yet, in the U.S. Navy mine countermeasures have often
been quick-fix solutions. Due to real competing needs, priorities, and lack of mine warfare
knowledge within the Navy, it has been impossible to sustain adequate priority and funding for
MCM. Important lessons learned, even when published by the participants, have been quickly
forgotten, and subsequent attempts to revitalize the service have often been predicated on the
wrong lessons. To date, no Chief of Naval Operations, Congress, or President has been
opposed to an effective mine warfare program, and some have actively championed one. Yet,
Lack of overall mine consciousness has often led us to remember the wrong lessons
from our mine warfare experience. The recent minings of KSamuel B. Roberts, Tripoli, and
Princeton remind us that even our most valuable and expensive warships can be easily stopped
by simple, cheap mines. When the Navy as a whole learns more about the reality and
potential of mines and their countermeasures, MCM will no longer be called the Cinderella of
the service and considered a subject about which much is written and less done. Only
knowledge will end the legends and reveal the truth about men like Farragut, who only
2-151
2-152
Naval Countermine Requirements
2-153
1
2-154
2-155
THE LITTORALS
2-156
POPULATION IN THE
LITTORALS
2-157
"Ten years ago 20 percent of presence missions were
conducted by Naval forces, today it's 50 percent".
- Adm W. A. Owens
2-158
• Since WWII:
- 300 Crises requiring U.S. resonse
- Only 5 required force build-up beyond forward
deployed forces (NATO, Korea, Cuban Missile Crisis,
Vietnam, Desert Storm)
- Only 3 required deployment of war termination forces
(Korea, Vietnam, Desert Storm)
• 27 CVBGs and ARGs forward deployed since 1992
(Average deployment 180 days)
• Average number of Marines deployed Is 24,000
• Since 1992, Marines have been involved in 29 of 41
operations (70.7%);since 1995, 18 of 24 (75%),
2-159
Operational Maneuver from the Sea (OMFTS)
is a marriage between maneuver warfare and
naval warfare ... [It] will couple doctrine with
technological advances in speed, mobility, fire
support, communications, and navigation to
identify and exploit enemy weaknesses across
the entire spectrum of conflict.
2-160
• OMFTS focuses on an operational objectives
and uses the sea as maneuver space. We will
generate overwhelming tempo and
momentum while pitting our strength against
critical enemy weakness. OMFTS
emphasizes intelligence, deception, and
momentum while integrating all organic, joint,
and combined forces.
2-161
/c CAPABILITY IMPRO VEMENTS
V
MOBILITY
INTELLIGENCE
COMMAND AND CONTROL
FIRES
AVIAT^N
MINE cfpUNTERM^^SURES (MCM)
COMBfeiMiel SUPPORT (CSS)
,®.
:saL,
2-162
COUNTERMINE VISION FOR OMFTS
RAPID FLOW OF COMBAT POWER AND
UNIMPEDED TACTICAL MOBILITY FROM
THE SEA TO INLAND OBJECTIVES ^
Mine counter measures will clearly play a critical role in our ability to
conduct OMFTS.
Because of their relatively low cost and pervasiveness, mines have
become a cheap means of limiting the mobility of ships and landing
craft in contested littoral regions. For that reason, we are rapidly
developing and enhancing our countermine and counter obstacle
reconnaissance, marking and clearing capabilities; precision
navigation; and in-stride breaching abilities to support maneuver at
sea, the transition across the beach, and movement inland.
Our deficiencies in mine counter measures can be considered the
"long pole in the tent." We are looking to industry for the technological
advances that will give us real time, seamless transition through a
mined area.
NEAR TERM REQUIREMENTS
• DOCTRINE & MINDSET THAT DOES NOT FOCUS
ON AVOIDANCE
2-164
NEAR TERM REQUIREMENTS (CONT’D.)
' VERTICAL LIFT CAPABILITY TO MANEUVER
SUFFICIENT FORCES TO HELP SECURE MCM AND
SURFACE LANDING OPERATIONS (ADVANCE
FORCE OPERATIONS)
2-165
MID TERM REQUIREMENTS
- IMPROVED CLANDESTINE MINE RECONNAISSANCE,
PREPARATION, AND MARKING
2-166
MID TERM REQUIREMENTS (CONT'D.)
> AMPHIBIOUS FIGHTING VEHICLES CAPABLE OF HIGH
SPEED, MULTIPLE OPTION SHIP TO OBJECTIVE
MANEUVER
2-168
FAR TERM REQUIREMENTS (CONT'D)
- CLANDESTINE RECONNAISSANCE CAPABILITIES
FOR ALL LOCATIONS AND BARRIER TYPES
2-169
"...WHEN YOU CANT GO WHERE YOU WANT TO,
WHEN YOU WANT TO, YOU HAVEN'T GOT
COMMAND OF THE SEA. COMMAND OF THE SEA
IS THE BEDROCK FOR ALL OUR WAR PLANS..."
■ CNO, Admiral
Forrest Sherman,
following the
Oct. 1950
Wonson, Korea
mine crisis
Participants’ ideas about what the littoral environment is like were given added clanity at the
Session at the Monterey Bay Aquarium, where attendees were privileged Jo hear from the
Oceanographer of the Navy and from operational personnel involved with very shallow water mine
countermeasures.
The Mining and Mine Threats Session then raised awareness of the magmtude and complexity
of the problem presented by the mine threat. Major King, in his graphic presentation on mine
clearance in the real world, showed vivid visual examples from personal experience m situations
encountered in the Falklands, Afghanistan, Cambodia and Bosnia. “Hap” Hambric re-emphasizd the
landmine dangers and demonstrated how new technology, combined with ingemous remvention of
existing technology, results in immediately fieldable, pragmatic solutions. Terry Kasey discussed the
commercial foundations of landmine proliferation and reminded the audience that the sea mine
problem is no less of a challenge. Finally, Prof John Arquilla presented a solution which may lessen
the need for stationary landmines altogether - using mobile, responsive, armed unmanned vehicles
to at least partially replace the military need for landmines used in their convention roles as means of
delay and diversion (because of its autonomous systems element, his paper appears in Chapter ).
3-1
MINE CLEARANCE.... IN THE REAL WORLD
INTRODUCTION
3-3
are being cleared manually using probes and cutting through it, and may be too soft to
locators. Although new technology and support heavy mechanical equipment.
equipment is tried from time to time, and used Whenever water passes through a mined area
within limited appropriate applications, there is a constant danger of mines being
operation managers have little time for rapid moved about, possibly moving several miles
clearance techniques. Their job is to ensure, to downstream. This is a regular occurrence in
the very highest level of confidence, that their the Falklands, and a major concern in
area is totally clear of mines when it has been Cambodia, where much of the ground is
swept. In the vast majority of circumstances, covered by water for long periods each year.
mechanical clearance simply cannot achieve Standing water renders most detection
this. Manual clearance is currently the only techniques useless and often prevents any form
solution: this requires each mine to be detected of mine clearance. Much of the grassland in
by probing and metal detection, and then former Yugoslavia is also covered by snow for
disarmed and removed or destroyed in place. several months of the year: this too prevents
the effective use of most detection and
Mine clearance project managers believe that clearance techniques.
many researchers do not understand the
process that is used, or the reasons that no
alternative currently exists. As the manager of
the Bosnian Mine Action Centre says:
"Unless you understand the process,
how can you find ways to improve upon it?"
MINEFIELDS
General
Minefields come in a variety of guises and
sizes, but they are rarely the flat, uninterrupted
grassy plains where so many of the
demonstrations and publicity shots take place.
Even ignoring the special circumstances of
Kuwait's oil lakes or the Falklands drifting Fig 2: Mined areas of grassland soon revert to a wild
sand dunes, minefields are never simple. When state, hiding mines. Water moves mines and often softens
the ground too much to support heavy equipment.
considering the following scenarios (which are
by no means comprehensive) it must be
remembered that they often appear in Vegetation
combination. In the Falklands, for instance, Many of the minefields in former Yugoslavia,
there are steep, rocky slopes with grassy Africa and South East Asia contain dense
patches, crossed by streams and littered with vegetation which has often been established
shrapnel and unexploded ordnance. for several years. In many cases the tangled
foliage simply prevents any access or view into
Grassland the mined area; where it has grown up around
To begin with, where mines are laid in fields tripwires, the situation is particularly difficult.
or open grassland (eg. the Falklands), grazing
animals do not keep it short and the grass soon Unless the vegetation can be safely burned
reverts to a wild state, overgrowing mines and away, deminers are faced with the prospect of
normally forming uneven tussocks. In many clipping and removing each twig individually.
parts of the world flat grassland has waterways Dense vegetation containing small trees totally
3-4
prevents the use of most mechanical Kuwait where mechanical clearance could not
expedients. be used: manual techniques became extremely
dangerous and one British operator was killed
whilst trying to uncover a mine in these
conditions. Terrain with steep slopes and large
outcrops of rock, common in Afghanistan and
the Falklands, clearly makes the use of any
vehicle-borne system impractical.
Battle Areas
Not surprisingly, mines are often found in
areas where battles have been fought,
contaminating the ground with the scrap of
war. At least, there are bound to be large
quantities of metal present: one shell can
produce thousands of steel fragments, each
large enough to dwarf the signature from a
minimum-metal mine. At worst, the area may
Fig 3: Dense vegetation will totally defeat most be criss-crossed with barbed wire and the
mechanical equipment and conceals tripwires. It makes guidance wires from missiles, cratered and
progress painfully slow. littered with unexploded ordnance (UXO).
Rocks
Rocky soil is also a problem in many parts of
the world. Small stones can make probing
almost impossible while larger rocks can
interfere with detection techniques and prevent
the use of ploughs or flails.
3-5
this is not always the case - particularly with tend to think of a pressure-operated blast mine
submunitions. Once armed, unexploded dual - probably plastic - buried in an open, flat,
purpose bomblets such as the American M42 grassy or sandy area. Of course there are a
or the Yugoslav KB-1 are far more pressure large number of different types of mine, and
sensitive than any AP mine. they are rarely presented in such ideal
circumstances
Urban Areas
When considering "cluttered" environments, it
is easy to overlook the fact that mines and Blast Mines
booby traps are also used in urban areas. Pressure operated mines, both anti-personnel
Clearance of houses and surrounding ground, (AP) and anti-tank (AT) are indeed often
for instance in former Yugoslavia and plastic cased; many have a minimal metal
Afghanistan, can be a very slow and complex content that makes them extremely difficult to
process. In most cases the presence of detect. Many are also "blast resistant" making
buildings, walls, fences, paths and roads makes them virtually immune to the effects of shock
the use of mechanical equipment impossible, or explosive overpressure. Increasingly, these
and detection techniques are hampered by the mines are scatterable and will therefore appear
large quantities of metal present. at irregular intervals on the ground -
sometimes in quite dense clusters. Some have
the explosive in the centre (such as the
Yugoslav PMA-3), while others have a central
fuze mechanism and a ring of explosive around
the outside (like the Italian SB-33). Some, like
the Russian PFM-1, contain liquid explosive.
3-6
steel, causing false alarms on detectors. The mines under the path of the tripwire to catch
explosive content of the mine is above the unwary deminers.
ground level and can be awkward to attack
using standard explosive blocks. Directional Fragmentation Mines
These mines use the detonation of high
explosive to project shrapnel in a
predetermined direction. The mines come in
two basic types; the "Claymore" type
rectangular mines (such as the US Ml 8 A1 and
the Yugoslav MRUD) project their shrapnel in
a horizontal fan, and can be lethal at over 50
m. The circular type (eg. the Russian MON-
100) have an effect similar to a large shotgun,
with a cone of fragments projected to ranges
often exceeding 100 m.
Bounding Mines
Bounding mines (like the Italian V-69 and the
Yugoslav PROM-1) are generally buried in the
ground and activated either by pressure or
using a tripwire. When initiated, they jump 2 -
Fig 8: Although bounding mines such as this Italian V-69
4 feet into the air before detonating: once are normally semi-buried, directional mines are often
again, they scatter steel fragments over the placed well above groimd. Both have substantial danger
surrounding area and can injure at significant zones and can contaminate large areas with steel
ranges. Although their high metallic content fragments when initiated.
makes them simple to detect, tripwire-initiated
bounding mines may kill or injure their victims Directional mines are almost always placed
some distance away from the mine's position. above the ground to take maximum advantage
of their range; in Bosnia they are often
The tripwires from bounding mines are positioned high up on tree trunks, overlooking
normally closer to the ground than those of fields of buried mines. Most can be initiated
stake mines and are often totally concealed by either by electrical command or by tripwire,
vegetation. Many have the ability to use and the vsdre may also be several feet above the
multiple tripwires running off in different ground. Deminers must adopt a special
directions. Because a tripwire could have a procedure to destroy these mines in place
mine on either end, or may initiate a device when they are well above the ground, this can
when it is cut, taut wires must always be be clumsy and dangerous. Once again,
followed to both ends during manual detonation scatters fragments over a large
clearance. It is common practice to bury blast area, making subsequent detector search far
3-7
more difficult. explosive. With such a formidable array of
potential traps, it is almost impossible to devise
AT Shaped Charge Mines universal manual mine clearance drills.
Modern AT mines are making increasing use
of shaped charges to enable a small explosive
charge to defeat vehicle armour. The most SHORTCOMINGS OF CURRENT
common principle is the use of a Misznay CLEARANCE OPTIONS
Schardin plate, which becomes a "self-forging
fragment" (sometimes known as an Hand lifting and probing
"explosively formed projectile" or EFP) when Greatly complicated by hard stony ground,
the mine detonates. The Yugoslav TMRP-6, booby traps and anti-disturbance mines.
a mechanically laid shaped charge mine, can be
electrically command detonated or operated by Metal detectors
pressure or tilt-rod. The Misznay Schardin Can be defeated by minimum-metal mines.
plate can also travel for a considerable Greatly complicated in ground heavily
distance, allowing it to be used horizontally as contaminated with shrapnel and scrap.
an off-route mine.
Flails
Scatterable shaped charge mines, like the US Can be defeated by resilient anti-shock mines.
BLU-91/B "GATOR", generally incorporate a Defeated by barbed wire, thick vegetation and
magnetic influence fuze. First-generation difficult terrain.
magnetic influence fuzes are often initiated by
movement, giving the mine and inherent anti¬ Rollers
disturbance function. GATOR is also fitted Defeated by double impulse fuzes, careful
with a self-destruct feature, though this proved positioning of mines, thick vegetation and
unreliable during the Gulf War. difficult terrain.
3-8
their detection. They can be used below, on, It is crucial that those working on new mine
or above the ground; among rocks, thick detection and clearance techniques take
vegetation, and in shallow water. Considering account of the real-world problems. Ideally,
that any combination could be encountered in they should be familiar with the nature of the
a single minefield, it is hardly surprising that technical threat and the environment in which
there is no panacea. it exists in order to understand the process that
is currently used for demining. In practice this
is difficult to achieve without extensive
operational experience; the obvious solution,
therefore, is to bring the demining community
together with the researchers. Without this
symbiosis, scarce resources will be wasted on
unrealistic solutions that simply cannot address
the problems of mine clearance in the real
world.
Colin King
Fig 9: The beaches of Kuwait with tripwire-operated V- Major Colin King left the British Army after 14 years, but
69 bounding mines on stakes in the water and pressure still undertakes operational work as the sole EOD analyst
operated VS-50 AP blast mines in the sand. A for the Ministry of Defence. His experience includes
combination of large rocks, steel stakes, barbed wire and work in the Falklands, Gulf, Afghanistan, Cambodia,
wet, stony sand ruled out everything except manual Croatia and Bosnia. He is a freelance consultant, and
clearance. Just two types of simple mine in a typical recently completed a detailed technical reference book
battlefield setting: parts of Bosnia are far more entitled "Mines and Mine Clearance" for the Janes
complicated than this! Information Group.
3-9
3-10
THE ANTIPERSONNEL MINE THREAT
ABSTRACT L INTRODUCTION
Mines are a major threat in all types of combat The wide i^read employment of landmines threatens
and will be the major threat in Operations Other Than to neutralize US advantages in firepower and mobility by
War (OOTW) which are expected to be the most likely severely limiting our ability to maneuver and disrupting our
missions for US forces in the future as well as post conflict tactical synchronization.^ Mines are a major threat in ^ types
humanitarian demining operations. Historically, mines of combat operations and will be die major threat in most
have always comprised a large part of the total threat, and Operations Other Than War (OOTW). Mines directly attack
their share is increasing. The clear historical trend is that the basis of our current doctrine by limiting our tactical
mines will be the dominant threat to the lives of US Army maneuverability and slowing our operational tempo. Like
personnel in the future. The current worldwide chemical warfare, mines are part of the battlefield environment
proliferation and widespread employment of landmines and effect everyone involved. Nonetheless, the US Army has
threatens to neutralize US technological advantages in fielded very little to counter the mine threat, while their fuzing,
conventional conflicts. In these conflicts, mines are a lethality, and emplacement technologies have continued to
threat to all US personnel (combat or support) who are evolve.
employed within 25 to 50 km of the combat zone. In The vulnerability of US personnel and tactical vehicles to
OOTW, the use of mines threatens the successful even the most primitive mines has resulted in a significant
completion of such operations by creating unacceptable number of American casualties in the last seventy years.^ The
casualties which undercut popular support. In this rate of losses due to mines has been rising since WW1. Mines
situation, mines are a threat to any individual that have caused almost 100,000 US Army casualties since 1942,
operates outside of small, carefully secured areas. enough manpower to create nearly seven infantiy divisions.
The principal global threat consists of very large Mines have also knocked out an estimated 1,200 US Army
numbers of unsophisticated (but effective) mines and a tanks, enough to equip almost 5 WWII period armored
virtually unlimited supply of improvised explosive devices divisions (Table 1 and Appendix A).
which may be used in mine roles. These low tech mines are The threat posed by mines will only worsen as the
the expected threat in OOTW. A much smaller, but worldwide proliferation of advanced landmines continues.
steadily increasing number of modern mines with Many mines currently being produced around the world are
advanced electronic fuzes are appearing in the inventories significantly more advanced than the current US inventory of
of conventional armies. Conflict with one of these forces is conventional mines.^ In addition to the mines being sold by
less probable, but in time, the electronic fuzed mines are former Warsaw Pact countries, several western and third world
certain to become a major problem for US personnel as countries manufacture and export mines as well, some of which
they find their way into the hands of unconventional are quite advanced (Table 2)."* The worldwide mine inventory
forces. is estimated to be several hundred million, with an estimated
This paper will provide an overview of the 2500 mine and fuze combinations and approximately 100
antipersonnel (AP) mines available worldwide, and in million emplaced.
doing so, outline the AP mine threat for the use of the Compared to other military technologies,
countermine and demining communities. The numbers countermine offers the Army the greatest payoffs in the areas of
and varieties of mines will at first seem daunting, thus it casualty reduction and increased battlefield mobility. To
must be emphasized that the mine threat can be effectively achieve these payoffs, the requirements writer/battle simulator
addressed through the development of appropriate must determine which mines constitute a significant threat to
technologies. the force in question. Then the countermine designer must
understand the nature of those mines and how to counter them.
3-11
TABLE 1: US ARMY MINE LOSS RATES* TABLE 3; AP MINE TECHNOLOGY EVOLUTION
WESTERN EUROPE 21% 2.4% 4.7% 4.1% CONTROL None -Remote control on/ofF
-Programmable self destmct
PACIFIC 33% 1.9% 1.9% 1.9% or self neutralization
KOREA (56%) 4.5% 3.8% 3.9% EMPLACEMENT Manual -Wide variety of scattering
or laying means
VIETNAM (73%) 28% 34% 33%
ANTIHANDLING None -Integral electronic
PERSIAN GULF 60% 34% (3.0%) (10%) -Electronic add-on
ADAM/PDM M3
PP-Mi-Srll PP-Mi-D
PMR-4 PMA-3
VS-SAPFM3 VS-MK2
BM-85 VS-SO
RANGER
UK PAD
C3
M-63 M-1951
K-2 DM-11
trip wire fijzes, these mines have proven extremely effective, and either the MON and OZM series mines.The VP series of
their greater effective coverage enables the emplacing unit to get devices were first employed in Afghanistan and have proven to
the same effect with significantly fewer mines per kilometer of be extremely effective. The Soviets and the French have also
front. fielded breakwire fuzes. The breakwire fiize, which can be
These highly lethal, area effect weapons have had used with both directional and bounding AP mines, is based on
their performance significantly increased through the use of a collapsing circuit. When the delicate wire is stepped on or
advanced fuzing and the use of pre-fragmented casings. cut, a circuit is broken, and the mine is activated.'^
Advances in fiize technology, such as seismic influence and
breakwire circuits, which can be found in some mines being DIRECTIONAL MINES
fielded today will make them even more difficult to counter. The first directional AP mine to enter production, the
For example, the Soviets have developed and fielded the VP Ml8 Claymore, first saw combat in Vietnam.''* The Claymore
series of mine control devices which are based on seismic has a lethal range” of 50 meters and covers a 60 degree arc. It
influence and possess an advanced processor for target is widely copied and is employed by many countries (Table 6).
discrimination. These devices can be employed with five of Two of the most effective directional mines currently in use are
3-13
the Soviet manufactured MON-100 and MON-200. They Sophisticated add-on fuzes such as the Italian VS-
produce a kill zone with a 4 to 5 degree arc and an effective APFl for bounding AP mines are becoming available today.
lethal range of 100 and 200 meters respectively. No other The VS-APFl is an improved electronic fuze for the Valmara
directional AP mine currently in use can match the ranges of 69, which was encountered in large numbers during the Persian
the Soviet MON-200.*^ Among the important concerns for the Gulf War. The VS-APFl has a ten minute safe arming delay.
countermine designer is the early detection and neutralization After this delay, it dispenses three highly sensitive tripwires.
of approach hazards (tripwires, breakwires, seismic sensors, This device also has a field programmable self-neutralization
and command detonation) and ballistic survivability. time, giving the user the option of having his minefield
TABLE 6; COMMON DIRECTIONAL AP MINES neutralize at a preprogrammed time. After the mines have self-
neutralized, they can be recovered and reused.^® This fuze
Origin Mine Fuze Lethal Explosive
Lethal allows the upgrade of “dumb” first generation in to modem
Arc Range
self-neutralizing “smart” mines. Among the important
US Ml 8 T,C 60 50m 682gC-4 concerns for the countermine designer is the early detection
and neutralization of approach hazards (tripwires, breakwires,
USSR MON-200 T,C, 4 200M 12kg TNT
B,S
seismic sensors, and command detonation) and ballistic
survivability.
MON-100 T,C, 5 lOOM 2kg TNT
B,S
TABLET; BOUNDING AP MINES
3-14
fragmenting AP mines (Table 9) which employ advanced provided through fire support channels to maneuver units or
electronic fuzing. These are grouped in a separate table their supporting engineers. This problem was vividly
because of the critical employment differences between the two illustrated during Desert Storm where objectives and the areas
types. These scatterable mines are frequently emplaced during immediately adjacent to them were found to be covered with
runway denial missions and to deny enemy access to his submunitions to a surprisingly great degree.^^
Nuclear, Biological and Chemical weapons storage facilities as
BLAST AP MINES
was done during the Persian Gulf War.^"* Among the important
Blast AP mines are descended from the large
concerns for the countermine designer is the early detection
underground mines that were dug under fortified positions and
and neutralization of approach hazards (tripwires) and ballistic
then detonated. Modem blast AP mines are produced by a
survivability. number of countries (Table 10). Significant improvements have
TABLES: COMMON SIMPLE FRAGMENTING AP MINES been made to "toe popper" mines, which, despite their inherent
simplicity, have been used with devastating effectiveness over
Origin Mine Fuze Lethal Explosive
the years, most recently by the Soviets in Afghanistan.
Radius
Examples include the PFM-1 and PMN. These improvements
US M-3 T,P 9M 408 g TNT include virtual elimination of metal (to decrease detectability),
blast over-pressure protection, low operating thresholds, integral
USSR POMZ-2M & T 4M 75gTNT
POMZ-2 TABLE lOt COMMON BLAST AP MINES
3-15
antihandling devices, and self-destruct or self-neutralization III. MINE EMPLOYMENT
options. Many of these are also suitable for scatterable mine In order for the requirements writer, the battle
laying. The most significant advances in blast AP mines can be simulator and the system designer to make informed decisions
found in Italian mines such as the SB-33 or the VS-Mk2. One on the mine threat definition for a given system, it is necessary
of the most unusual blast AP mines is the Yugoslav UDAR, a to understand how and where mines will be employed in
command detonated bounding FAE (Fuel-Air Explosive) mine.^^ different situations. From this, it is possible to determine rough
Among the important concerns for the countermine designer is probabilities of encounter against the different mine threats.
the safe detection of low metallic content mines. These probabilities will be a function of the threat’s mine
inventory, doctrine and situation. These factors will evolve with
CHEMICAL MINES time and must be periodically reconsidered, particularly in light
The British developed Livens Projector was first of the availability of many advanced mines on the open market.
employed in 1917 and is arguably the first chemical mine.^^ Threat doctrine can be used to predict where mines will be
Except for the introduction of nerve agent fills, no significant encountered and in what manner (i.e. density, mix, emplacement
improvements have occurred in the design of chemical mines techniques, etc.). This information should help decrease the
since WWII. Only the US and the former Soviet Union have probability of over/under design and limit the needless
been identified as producing them (Table 11).^^ However, It expenditure of lives and money.
should be noted that most blast AT mines can be readily Although no two minefields are exactly alike, there
converted to chemical mines by removing the main charge and seems to exist three basic mine warfare doctrines in the world:
replacing it with the desired chemical agent. Chemical mines US/NATO, Russia/former Warsaw Pact, and Guerilla (such as
are typically identifiable by their color or markings. These are Viet Cong). The differences are described below. The first two
intended to be integrated within normal minefields.Flame doctrines are focussed on battlefield countermobility. Guerilla
mines such as the improvised flame fougasse occasionally mining activity is focussed on interdiction and harassment, in
employed by American engineers are also technically classified effect, a cheap substitute for artillery. Mines are also used by
as chemical mines. guerillas to control refugee movement and to undermine the
TABLE 11: CHEMICAL MINES
political stature of their opponents. Because of the significant
differences in how mines are used in conventional operations
Origin Nomenclature Fuze Fill Remarks
and OOTW, they will be discussed separately.
US M-1 C H, SCHEDULED FOR
HD DESTRUCTION CONVENTIONAL OPERATIONS
North Korea and Iraq are the two threats most
M-23 P,C VX SCHEDULED FOR
DESTRUCTION
frequently cited as the basis for the current US strategy requiring
the capability to win two nearly simultaneous medium regional
USSR KHF-l/KHF-2 C H conflicts.^^ Both countries possess large mine inventories which
CHINA ? would figure prominently in a war with either (Appendix C).
- -
North and South Korea have emplaced large numbers of mines
Fuzes - P=pressure, C=command between them along the DMZ (Demilitarized Zone)^^ and Iraq
still retains a large inventory of mines.^^ Many former Warsaw
BOOBYTRAPS Pact members and former client states of the USSR continue to
Boobytraps and Improvised Explosive Devices (lEDs) employ Soviet doctrine (Figure 1).
are an adjunct to AP mines and are frequently employed in an
Iraq dramatically demonstrated the proliferation of
urban environment. They can utilize any of the kill mechanisms
advanced mines during Operation Desert Storm, when they
discussed earlier. The primary advances in boobytraps have
employed mines from nine different countries of origin.^^ Their
come in the area of fuzing. Some examples include light
diverse inventory of advanced mines provides an excellent
sensitive devices (based on photovoltaic cells) that detonate
example of what is available on the open market.^^ US forces
when light of sufficient intensity strikes a sensor, and anti-probe
involved in future conventional operations must be prepared to
pads that initiate the main charge if a soldier using a probe
face a mine threat similar in scale and sophistication to that
pushes against it.^® However, these devices are rare. The vast found during the Persian Gulf War.
majority of boobytraps the soldier can expect to encounter will
Minefields laid for conventional operations are
be relatively simple but cunning devices that use mechanical or
routinely part of a complex obstacle system that typically
electro-mechanical firing devices.^* Some of the most advanced
includes a mix a AT and AP mines, wire obstacles, antitank
boobytrap devices were manufactured in Yugoslavia. These are
ditches, and restrictive terrain (Figure 2). These obstacles are
the “Superquick Family of Electronic Fuzes” and they have normally covered by direct and indirect fire.
seven different activation options.^^
3-16
(StMPUnEDDUGfUM. NOTAU
DETAJLS OR RfARONS SHOWN.)
NOTES;
1. Main drferttwi M is oryanind into
tM> edtelons and a cserve
• FnletfidonaAds enemy tocsav
forcing han to oonoenMe and Cto^afat
^itn into Ine sadto
MR RE6T j.^?...E- • Second echetoirs mission is to desroy
IMS KM irtoffly or reinforaneptooe Irst ectietan
r^O"' %, LEGEND:
,irl
MRdMotorized Rk. 6N=ea8Mori
DAG=i>v»siot anery Gnxf. TKsTank.
HR REGT REGT=flegimeoi CPsCommand Post
10 IS KM
TK RECT i*| ^ fntUmtSwWtr,
EH2 Mbed tineield (anipersonnif
andanitank)
WW Bsitof
Minefield characteristics will also vary with their of these two systems. For example, the Iraqi minefields used
relative location in a dynamic battlefield environment. NATO type mine clusters with a mix of AT and AP mines,
US/Allied situational obstacles consisting mostly of while the North Koreans employ Russian patterns with slight
artillery/MRL (Multiple Rocket Launchers) and aircraft variations on mine spacing.
delivered scatterable mines can be placed well behind enemy Among advanced nations, the employment of manually
front lines during deep battle operations. The main battle area emplaced minefields by conventional militaries is expected to
should see the extensive employment of organic mine continue to decrease in favor of the use of scatterable mines.
emplacement capability (manual, vehicle dispensers, and tube However, they will remain a significant presence on the
artillery) by either side. While US rear areas may see the battlefield for the foreseeable future. These minefields can be
emplacement of nuisance mining of LOCs (Lines of laid by the unit on the spot from resources at hand, in a manner
Communication) by enemy scatterable mines, special forces that can be specially tailored to the situation.^® The
and/or guerrillas^^ as well as their own protective minefields characteristics of these will vaiy with doctrine (Figures 3 & 4),
around key installations. As the battlefield shifts, rear area units time, equipment, training, available mines (Table 12 provides
will also encounter the remains of obstacles in the old deep and information on the probability of encountering different mine
main battle areas (such as breached/marked minefields, types in a manually emplaced minefield), and the situation.
bypassed/unmarked minefields, and dud submunitions). They are either tactical/protective nature or are employed as a
nuisance. Manually emplaced tactical minefields will appear
CONVENTIONAL MINEFIELD CHARACTERISTICS more frequently when the opierational tempo slows and later in
war, after the depletion of high tech mine stocks. The automatic
MANUALLY EMPLACED MINEFIELDS self destruct/self neutralization features of advanced mines will
Although the characteristics of scatterable minefields prevent their use in barrier minefields (like those around
are mainly a fimction of the delivery system, the doctrine for Guantanamo Bay, the Korean DMZ, the Iraqi minefields placed
manually emplaced minefields differs l^tween the US/NATO in Kuwait, and along the old Inter-German border) that require
and Russia/Former Warsaw Pact. Considering that most third an indefinite service life. Also obsolete, second generation mine
world militaries have been trained by one of these two parties, laying systems such as GEMSS (Ground Emplaced Mine
most conventional minefields encountered will be based on one
3-17
Figure 2: Obstacle types^
STANDARD BLOCK
3-18
nURWIRI
lit ROW
U ROW
U ROW
li ROW UROW
» ROW SO ROW
PRESENT FUTURE^^
BLAST 1 1 1 1 - - 1 ■
DIRECTIONAL 1 4 - 1
B 2 1 1
BOUNDING 1 1 2 1 1 1 1 -
STAKE 1
Scattering System), the M-56 helicopter dispenser, and the M-57 SCATTERABLE MINEFIELDS
towed mine planter may have to be employed in a long war, if Scatterable mines were first introduced by the Italians
they are still available. Manually emplaced minefields are and Germans early in WW II. Since then, a wide variety of
generally located within direct fire range of defensive positions, means have been developed to emplace them. Frequently, the
althou^ covering forces units may also employ them as part of mines employed by these systems are of the most advanced type
their deception operations.^* (Tables 14 through 16 provide information on the probability of
encountering different mine types in a variety of scatterable
MINE EMPLACEMENT SYSTEMS minefields). Many of the minefields emplaced by these delivery
Many countries possess a variety of mine emplacement systems cover large areas (Figures 5 & 6) in a very short time.
systems (Table 13). The availability of these systems will Additionally, these systems allow units to rapidly emplace
significantly increase the influence of mines on the battlefield. considerably more mines than they could using just traditional
manual mefliods (Table 19).
3-19
TABLE 14; VEHICLE EMPLACED MINEFIELDS*
PRESENT FUTURE^^
BLAST - - - - - 2 -
DIRECTIONAL ■ - > - - - - -
B
BOUNDING - - - - - - - -
SIMPLE FRAG 1
Scored from 1 (most common) to 4 (very rare)
- - - B 1 1 1
♦This table is based on educated guesses and provides only very rough approximations.
♦♦Through about 2005
PRESENT FUTURE**
BLAST - - - - - - - H'
DIRECTIONAL - - - - - - -
Bl
BOUNDING - - - - 1 2 - 1
SIMPLE FRAG 1 - - - 1 1 1 1
Scored from 1 (most common) to 4 (very rare)
♦This table is based on educated guesses and provides only very rough approximations.
♦♦Through about 2005
PRESENT FUTURE**
BLAST 1 1 - -
B 3 1 -
DIRECTIONAL - - - - - - "
BOUNDING • - - • - - -
B
SIMPLE FRAG 1 - - - 1 1 1 1
Scored from 1 (most common) to 4 (very rare)
♦This table is based on educated guesses and provides only very rough approximations.
♦♦Through about 2005
Expected minefield densities (Table 17) vary from .2 emplaced with minimal risk by the guerrilla, while achieving
to 2.17 mines per meter of front, with an average of about .8 significant disruptions to military operations and the civilian
mines per meter of front. For a historical perspective, see Table economy. The mine has become the guerrillas' weapon of
18 for the minefield densities found during some critical battles. choice.'*^ They provide the guerrilla with an ideal "economy of
force" capability and serve as an "equalizer" against a more
IV. OPERATIONS OTHER THAN WAR (OOTW)'" technologically sophisticated opponent. It is expected that mines
Since World War II, 70% of US personnel casualties will continue to rank at the top of the guerrilla's list of preferred
in combat have occurred in (DOTW."*^ In these conflicts, mines weapons. 400 to 600 million mines have been emplaced in the
have also accounted for about 33% of our personnel losses. last 55 years of which 85 to 200 million remain active
Most of the mines are emplaced in a manner similar to that used throughout the world.'*® These will constitute a serious threat to
by the Viet Cong against the US during the Vietnam War. This any US forces committed in these areas (Appendix D).'*^
guerilla "doctrine'*'’" was been heavily exported by the old Military forces responding to peace keeping or humanitarian
Communist block with examples of translated manuals turning missions require great mobility. Keeping the region's lines of
up in Africa, Asia, and Central/South America. Mines can be communication and economic infiastmcture free from danger are
3-20
I-m M Indicates Mine Encounter
20 10 5 1M
3-21
TABLE 17: MINE DELIVERY SYSTEMS^
US Gator F N/A BLU-92/B .66*' or .2 132 AP mines in an area Approx. 200 X 650m per aircraft
Volcano V.H N/A BLU-92/B .14 Two 1,110 m strips, 35m deep with 70m (air delivered) or
50m(ground dispensed) between strips, 960 mines capacity*’
155mm How A 17km ADAM .lto.8 Emplaced in modules 400 X 400m or 200 X 200m
* A=artillery, H=helicopter, F=fixed wing aircraft, R=rocket, V=vehicle dispensed, MD=manportable dispenser
**In mines per meter of front
TABLE 18: HISTORICAL MINE DENSITIES as part of an ambush) as well as protective minefields around
BATTLE DATE MINES PER KM OF base camp areas, and denial mining of villages and agricultural
FRONT*’ areas7® The primaiy threat to personnel is from blast AP mines.
3-22
are well ahead. Countermine equipment is increasingly
TABLE 19: DAILY DIVISIONAL OBSTACLE EMPLACEMENT
inadequate to deal with the threat, both in terms of capability
CAPACITY""_
and quantity. For this reason, it is necessary to carefully
RUSSIA US consider the mine threat during system design/selection.
OBSTACLE ASSETS LINEAR ASSETS LINEAR
TYPE OUTPUT OUTPUT VI. DEMINING
Some of the earliest examples of area mine clearance
Manual MF All Cbt 2.2 km” All Cbt 14.7 km
occurred during the American Civil War when irate Union
(Protective) Units Units
soldiers under the commands of Generals McClelland and
(Tactical) SEngrPlt 9.3 km 27Engr 75.0 km” Sherman forced Confederate POWs to clear mines (“land
Pit torpedoes”) around Williamsburg, Virginia and Ft McAllister,
Mechanical GMZ 2.4 km Volcano 6.6 km"* Georgia. Some emplaced Confederate landmines have been
MF* X3” X6 found as recently as 1960 near Mobile, Alabama.^
The first post-conflict demining operations were
UMZX4 1.0 km
conducted in France following World War I. At this time, US
PMZ-4 6.4 km” engineers cleared thousands of German land mines. The 108th
X8 Engineer Regiment, 33rd Infantry Division alone cleared over
6,000.®^ After World War II, an estimated 45,000 man-days
Artillery MF* BM-21 6.0 km M-109 23.0 km“
X18 X72 were required to clear 8 million mines from France, Germany
and Belgium alone.®® The British had cleared 280,000 of their
Helicopter MF* PKPI Volcano 3.3 km mines from their beaches by March, 1946. This work continued
(Mi-24 X3
X6) through at least 1958.®^ In the last twenty years, significant
demining operations have been performed along the Suez Canal,
VMR-2 2.1 km" Afghanistan, Kuwait, Cambodia, Somalia, El Salvador, Angola,
(Mi-8
Somalia, and Bosnia to name just a few. Emerging US doctrine
X4)
is contained in TC 31-34 Demining Operations (Initial Draft).^
Fixed Wing MF N/A N/A”
VI. CONCLUSIONS
TOTAL 31.6 km 122.6 km
MINEFIELDS Mines directly attack the basis of current US doctrine
by limiting tactical maneuverability and slowing operational
ANTITANK MDK-3 10.1 km M-9 16.2 km tempo, yet the US has fielded very little to counter this threat. If
DITCHES X6 ACE
X63 US forces are to accomplish their assigned missions with
minimal casualties, it is essential that mines be recognized for
i
TOTAL 1 41.7km 138.8 km the threat that they represent. US military systems must be
OBSTACLES
designed accordingly, since mines typically account for a
* These capabilities are frequently held in reserve to provide rapid situational
significant portion of our casualties in both conventional
obstacle emplacement
MF-minefield operations and OOTW.
Many nations have developed and are exporting
Gill, Commandant, US Army Engineer School asked, '‘Why scatterable mines that can be placed throughout the whole area
does something have to get broken (I mean really broken, like of operations even deep in US rear areas where logistic units are
countermine) before we focus attention and rally resources?” not supported with any significant countermine capability.^^
Furthermore, the US Army remains poorly equipped One thing is certain: US deficiencies in countermine threatens
and trained to deal with mines in a low intensity conflict the mobility necessary to successfully execute current doctrine,
scenario. The capabilities of US light forces have not possibly jeopardizing the successful completion of many combat
significantly changed since the Vietnam War.®^ In fact, US operations and OOTW.
dismounted soldiers are using essentially the same technology
that was available during World War IL®'* This is because
countermine equipment was a low priority during the Cold War
when the US was preparing to fight a defensive battle in Central
Europe.
The technical advances in AP mines have significant
operational and tactical implications for future US combat
operations.^^ Currently, in the never ending spiral of
measure/counter-measure/counter-counter-measure, the mines
3-23
TABLE A-1: US ARMY MINE CASUALTY SUMMARY
Column (5) plus column (7), percentage calculated by dividing column (9) by column (8)
TABLE A-1; US ARMY MINE CASUALTY SUMMARY (CONTINUED)
3-25
Tank chassis based vehicles only, does not include APCs/IFVs or other tactical vehicles.
TABLE A-2: US ARMY GROUND PERSONNEL MINE KIAs (WORLD WAR U TO PRESENT)
3-26
This is consistent with Warfare and Armed Conflicts. A statistical Reference. Volume IL by Michael Clodfelter, McFarland and Company, 1992, page 959, which gave an overall estimate of
a
d^
> o
d^
CM
d^
m
00
t-
ci
(S-2 w s
o r- m CM VO cn
m o VO m rt o
< <s & fS o m o
00 (S in’
o S\ S <s VO
f-<
H
S
y
/<-s CU
»n d^ d^
CM O CM VO
'S^ 00
00 m (S cn
00 <S m C7\ 00 VO
Ol
S m
fO cs
VO irT oo'' fO C'*'
O fO *-( CM
TABLE A-3: US ARMY CASUALTIES KILLED IN ACTION IN WORLD WAR D, BY MUNITION'
sS
o (M O
S' $ o^
d^
o
00 VO fO m 00
<s Tf t-
1
VO
fO
r-
C'l
00 oo
00
oo^ 2
< ri ro o
a.
s
u
<
Ph
Sw
fNj 00 e
? m
C^J m
o Ov Ov CM o\
VO
fS rf 0\ m 0\ § 00
Cv R m R »n c^
oC o' vo"
P cn VO
I 1
1■
o
(S ^ Sf
Pi ci vn o Ov
Ci
P VO 00 o
752
376
o> VO CM
oo <s VO
Q o
00^ r;
o
s
■
C*'*' cm''
in 00*' (S fO CO
fO
Oi
I sP
d^
(o' ■■ d^
■■
cs 00
o
m 00 CM
sH fO
Tf
m
m
VO
fO
w
r-
r- fO
KH
CM
oo
CO
O
ro
2.7% KIA by mines plus 0.2% KIA by boobytraps.
*n m m vo^
CO
cT rn* VO i> s itT in cm"
cs m CM o
»—1
I■ :z;
i
H
•3
•3
a
o i
H § o 1
1
o
<
9 o
s
§
1 1
PQ
Q
CO CO
d
TtI
1 I
CO
00
CO
H
w
tP
i
p
CO
w
s
HH
P <3 5
O H
ffl m
1 d s 5
H H H
3-27
TABLE A-4; US ARMY CASUALTIES KILLED IN ACTION SINCE WORLD WAR n, BY MUNITION
00
US 4,257 516 614 20.9% 783 3,288/6,251 174/604
2
US 277 118 26 20.7% 52 221/419 12/40
69Z
ALLIED (UK, FRANCE) 1.757 1,757 269 15.3%
00
00
M
00*
SICILY TOTAL 109 72 28
NT
1
cs
r3
US 58 1/5
2
ALLIED (UK, CANADA)
8
US 588 150 137 167 470/894 37/129
,
FJ
'a;
S §
847
1 1
MEDITERRANEAN TOTAL" 4,442 718
3-29
LZL
US 1.020 180 158 243 738/1403 50/174
§
ALLIED (UK, CAN, FR) 3,422 302 20.0% 624
The Mediterranean Theater is the sum of North Africa, Sicily, and Italy.
‘Task Force Eagle Mine Strike Summary” as of 5 August 1996. The US has lost 1 M728 CEV, 2 Bradleys, 2 Panthers, and 1 HMMWV to mines. Except for the HMMWV, all vehicles were
TABLE A-6: US ARMY GROUND PERSONNEL WOUNDED IN ACTION (WORLD WAR H TO PRESENT)"
vS
§
WWII (OVERALL) 599,724 507,873 8,247
i
WESTERN EUROPE 393,987 334,889 6,251 15,759
S!
1.529
E
U
78.886
1
93,202
3.6%
r-
2,451
TT
VO
Wk
65,170
'd
KOREA
§
VO
VO
PERSIAN GULP 467
o\
o
6.5%
o>
SOMALIA 137
o
100%
•T
BOSNIA
00
TOTAL 864,437 740,883 8,247 11%
3-31
Column (7) divided by column (3).
TABLE A-8: US ARMY CASUALTIES WOUNDED IN ACTION SINCE WORLD WAR B. BY MTOOTION
3-33
The following are not included in these totals, Grenada: 119 wounded, Panama: 320 wounded.
DISCUSSION this breaks down to 7,246 US Army mine KIAs and 56,783 US
There are a variety of potential inaccuracies in this Army mine WIAs. Grenades are assumed to account for the rest
estimate of US mine casualties. The estimates, including of the US Army KIAs (4,225).
accuracy of diagnosis of cause of death, identification of the The WIA fraction for each munition type was
ordnance responsible for the casualty, and data manipulation estimated to be shells-.26, bullets-.25, grenades-. 15. The
techmques used to obtain statistical estimates are of number of WIAs by type of munition was determined by
questionable reliability. multiplying the appropriate fraction by the total US Army WIA
figure of 167,771.
VIETNAM WAR MINE CASUALTY ANALYSIS
The Mine Warfare Center in Vietnam reported “An
important observation is that the share of casualties rep)orted as
caused by mines and boobytraps is believed to be well l^low the
actual fact. Specifically, it is believed that the classification of
“fragmentation” as the cause of casualties in many cases
obscures the fact that the fragmentation resulted from a mine or
boobytrap explosion. Infantry divisions in the field experience
a higher percentage of casualties resulting from mines and
boobytraps. ...figures released by the 1st Marine Division and
the 9th Infantry Division indicate that the mine and boobytrap
share of their casualties fluctuates between 40% and 60% of the
totals.”^^
Based on the primary source information provided by
the Mine Warfare Center in Vietnam, medical corps estimates
for mine casualties (contained in Assessing the Effectiveness of
Conventional Weapons, Table 2-15, page 66) appear to be
mistaken. There are a variety of reasons for the discrepancies.
By their own admission, the medical corps acknowledges that its
is often times difficult for medical personnel to properly identify
the cause of a casualty. It would appear to be for this reason that
U.S. CASUALTIES IN SOUTHEAST ASIA. Statistics as of
November 11, 1986, page 10 lumps mines and grenades
together and has a additional category labeled “multiple
fi*agmentation wounds.” The total number of KIAs given for
these two categories is 11,471. Considering the large number of
fragmenting mines encountered by the US Army in Vietnam
(40% of the total encountered)^^ and the fact that this type of
munition frequently causes multiple casualties, it would seem
reasonable to reevaluate the breakdown of the causes of
casualties in this light.
ESTIMATION METHOD
Total US Army combat casualties (KIA plus WIA)
were assumed to be 26,259 KIA plus 303,659 (total US DOD
hospitalized) times .65 (fraction of Army KIA of DOD total)
times .85 (fraction of combat casualties to enemy ordnance).
This produces an estimate of 194,030 total US Army combat
casualties and 167,771 total US Army WIA.
Based on information from the Mine Warfare Center
in Vietnam, total US Army mine casualties were estimated as
.33 (fraction of combat casualties to mines) times 194,030 total
US Army combat casualties. This produces an estimate of
64,029 total US Army mine casualties. Using the ratio .113
mine KIAs per mine WIA from the Mine Warfare Center’s data,
3-34
APPENDIX B
THE ORIGINS OF MILITARY MINES
Early mining techniques were developed in response mining. These mines were defined by the depth and size of the
to walled cities in the Middle East which were themselves charge as follows:
developed as protection against raiders and other threats. Jericho -For depths greater then than 3 meters, it was called a “mine”
is the oldest known example of a walled city (dating from -For depths less than 3 meters underground, it was called a
approximately 7,000 B.C.). Prior to the introduction of mining, “fougasse” (or contact mine)
the attacker’s options were limited to blockading the city (starve -When used as a “countermine” against an enemy mine, it was
them out), scaling the walls, breaching the walls with a battering called a “camouflet”
ram (which first appeared in Egypt about 2000 B.C.),or by -When intended to destroy an entire fortification (using 2,500
stratagem (for example, the Trojan Horse). Early mines were kilograms of powder or more), it was called “pressure balls”
used both offensively and defensively. Their use has evolved (globes de compression)^
considerably through the 3000 years of warfare since their
introduction. This evolution includes the emergence of explosive
TUNNEL MINES
The effectiveness of tunnel mines was significantly
tunnel mines, fougasse, self-contained mines (both antipersonnel
increased by exploding large charges of black powder at the end
and antitank), boobytraps, and sea mines.
of the galleries driven under fortifications. The first recorded use
of such a “mine” in Europe was in 1403 during a war between
THE EARLIEST MINES
Pisa and Florence, when the Florentines exploded a charge in a
The Assyrian Army, with its iron pioneer tools, is
forgotten walled up passage in the walls surrounding Pisa. In
credited with the first known use of mines in warfare. This
his famous work on siege warfare (published in 1770) Sebatien
occurred around 1000 B.C. and consisted of tunnels (mines)
Le Prestre de Vauban (French Marshal, 1630-1707) codified the
driven beneath the foundations of walls and fortifications.
principles of military mining that remained valid well into the
These mines could then be used by soldiers to gain access to the
nineteenth century. The number and locations of demolition
interior of a fortified area or they could be used to create a
chambers were dictated by the type of fortification. According
breach large enough for a full scale attack by collapsing a
to Vauban’s tables, explosive charges for mining could range up
section of a wall. This was done by excavating a chamber under
to 26,690 pounds. The purpose of the mine was not only to
the wall while bracing the ceiling with timber supports. These
cause destruction, but also - with the rocks and soil ejected - to
supports were then burned, which caused the collapse of the
form a breaching ramp that the assault troops could use.
chamber and the structure above it. Attacking soldiers then
Moreover, the demolition often came as a surprise to the
assaulted through the resulting breach. Many such mines have
defending forces, causing panic and confusion among the
been mentioned in history, notably the successful mines used by
Alexander the Great at the sieges of Halicarnassus (334 B.C.) defenders.
Tunnel mines were very time consuming to employ.
and Gaza (332 B.C.)^^ as well as Julius Caesar during the siege
Military mining during a siege could last 30 days or more.
of Marseilles in 49 B.C,^^ Effective mining (and other
Furthermore, specialists were required for the job. At first, coal
engineering skills) was critical to the military successes of both
miners were hired. It was not until standing armies were raised
men.
by the absolute monarchs of the 17th centuiy that formal mining
units were formed (1673 in France, 1683 in Austria). Their
EARLY OBSTACLES
work demanded courage and special caution. Lack of oxygen
During the siege of Alesia (52 B.C.), Julius Caesar’s
and possible flooding made their job difficult. Eighteen miners
engineers emplaced a 100 meter deep combination of towers,
and 36 unskilled workmen were normally employed in three
pallisades, ditches, abatis, and caltrops to slow down attacking
eight hour shifts to construct an assault mine.
Gauls so that they could be more effectively engaged by Roman
Mining was begun as soon as sappers had completed
missile weapons. These obstacles also gave Caesar the time
the last parallel in front of the glacis of a fortress or fortified
needed to successfully deploy reserve forces to threatened areas
town. The besieging miners dug galleries, about 1.25 meters
along his 13 mile long perimeter.^’ Other examples of early
high and 1 meter wide, lined with wood. Once the miners had
obstacles include the abatis emplaced by the English longbow
reached the selected site for the explosion, they dug out the blast
men for protection against the mounted French Knights at Battle
hole perendicular to the previous direction of the gallery. This
of Agincourt in 1415
mine chamber was then filled with the amount of black |x>wder
determined by the siege engineer.
EXPLOSIVE MINES
To ignite the mine, an ignition “sausage” was fed out
The advent of the capability to manufacture and
of the mine chamber. This sausage was a tube made of linen
detonate black powder (in Europe this occurred in the 14th
and filled with granulated powder, leading back to the point of
century) resulted in the next major improvement in military
3-35
ignition (minenherd). This ignition sausage was laid in 6 cm casualties as occurred during the sieges of Ciudad Rodrigo,
wide wooden duct, covered with a board, to protect it from Badajoz, and Santander in the Peninsular Campaign of the
moisture on the floor of the mine galleiy, or other damage. The Napoleonic Wars.
gallery was finally tamped with sod or earth, over a length of 6 Fougasses were employed by George Washington’s
to 10 meters. The miner ignited the granulated powder in the engineers at Forts Mifflin and Mercer on the Delaware River
ignition sausage with an ignition sponge at the appointed time, during our Revolutionary War. * * * The Mexicans also attempted
and then retreated quickly before the ignition sponge had burned to employ them on the approaches to Chapultepec during the
down to the granulated powder. Mexican-American of 1845.**^ Fougasses are still occasionally
Immediately after the explosion, the besiegers could employed by irregular forces, such as the Viet Cong,Central
assault the fortress or extend their sap trenches into the crater American guerillas,**^ and Bosniarfi"* that lack access to
and reinforce them with gabions. If necessary further mines modem landmines.
were used to take the pallisades of the covered passage, the
supporting walls of the counterscarp or the scarp thus facilitating SELF-CONTAINED ANTI-PERSONNEL MINES
entry into the fortress. Military engineers in China produced and employed
While working in the tunnels, attention had to be paid the first self-contained explosive antipersonnel mines for use
at all times to listening tunnels and the countermines of the against the Mongol invaders in 1277. These mines were
defender. The attackers tried to deceive the listening posts by manufactured in many shapes and sizes.
constructing phony galleries, in which workers produced a lot of The introduction of the first European flintlock in 1547
noise (noise gallery). was the basis for the first fuzed anti-personnel mine in the west.
As they became available, military engineers This was the fladdermine which was developed by Samuel
incorporated the latest technology from civilian mining, more Zimmermann of Augsburg in 1573.**^ It consisted of one or
eflScient munitions (picric acid in 1843, nitrocellulose in 1845, more pounds of explosive buried at a shallow depth in the glacis
and TNT in 1904), galvanic ignition (1860s), and ventilation of a fortress and was actuated by somebody stepping on it or
systems. During World War I, both sides employed mechanical activating a trip wire stmng low along the ground This released
tunnel boring machines. a flintlock igniter which fired the main charge. Like the
This type of tunnel mining has continued into the Fougasse before it, these devices were highly vulnerable to
modem era and has been used by Napoleon at Acre (1799) , in dampness and were only practical for use around fixed
the American Civil War (Vicksburg*®^ and Petersbur^®^), fortifications.
Russo-Japanese War (Port Arthur*®"*), World War I (Western The introduction of the explosive shell in the 1700s
Front*®^ and the Isonzo Front®^ ), World War II (Russian (1221 in China) and the percussion cap by the British in the
Front*®^, French-Indochina War (Dien Bien Phu*®®), and may be 1820s made possible the next important step in the development
employed by the North Koreans in some future war considering of mines. Confederate soldiers, under the leadership of General
that some tunnels under the DMZ (De-Militarized Zone) have Gabriel Raines improvised the AP mines from artillery shells
been discovered and more are suspected. near Yorktown, Virginia during the campaign of 1862. * By the
end of the Civil War, the Confederates had emplaced thousands
FOUGASSE*®^ of “land torpedos” around Richmond Virginia, Charleston South
Frederick the Great, King of Prussia, remarked Carolina, Mobile Alabama, Savannah Georgia, Wilmington
“Fougasses formed into a T like mine, in order to blow up the North Carolina, and Atlanta Georgia producing hundreds of
same place three times, can be added to the intrenchments. casualties. Their use was advocated by such famous soldiers as
Their use is admirable: nothing fortifies a position so strongly Robert E. Lee, John Mosby, and J.E.B. Stuart.
nor does more to ward off attackers.”**® These fougasses were The tretmine (step-on mine) was the next mine of this
simple black powder devices that were first developed for the type to appear. It went into production before World War 1.
defense of permanent fortifications. They were supposed to be However, the near domination of infantry by artillery and the
detonated in the face of an enemy assault. A black powder machine gun meant that the need for AP mines received little
charge was placed in a chamber excavated in the face of a attention from the warring powers.
fortification or in front of it. The chamber was then packed with The origin of each specific type of AP mine is
a large amount of fragments (normally just rocks or scrap iron). discussed under the appropriate heading in the text. The
If properly emplaced, it functioned as a crude claymore type antipersonnel mine reached full maturity during World War II
mine. The fougasse was command detonated by manually and has been a facet of almost every conflict since.
igniting a powder train from a protected position at the
appropriate time. Fougasses suffered from obvious defects, not ANTITANK MINES
the least of which was its vulnerability to the elements; even German combat engineers improvised the first antitank
moderate dampness would render the fougasse inoperative. In mines during WWI in response to the appearrance of the tank.
the right circumstances they could cause a large number of Initially, they used existing artillery and mortar shells with
3^36
sensitive fuzes. Later, they improvised wooden box mines, each in use appeared before or during World War II. One of the
weighed about 12 pounds and consisted of 20 powder charges most highly developed countermine organization was the British
of200 grams each. These were placed in boxes approximately 79th Armored Division which consisted of nothing but special
14X16X2 inches and were concealed about 10 inches deep. purpose armored engineer vehicles.
Detonation was caused by a hand grenade placed inside and
against one of the walls so that the primer passed through the SEA MINES
TTie Chinese first employed sea mines in the fourteenth
wall. It could function automatically as the tank passed over it
or by command detonation (which was greatly facilitated by the century. The oldest known European plan for a sea mine was
presented by Ralph Rabbards to Queen Elizabeth in 1574.
use of electric blasting caps which first appeared in 1900)."’
These AT mines were scattered at random to reinforce wire The first known employment of sea mines in the west occurred
obstacles and antitank ditches in front of the trench lines. The in 1777 when an American Army engineer, David Bushnell,
Germans also began to manufacture antitank mines in 1916 and attacked British ships on the Delaware River with floating
mines. Robert Fulton and Samuel Colt both became
produced almost 3 million before the Armistice of 1918.
The Germans developed and fielded the first full width interested in sea mines but lost interest when their experiments
attack mine toward the end of World War 11. It employed a tilt were not well received. Although, floating mines were used
rod and shap)ed charge kill mechanism. Improvised side attack during the Crimean War by the Russians in 1855 and at Canton,
China in 1857-58, their first significant employment occured
AT mines were first employed by the Germans and Russians on
during the American Civil War, where they were responsible for
the Eastern Front in World War IL Like the anti-personnel
mine, the antitank mine reached full maturity during World War most of the Union ships suilk.^’®
n and has been a facet of almost every conflict since.
antiaircraft mines
The first improvised anti-helicopter mines appeared
BOOBYTRAPS
The first explosive boobytraps were employed by the during the Vietnam War and were used to cover potential
landing zones. Many manufacturers now offer this type of
Chinese against the Mongols in 1277."® The first appearance of
mine. During the Cold War, the Russians developed an
explosive boobytraps in the West occurred during the Seminole
antiaircraft mine based on their SA-7/14.*^® The British and the
War of 1840."^ These were also employed in limited numbers
Americans are developing “smart” anti-helicopter mines that can
by the Confederates during the Civil War. The Confederates
be deployed to engage low flying helicopters.*’^ Some of the
employed a variety of devices including pull firing devices, timer
technologies being develop>ed for the Ballistic Missile Defense
run, and coal and wood “torpedoes” which detonated when
burned in a boiler etc. With the introduction of reliable Office could, in fact be consider orbiting mines.
mechanical devices during World War II, the boobytrap reached
full maturity and has been a facet of almost every conflict since.
COUNTERMINES
The original countermines were mines dug by the
defender to disrupt enemy mining efforts. Countermines were
employed frequently to defeat enemy mining efforts when they
were detected. Before the advent of black powder, a successful
countermine resulted in the interception of an enemy tunnel and
produced a confused, close quarters underground fight, as the
two sides fought for control of the tunnel.
John Vrano was the first to use black powder in a
countermine against the Turks during the siege of Belgrade in
1433."^ In this application, the intent was to dig down close to
the enemy’s mine gallery and emplace/detonate a charge that
would collapse his tunnel and kill the miners. During the Thirty
Years War, poisonous antimony gas was released into the
tunnels to kill the miners. The use of this t5q)e of countermine
has continued up to the First World War.
Modem countermine equipment first appeared at the
end of World War I as the British and French attempted to find
a countermeasure to protect their tanks from German antitank
mines. Except for some of the advanced electronic systems
cuirently in development, most countermine concepts currently
3-37
APPENDIX C
AP MINE EFFECTIVENESS
The recent debates on the effectiveness of AP mines where the limited number of AP mines available were employed
have focussed strictly on their significant attritional effectively, the attacking forces were either stopped (7th
characteristics. As shown in Appendix A, they have proven Armored Division) or severely retarded (1st South African
highly effective in this area.’^° In Vietnam, they accounted for Division). After the defeat of Rommel’s badly outnumbered
33% of US combat casualties. As a result of this, AP mines are Affika Korps at El Alamein, the skillful use of mines by the
often viewed as defensive weapons suited only for old fashioned German pioneers (combat engineers) to exploit terrain conditions
attrition warfare. Indeed, unskilled armies often use them in this was critical to the successful escape of the Afrika Korps from
fashion. There is also an important synergistic effect between Montgomeiy”s 8th Army.*^® Later, when faced with the
mines and other weapon systems. By slowing down and daunting task of defending the Atlantic coast, Rommel requested
channelizing the enemy, mines increase enemy exposure to the 50 million mines but had only received/emplaced about 5-6
direct and indirect fires that normally cover an obstacle. In the million by D-Day.Nevertheless, the critical landing on
same vein, AP mines are frequently used to protect AT mines Omaha Beach nearly failed because the US 1 st and 29th Infantry
from breaching by dismounted sappers. This technique was Divisions could not get off of the beach in j>art because of the
used by one USMC battalion during the Gulf War. AP mines.
Nevertheless, this narrow view overlooks the primary At Kursk, the Russians skillfully integrated AP mines
benefit of integrating AP mines with the combined arms team, within their defensive zones to separate the General Model’s
which is their ability to decrease the operational tempo of Panzers, particularly the Tigers, Panthers and Elephants from
dismounted enemy forces by undermining his moral. As their accompan}ang infantry and combat engineers, thus
General Patton observed, “The effect of mines is largely breaking up the German combined arms team. By slowing
mental.This is a very important effect because the US down the German advance, the Russian were able to mount
doctrine of victory through maneuver warfare rests on exploiting counterattacks that eventually halted the German spearheads.*^*
the enemy’s key psychological weaknesses. As Napoleon The fielding of long-range mine scattering systems has
remarked, “The moral is to the physical as three is to one.” A enhanced the ability of ground combat forces to use mines
key result of making the enemy fear mines is the increased time offensively, the decisive form of combat. These systems can be
it takes him to move. It has been estimated that for conventional used to separate dismounted elements from mounted units, to
maneuver units conducting mounted operations that 100 anti- delay the arrival of reinforcements and disrupt the
track mines per km of front decrease the rates of advance by s3/nchronization of an opponent. They can also be used to deny
40%, 500 anti-track mines per km by 50% and 1000 anti-track an opponent access to key facilities such as airfields and
mines per km by 60%. This gives US forces more time to Nuclear, Biological and Chemical weapons storage sites (as was
exploit fleeting opportunities as they appear on the battlefield. done during the Persian Gulf War) which could contaminate a
As Napoleon remarked, “The loss of time is irreparable in large area if they were attacked with precision guided munitions.
war.”*35
Considering that the only conflicts in which the US
Consider the following example, “(T)he road out of Army has not achieved a resounding victory in the 20th century
Normandy was indeed mined, considerably more than the were against dismounted opponents in Korea and Vietnam, the
Carentan area had been. I don’t remember anyone getting killed willingness of some individuals to deny the use of AP mines to
on the road, but we lost two trucks and a taiftc had to be our soldiers is morally and militarily questionable. Furthermore,
retreaded. The worst part of these explosions was that they our most likely military operations for the foreseeable future will
made us acutely aware of their potential and their probable be against just such dismounted opponents. As the great
numerousness. This was the excruciating aspect of those first military philosopher Carl Von Clausewitz noted “Kind-hearted
few days. You stare at the ground and wonder where not to people might of course think there was some ingenious way to
walk. What part of this dust, or this rich loam, carries death disarm or defeat an enemy without too much bloodshed, and
within it?”^^^
might imagine this is the true goal of the art of war. Pleasant as
Field Marshal Rommel, the great “Desert Fox,” it sounds, it is a fallacy that must be exposed: war is such a
understood the effects of mines better than most as he dangerous business that the mistakes which come from kindness
demonstrated in preparing for the Second Battle of El Alamein. are the very worst. The maximum use of force is in no way
“We wanted to ensure that the work of clearing the minefields incompatible with the simultaneous use of the intellect.”*^^
proceeded at the slowest possible speed and not until after our
outposts had been eliminated. Most of the mines available in
Africa were unfortunately of the anti-tank type, which infantry
could walk over without danger. They were, therefore,
comparatively easy to clear.” However, in some of the sectors
3-38
APPENDIX D
AP MINE INVENTORY PROFILES
3-39
ENDNOTES AND REFERENCES
fl 1 ‘hisVietnam. The Iraqis also attempted to exploit our vulnerability to mines but failed to secure their open
„. “Th® Vehicular Mine Threat,” by Harry Hambric and William Schneck. Proceeding ofthe Sixth Annual
i ARDEC Combat Vehicle Survivability Symposium (tn. May 1995, Volume 1, p^gcs 53-99. ~
2. Although mines are viewed by the Army as an "engineer or EOD problem," the vast majority of mine casualties occur in other combat arms and
support units. Anote item not reflected in these statistics is the synergistic effect of mines, whereby the presence (or fear) of mines slows down US
torc^Md increases their exposure and casualties due to other enemy weapons. See Antiarmor Tactics and Techniques for Merhan.V^t Tnfi.ntr', tC
7-24, Headquarters, Department of the Army, 30 September 1975, page E-4. --
Luntem^b^tyT Operations Desert Shield and Desert Storm," US Army Engineer School, Ft Leonard Wood, Missouri, July 1993, page
Milita^ Vehicles and Logistics 1991-1992. Jane's Defence Data, 1991, pages 143-206. It can be argued that advanced mines represent
, ® weapons, ^ey do not detonate until such a time as they can damage a target, are extremely difficult to counter, can be fitted with
IFF (Identify Fnend or Foe), self destruct/ neutralization, while minimizing the exposure ofthe emplacing force to enemy countermeasures.
6. "Ady^c^ in Mine Watfye; An Overview," by William C. Schneck, Malcolm H. Visser, & Stuart Leigh, Engineer Professional Bulletin, Ft
Leonard Wood, Missoun, Apnl 1993, page 3.
7. The section on mines is adapted from "Advances in Mine Warfare: Anti-personnel Mines," by William C. Schneck, Malcolm H. Visser & Stuart
Leigh, Engineer Professional Bulletin, Ft Leonard Wood, Missouri, August 1993, pages 26-33. ’
OP CIT Jye's Military Vehicles and Logistics, 1991-1992. pages 143-206. Mines rarely introduce "cutting edge" technologies, rather they
exploit technologies developed for other higher profile weapon systems.
Mme Recognition and Warfare Handbook. US Army Engineer School, Fort Leonard Wood, MO, November 1990. Contrary to USAF
^atements, attempts to breach minefields by bombing (including the use of Fuel-Air Explosives (FAE)) were not effective. It should be noted that
FAE was intended to expose shallow buried mines through their differential inertial response, and then deal with them by some other means such as
direct fire. See [Hie United States Army Engineer After Action Report for Operations Desert Shield and Desert Storm, page MobiHty-18 and "After
Action Report, Operations Desert Shield and Desert Storm," William C. Schneck, Belvoir RD&E Center, 12 November 1991, page 10.
10. Land Mines, TM 9-1940, Headquarters, Department of the Army, Washington, D. C., 15 July 1943.
yjet Cong Boobytraps, Mines, and Mine Warfare Techniques. TC 5-31. Headquarters: T^pjirtmpnt A^y pocfmber 1969.
Mine/Countermine Operations, FM 20-32, Headquarters, Department of the Army, September 1992, page D-18.
gisnificant Landmines and Booby Traps Employed by US and Allied Forces. 1940-1970. Landmine and Countermine Warfum Engineer
Apncy for Resources Inventories, Washington, D.C., June 1972, page 6. The first directional mine was prototyped by the Geimans in WWU see
Claymore Mines, Their History and Development, by Larry Gnipp, Paladin Press, Boulder Colorado, 1993, page 29.
15. The US defines the lethal range (or radius) of a munition as that distance where there is one letlial fragment (38 ft-lb) per square meter One
square meter is the approximate frontal area ofthe average standing soldier. Non-directional mines that produce fragments in all directions are defined
in temis of lethal radius instead of range. See also ^und Ballistics. CMH Pub. 81-34, Office ofthe Surgeon General, Department ofthe Aimy
Wasmngton, D.C., 1962, page 93.
Innn X 12mm mild steel cylinders (approx. 165 grains each) with a maximum velocity of about 4500 fos (about
4000 fps average) T^e MON-100 projei^ 400 lOmirt X 10mm mild steel cylinders (approx. 95 grains each). The MON-90 projects 2000 7mm X
7i^ mild steel cylin^rs (approx. 33 grams each). The MON-50 projects 485 5mm X 5mm mild steel cylinders (approx. 12 grains each ) or 540
spherical fragments. The existence of a MON-500 has also been reported. Classified mass/velocity profiles for the US M16 Ml 8 and M-26 are
W?ai p^^^^ Terminal Data for Surface to Surface Weapons OTl, FM 101-62-3, Joint Munitions Effectiveness
Equipment, TM 5-223C, March 1952, pages 123-129. This mine was fii^ encountered by the West during the “Phony
War m J[939 when French patrols began to suffer unexplained casualties. The French dubbed their nemesis the “silent soldier.” Engineers in Battle
by Paul Tiiompson, The Military Service Publishing Company, Harrisburg, PA, 1942, pages 64-65. ’
18. OP CIT, Mine Recognition and Warfare Handbook, page 196 & 197.
3-40
21 Kastem Europe. World War IT. Landmine and Countermine Warfare, Engineer Agency for Resources Inventories, Washington, D.C., August
1973, page 155.
22. OP CIT, Jane’s Military Vehicles and Logistics. 1991-1992, page 199.
23, The Soviet Armv: Troops. Organization, and Equipment, FM 100-2-3, page 5-166.
24 Gulf Air War Debrief. Edited by Stan Morse, World Air Power Journal, Airtime Publishing Inc., Westport, Connecticut, 1991, page 64.
25. OP CIT, ’’After Action Report, Operations Desert Shield and Desert Storm,”, pages 13-15. See also Multiservice Procedures for Operations.in
an Unexploded Ordnance Environment. FM 100-38, July 1996, Appendices C, D, & E.
26. OP CIT, Min. Recopnition and Warfare Handbook, pages 70, .80, 82, & 138-143. It appears that the UDAR is intended to breach its own
minefields to open up counterattack routes.
27. rhpmir.l Warfare in W»rM War I: The American Experience, 1917-191.8. by Charles Heller, Uavenwotth Papers No. 10. Combat Studies
Institute, Ft Leavenworth, Kansas, September 1984, page 20-21.
28. r^TT Miliran, Vehicles and Loristies. 1991-1992. pages 191 & 201. Germar^ al» pr^u^ (but did
during WWII. See Handbook on Oerman Military forces, TM-E 30451, War Department, Washington, D.C., 15 March 1945, p g
30. Principles of Improvised Explosive Devices, Paladin Press, Boulder, Colorado, 1984.
32. Engineer Contingency Handbook ^Former Yugoslavia), US Army Engineer School, Ft Leonard Wood, MO, 1993, pages 1-35 to 1-37.
34 North Korean People's Armv Handbook. FC 100-2-99, Battle Command Training Program, Ft Leav^worth, Kar^, 17 ^ril 1992, pages 10-27
to 10-36. See also North Korean Military Forces. FM 34-71, Headquarters, Department ofthe Army, 5 February 1982, page 13-4.
36 OP CIT "After Action Report, Operations Desert Shield and Desert Storm," William C. Schneck, Belvoir Research, Development md
Engineering Center, 12 November 1991, pages C-1 & C-2. The intelligence community actually estimated that approximate y 75 mme 1^ from 13
countries constituted a threat to US forces during the Gulf War, indicating how unsure the US is of many potential opponents imne mventones.
37. "Advances in Mine Warfare: Antitank Mines," by William C. Schneck, Malcolm H. Visser, and Stuart Leigh, Engineer Professional Bulletin,
November 1993, pages 38-45.
38 The type of mining employed by guerillas during OOTW may also be encountered by rear area units conducting conventional op^ations (for
example: Sng operations by French and Russian partisans against the Germans durmg WWII or by the North Koreans against the US dunng
Korean War).
39. Protective minefields are normally present around all defensive positions that have been occupied for more than a couple of hours.
40. A detailed discussion of cuirent US obstacle doctrine is beyond the scope of this paper. For more nnation s^ ^mbin^^O^ls
Integration, FM 90-7, Headquarters, Department ofthe Army, 29 September 1994, pages 24 to 2-7. The obstacle mtent must be confirmed through
war gaming.
41 Manually emplaced mines may also be encountered in nuisance minefields that are laid by withdrawing units or as part of cross-FLOT
Line of Troops) options. Additionally, many countries employ dummy minefields to stretch available mme stocks or confuse the enemy about
possible counterattack routes. For dummy minefields to be effective, the eneniy must be ’’sensitized" to the mme threat.
42. The best available references for mine warfare in OOTW are TC 5-31 and the four volum^ on Vietn^ in the Mine/^urte^ne
urban mining, see "French Army Combat Engineer Eiqjerience in Beirut, Lebanon, February 1985 Fr^chW^ College Report 01-85, and for
Mogadishu, "After Action Report, Operation Restore Hope," by William Schneck, Belvoir RD&E Center, 13 June 1994.
43. Although Vietnam occasionally had conventional features, it would be treated as an OOTW under current US Army doctrme.
44. Admittedly a rather generous description of the opportunistic route mining and other techniques they employed.
45 Rapnffr r.niinfprmin<^ fiiiide BRDEC Pamphlet 350-4, Belvoir Research, Development & Engineering Center, 30 November 1990, page 9.
46 Tt i. Pl.imed in landmines. A Deadly Legacy, page 155, that the "great majority" are antipersonnel. This may well be true in ru^e^untara
like Cambodia and Nicaragua where internal conflicts have been fought by dismounted soldiers but it is of questionable accuraiy for des^ " “
more advanced belligerentsusing motorized forces. This exaggerated claim lacks perspective and may be part of a disinformation campaign m support
3-41
of efforts to ban the antipersonnel mine. For example, see Tchad Libaa 1986-1988. Deminage-Depollution-Deoiegeage, 17eme Regiment du Genie
Pmachutiste, page 45. Over 80% of the mines cleared by the French engineers were AT mines. In conventional operations, the mix of AP mines in
US scatterable minefields typically ranges from 17% to 25%, depending on the emplacement system used.
47 pOD Scientific and Technical Intelligence (S&TI) Support to International Mineclearing Programs," Briefing by Tom Reeder, Foreign Science
and Technology Center, Charlottesville, Va., view graph #3.
48. Typical of downward ejected helicopter emplaced minefields like the other Italian systems, the US M-56, and the Russian PKPI.
49. Typical of vehicle dispensed minefields like Istrice, Minotaur, and Volcano.
50. Other MRLs in service/available that can scatter mines include the Astros II (Brazil) and one from Egypt. The Chinese have three other MRLs in
service that can lay mines (Type 74, Type 79, and Type 81).
51. The VS-MDH and S Y-AT systems are similar but carry 2080 or 3744 AP mines each respectively.
52. Minefield depth should be similar to that of the US M-56 system, about 40m. The typical length is about 270m.
56. Minefield depth is probably similar to the US GATOR system, about 200m. The JP-233 is intended primarily for runway denial.
57. 300 ordered by the German army. Mine densities vary from. 1 to .6 mines per meter of front. A UH-1 helicopter based system is also in
development. This system can lay a 500m minefield in 20 seconds.
58. Germany has procured 1000 dispensers and Italy 100 Jane’s Weapon Systems. 1987-88 Jane’s Defence Data, 1987, page 767.
60. Licensed copy of the US MLRS. 59 launchers have been ordered by the British army. 150 launchers and 350,000 AT-2 (DM-1399) mines in
MLRS rockets have been ordered by the German army. There are 28 mines per rocket, 12 rockets per launcher, & 9 launchers per battery.
61. GATOR density depends on whether the minefield is approached down the long axis or short axis. See FM 20-32 (1992) pages 6-11 to 6-13
See also FM 101-50-20, chapter 9. v -
62. Volcano minefields with a depth of 1.4 mines per meter are laid with a depth of320m.
41 systems were procured and deployed to Europe. They were consolidated in the Armored Cavalry Regiment and will be replaced by Volcano
The dispenser is carried by a UH-1 helicopter.
64. 69 procured and deployed with units in Europe and the US. Issued one per mechanized/armored division combat engineer company and
au^iented with the M-138 Flipper (174 procured) at a rate of 1 per platoon per company with GEMMS. Also issued separately to Airborne
engineers. It is to be replaced by Volcano.
65. These numbers represent averages over the battle area and include successive belts of minefields. Local densities could vary significantly.
67. 800,000 total mines (48% AT mines). This was 6 times the density employed in the battle for Moscow (Nov/Dec 1941) and 4 times tlie density
employed at Stalingrad (Oct/Nov 1942). See "Kursk, the Clash of Armor," by COL Koltunov, History of the second World War, Marshall Cavendish,
page 1381. ^
68. For Omaha Beach, these mines were concentrated at the critical exits from the beach. See Breaching Fortress Eurooe The Storv of 11 S
Engineers in Normandy on D-Dav. by Sid Berger, Kendall/Hunt Publishing Company, 1994. -^
From Shield to Storm, by James Dunnigan and Austin Bay, William Morrow & Co., New York, 1992, page 359.
71. OP CIT, After Action Report, Operation Restore Hope," page 56. The concepts for most of it dates from dates from World War I or II.
w' System Threat Assessment Report (STAR) For Family of Countermine Equipment (FACMS) (U). chanter 2. See also “The Vehicular
Mme Threat,” page 79 (Appendix E).
73. OP CIT, "After Action Report: Operations Desert Shield and Desert Storm."
3-42
74 Assuming the following standard equipment is on hand and the standard stockage of mines .s available m the division Additional support from
higher headquarters would increase these figures. The typical Russian UBL for mines is 13,000 to 20,000 conventional AT mines and 20 000 to
30 000 AP mines per division. These figures do not include artillery and aircraft scattered mines. This equipmert list is based on fte st^d^d Russian
or Motorized Rifle Divisions (TD, MRD) and the US Armored or Mechanized Divisions. Information for this table is taken from Battle Book,
ST iOO-3 Center for Army Tactics, CGSC, Ft Leavenworth, Kansas, page 6-1; Soviet Engineer Operations, Special Text Ft Leonard Wood,
Missouri '5 January 1990, pages 3-1 to 3-43; and Engineer Systems Handbook, pages 38 to 63. There would be a natural tendency to ej^aust the
stock of high tech mines early in a conflict, thus forcing units to employ lower tech stocks later in a campaign. The t^ical frontage for a Russian
MRD/TD attacking/ defending, 20 km and 20-30 km respectively. Minefield output for the Russians is based on 750 anti-track mines per ^and 750
mines per kilometer of front (US) with a 50/50 mix anti-track/anti-hull. See also Countermobility, FM 5-102, Headquarters, Department of the ^y,
Washington, D.C. 14 March 1985. Other types of manmade obstacles (wire entanglements, blown bridges, road craters, log obstacles, etcO ^e also
possible (see Engineer Field Data, FM 5-34, Headquarters, Department of the Army, Washington, D.C., 14 September 1987, pages 3-1 to 3-15)
however tlie probability of encounter is difficult to estimate. Although wire obstacles are generally considered to be antipersonnel m nature Aey are
also capable of impeding vehicular movement. The effects of natural obstacles are covered in FM 90-7 (Final Draft), 1977, pages 3-5 to 3-19.
76. The 3 combat engineer battalions assigned to US mech/armored divisions are capable of laying approximately 96 km of minefield per day but
logistic capacity within the division will limit this to 75 km without significant corps level augmentation.
77. The Russians carry three loads per GMZ, UMZ, & PMZ-4.
78. A logistic surge of four reloads per day would yield 26.4 km per day.
79 The Russians typically attempt to keep 1.0 km of minelaying resources in reserve within a POZ (mobile obstacle detachment). The Russia otc m
the process of replacing the PMR-3s with GMZ armored mineplanters. The typical Russian MRD/TD will be able to mechanically emplace minefields
at a rate of 11 km/hr.
80. Assumes a UBL of456 rounds of RAAM per 155mm howitzer bn (battalion) (3 bns per division), 8 mines per round, the normal CSR
(Controlled Supply Rate) is 8 rounds per bn per day during offensive operations and 16 rounds per day during defensive operations, ^girtic surge
capability or shortages could significantly alter this. The US is planning to employ WAM from either standard MLRS rockets or by ATACMS.
81. It is believed that the VMR-1 and VMR-2 have been retired from front line Russian divisions but may still be found in some client states.
82. The majority (nearly 100%) of the mines laid by the US and UK during the Persian Gulf War were by fix^
dropped approximately 22,000 GATOR mines on 50 sorties during the first week of the air campaign. The US Navy also used GATOR to block the
IraqiVetreat from Kuwait along the "Highway of Death” to establish this large kill zone at the end of the ground phase^ ^ w u
approximately 43,000 HB-876 on 100 sorties during the first few days of the air campaign. See Gulf Air War Debrie£ Edit^ by St^ Morse, World
y^r Power JoLal, Airtime Publishing Inc., Westport, CT, 1991, page 218. CMS Inc. reported clearing 185 ^^AM and 7^ RAAM mmes m ^
sector of Kuwait. This included the area through which the USMC maneuvered during the ground war. See Unexploded Ordnance (UXO) Study.
83 Vietnam Lessons learned. 1965-1968. Landmine and Countermine Warfare, Engineer Agency for Resource Inventories, Washington D.C., July
1972. Some veterans of this conflict have argued that the US is in fact worse off in this respect as a number of specialized it^s and traimng couraes
used in Vietnam are no longer available. Currently there is little protection available to the "light” forces selected for operations other ton war. A few
special purpose mine resistant vehicle kits have been made available, primarily to special operations forces. Several mine resistant vehicle progr^
are cuirently underway. Their products are designed to significantly reduce crew casualties m both OOTW and conventional operations on a nonlinear
battlefield. Some of these programs have demonstrated great promise.
84. Western Europe. World War II. Landmine and Countermine Warfare. Engineer Agency for Resources Inventories, Wash. D.C., 1973.
85. Advances in C4I (the "Information War") will mean little if the US can not maneuver to exploit it.
86. Infernal Machines. The Story of Confederate Submarine and Mine Warfare, by Milton F. Perry, Louisiana State University Press, 1965, pages 22,
166, and 205 (chapter V, note 1).
87. Historical Report of the Chief Engineer. American Expeditionary Forces. 1917-1918, Washington, D.C., 1919, page 241.
88. Mine/cnuntermine Onerations at the Company Level. FM 20-32, US Army Engineer School, Ft Belvoir, VA, September 1976, page 135.
89. I Inexnloded Bomb. A History of Bomb Disposal, by A B. Hartley, W. W. Norton & Company, New York, 1958, page 237.
90. Demining Operations rinitial Draft-). TC 31-34, Headquarters, Department of the Army, Washington, D. C., July 1996.
91 And possibly stretch our limited countermine assets beyond the breaking point and producing battlefield paralysis.
92. “Mine Warfare in Vietnam,” The Mine Warfare Center, Engineer Section, Headquarters, US Army Vietnam, August 1969, page 66.
94. The Art of Warfare in Biblical Unds. Volume 1. by Yigael Yadin, McGraw-Hill Book Company, New York, page 317.
3-43
Be Generalship of Alexander the Great by J.F.C. Fuller, Rutgers University Press, New Brunswick, N. J., 1960, pages 200-218.
War Commentaries of Caesar, by Julius Caesar, translated by Rex Warner, The New American Library, 1960, pages 259-266.
The Face of Battle, by John Keegan, Penquin Books, 1976, pages 90-91.
99. Siege, by Gert Bode, International Military and Defense Encyclopedia, Volume 5, Brassey’s Inc., Washington, D.C., 1993, page 2421.
A Manual of Siegecraft and Fortification, Sebatien de Vauban, translated by George Rothrock, The University of Michigan Press, 1968.
102. “Engineer Operations During the Vicksburg Campaign,” by Robert Puckett, AD-A255-141, Ft Leavenworth, KS, 1992, pages 124-132.
Thg Siege of Petersburg, by Joseph P. Cullen, Eastern Acorn Press, 1970, pages 17-23. The mine exploded by Federal troops under the
Confederate earthwork at Elliot’s Salient at Petersburg, Virginia, on 30 July 1864 was charged with 8,000 pounds of powder and produced a crater 9
nieters (30ft)deep, 18 meters (60ft) wide, and 52 meters (170ft) long. The subsequent Federal assault, however, was unable to exploit the temporary
advantage gamed by the explosion and the surprise. ^
The History of Fortification, by Ian Hogg, St. Martin’s Press, New York, pages 185-189.
Underground, The Tunnellers of the Great War, by Alexander Barrie, Tom Donovan, London, England, pages 243-261. On 7 June 1917
Bntish engineers fired nineteen mines with 430 tons of Ammonal at a depth of 40 meters at Wytschaete Salient south of Ypres destroying three
German battalions.
106 13 March 1918 Austrian engineers blew up part of Mount Pasubio, which was occupied by the Italians, using 50,000 kilograms r55tonsl of
explosive killing 485 men. ^ \ j
107* Small, Unit Actions During the German Campaign in Russia, CMH Pub 104-22, Center of Military History, Washington, D.C., Facsmimile
edition 1988, pages 165-168). The “Blitzkrieg” oriented German Army of WWII maintained special “Minier Pioniere” units throu^out the war
(Piopiere, Entwicklung einer Deutschen Waffengattung, by Dietrich Petter, Wehr und Wissen Verlagsgesellschaft MBH, Darmstadt, Germany, 1963,
Hell in a Very Small Place, by Bernard Fall, J.B. Lippincott Company, 1966, pages 384-386.
109. Not to be contused with the improvised flame mine that US Army engineers occasionally employ and call a ‘Tougasse.”
Frederick the Great, On the Art of War, by Frederick II, edited and translated by Jay Luvaas, copyright 1966, The Free Press, New York, page
M Engineers of Independence, A Documentary History of the Army Engineers in the American Revolution, 1775-1783. by Paul K Walker
Historical Division, Office of tlie Chief of Engineers, Washnigton, D.C., page 158-159. Such devices are still occasionally encountered in modem
warfare for example, see Engineer Field Manual, Parts I-VII. Professional Papers of the Corps of Engineers, U.S. Army, No 29 Washnigton. D C
1 ^2. Tlie War with Mexico, by Donald Chidsey, Crown Publishers, New York, pages 161-163.
^ Ayiidas de Instruccion Contra Minas, Trampas Y Artefactos Explosivos. Guatemalan Corps of Engineers, undated, page 60.
The Genius of China, by Robert Temple, Simon and Schuster, New York, 1986, page 235-237.
^Eees Lieutenants, Volume 1, Douglas Southall Freeman, 1942, pages 268-269. See also Southern Historical Society Papers. Volume III.
MMary to June 1877, Broadfoot Publishing Company, 1990 edition, pages 38-39. The sheels used were ordinary 8 or 10 inch mortar or columbiad
^ells. Such “land torpedoes” were also employed around Atlanta, Fort Mcallister, Fort Wagner S.C, Fort Fisher, N.C., and Richmond (Battery
Hamson and Ft Gilmer). Sherman’s troops also employed mines around damaged railroad supplies to prevent their salvage during the Atlanta
^ The Fighting Tanks Since 1916, by Ralph Jones, George Rarey, and Robert Icks, The National Publishing Company, Washington, D.C., 1933
page 262. The British also employed antitank mines. These consisted of lengths of pipe filled with explosives. It worked quite effectively when
buned across a road. See FM 20-32 (1976), page 133. n j
118. ^cjent Inventions, by Peter James and Nick Thorpe, Ballantine Books, New York, page 207.
Southern Historical Society Papers, Volume X, January to December 1882. Broadfoot Publishing Company 1990, pages 257-260.
120. Medieval Warfare, by Terrence Wise, Hastings House Publishers, New York, 1976, pages 168-169.
3-44
121. OP CIT, The History of Fortification, pages 99-100.
122. “Mine and Countermine in Recent History, 1914-1970,” by Russel Stolfi, BRL Report 1582, Ballistic Research Uboratories, Aberdeen
Proving Grounds, Maryland, April 1972, page 91.
123. Vanguard of Victory, The 79th Armored Division, by David Fletcher, Her Majesty’s Stationery Office, London, 1984.
126. OP CIT. Infernal Machines. The Storv of Confederate Submarine and Mine Warfare, page 4.
128. On Air Defense, by Janies D. Crabtree, Praeger, Westport, Connecticut, 1994, pages 183-184.
129. OP err, Jane’s Military Vehicles and Logistics. 1991-1992, page 198 and “Anti-Helicopter Mine” Brochure, Ferranti Inter., undated.
130. The Russians have even developed formulas for calculating the casualty causing potential of a given minefield. See Soviet Engineer Operatioiis,
Special Text, US Army Engineer School, Ft Leonard Wood, Missouri, 5 January 1990, pages 3-5 to 3-7.
131. “The 3d Marines in Desert Storni,” by John Admire, U.S. Marines in the Persian Gulf. 1990-1991: Anthology and Annotated BibliograjM
compiled by Charles Melson, Evelyn Englander, and David Dawson, History and Museums Division, Headquarters, USMC, Washington, D.C., 1992,
page 168.
132. “Defeating the Enemy’s Will: The Psychological Foundations of Maneuver Warfare,” by David A. Grossman, contained in Maneuver Warfee
Anthology, edited by Richard Hooker, Presidio Press, 1993, pages 142-190.
133. War as I Knew It By George Patton, Pyramid Books, New York, 1972 edition, page 350.
134. Numbers. Predictions. & War, by T. N. Dupuy, Bobs-Merrill Co. Inc., New York, 1979, page 215.
135. The Campaigns of Napoleon, by David Chandler, MacMillan Publishing Company, New York, 1966, pages 149 and 155. For a more recent
perspective on tlie criticality of time in modem warfare, see Fighting by Minutes, by Robert Leonard.
136. Patton’s Best. An Informal History of the 4th Armored Division, by Nat Frankel and Larry Smith, The Berkley Publishing Group, New York, 1978,
pages 17 and 18.
137. The Rommel Paoere. edited by Liddell Hart, Harcourt Brace and Company, New York, 1953, pap 300. The original plan called for 25% AP
mines. RommePs engineers were only able to emplace about 14,000 AP mines (3%) of the 500,000 mines emplaced.
138. N»rth Afriea T ^iHmine and Countermine Warfare. 1940-1943. Engineer Agency for Resources Inventories, Washington, D.C.. June 1972,
pages 103-160.
139. Western Europe »snrtmine and Countermine Warfare. 1940-1943. Engineer Agency for Resources Inventories, Washington, D.C., July 1973,
page 9.
141 Citadel The Battle of Kursk, bv Robin Cross, Sarpedon, New York, 1993, pages 164-167. For information on Russian preparations for Kursk,
see “Soviet Defensive Tactics at Kursk, July 1943,” by David Glantz, CSI Report No. 11, Conibat Studies Institute, Ft Leavenworth, Kansas, 1986.
Another good account of German engineer operations at Kursk is contained in “If You Don’t Like This, You May Resign And Go Home:
Commanders’ considerations In Assaulting A Fortified Position,” by Michael Woodgerd, AD-A244-373, Naval Post Graduate School, Monterey,
California, pages 52-82.
142. On War, by Carl Von Clausewitz, translated by Michael Howard and Peter Paret, Princeton University Press, Princeton, New Jersey, 1976
edition, page 75.
143. Tliis data is based on the frequency for which a given mine was mentioned in Hidden Killers, The Global Problem with Uncleared Landmines,
Department of State Publication 10098, July 1993. Admittedly, not very scientific, but this is the most comprehensive data avatlable.
144. The New York Times gave the following figures for mines presently emplaced around the world: L-9 Barmine, 6 million, Kuwait, Iraq; VS-2.2,
10 million Afghanistan, Iraq, Iran, Kuwait; PT-Mi-Ba-III, 11 million, Iran, Iraq, Kuwait, Mozambique, Somalia; Type 72A, 20 million, Af^ianistan,
Angola cLnbodia, Iraq, Mozambique, Somalia, Thailand, Kuwait; M-18,6 million, Angola, Mozambique, Central/South America; POMZ-2,16
million, worldwide; PMN, 20 million, worldwide; PRB 409,11 milUon, Afghanistan, Iraq, Iran, Mozambique, Somalia, Lebanon; "New York Tunes
Magazine," January 1994.
3-45
3-46
The Proliferation of the
Mine Threat
and the PRORYV System
The use of mines as an essential part of arsenals around the world is expanding. Atjewildenng
array of mine types floods world arms markets these days. Countries without the means to
directly engage the armed forces of the US and its allies recognize that the use of mines is a most
efficient and effective means of limiting our mobility of forces. In order to retain our mobility and
preserve our mission capability in spite of the rapid proliferation of mines, we must continually
develop mine countermeasures systems which are responsive to the widening threat. Finally,
when a conflict is ended, the battlespace is likely to be littered by mines. Some naval mines will
sterilize (self-destruct) when their timers run out or their batteries die. Naval contact mines and
land mines rarely have any mechanism for sterilization. These mines pollute lines-of-
communication years beyond a conflict.
Since the industrial revolution, land and sea forces have made great strides in developing weapon
systems and tactics that allow higher and higher levels mobility and flexibility. The US has made
massive capital investments to make forces that are extremely mobile and flexible. Mobility and
flexibility give the armed forces of the US the ability to project massive amounts of fire power
around the world, within a matter of days. It is essential, therefore, for countries attempting to
act contrary to the interests of the US and its allies to be able to limit the mobility and flexibility of
our forces. A significant response to these increases in mobility and flexibility has been the
proliferation of countermobility technology, primarily mines.
Mine manufacturers build mine types based on the features of their desired targets. Fusing,
warhead size and method of employment are all in response to the target. There are four basic
targets for mines;
• Personnel,
• Tanks,
• Naval Ships.
3-47
Killing targets outright is not necessarily the intent of the miner. Rather, miners nominally use
their weapons to inhibit movement by presenting a credible threat to the combat efficiency of the
target. This philosophy of use has evolved from the concept of mass barrage mining. Defending
forces will place a mass barrage of mines across the extent of the front that enemy forces could
exploit. Mines designed to kill their targets outright are considerably larger than the sort that
cause minimum effective damage. The larger mines require more logistics support, more laying
personnel and more time to lay than smaller mines. Since the goal is area coverage to reduce
mobility, the smaller mines placed at higher rates are preferable. This concept has resulted in
mines that are smaller, lighter and easier to use than their W.W.II counterparts. The smaller
mines are also cheaper and used more freely.
In addition to mines that users can deliver easier and faster, Current mine technology has also
provided mines that are more resistant to countermeasures. Mine manufacturers have
accomplished this by technologies ranging from the use of materials resistant to current detection
means, to the use of microprocessors to achieve better target discrimination. Ironically, as mines
become a more complex problem for mine countermeasures assets, they are becoming easier to
prepare and deploy requiring very little specialized training for personnel trained in ordnance
handling.
The mine-proliferation problem is not an abstract one based on some future conflict. It is a real
problem that currently faces us right now. Antipersonnel (AP) and antitank (AT) mines sell for
pennies apiece. The International Red Cross (IRC) and UNICEF estimate that there today there
are about 110 million AP mines on the ground. Additionally, mines cause injury or death for over
70 people every day. According the IRC, there are 14 countries that have enough mines on the
ground to present a serious threat to the civilian population.
Naval mines sell for between $1,000 and $2,000,000 depending on the complexity and size of the
system. Mines ranging from simple contact mines to complex “rising” mines are available on
world markets. These mines potentially present a grave risk to naval assets. In terms of naval
mines, since the end of WWII, the world has seen at least seven major mining incidents. The US
Navy has had a total of seventeen ships damaged due to weapons other than guns. Of the
seventeen ships damaged, mines damaged fourteen ships. All of the damage was extensive. Some
damage was (like the USS WARRINGTON with 35 dead) tragic. In that same time frame, mines
have struck at least 25 noncombatant ships.
There are at least 24 mine exporters in the world; and an uncounted number of countries have
acquired the technology base required to produce their own mines. The size of mine systems
makes them easy to move and extremely difficult to detect before the first firing occurs in a mine
field. As the proliferation of mines expands and the number of producer nations increases, the
knowledge-base of the US MCM forces lags further and further behind. To be able to respond to
the mine proliferation problem, MCM systems must be robust enough to anticipate as yet
unconfirmed mine technologies. Mine proliferation requires an ever-expanding effort at acquiring
and exploiting the mine systems. Mine acquisition and exploitation are the only means by which
we can determine true mine performance values and develop the tactics and methods to counter
those systems. The ability to counter mines may be critical to our ability to operate our forces
during conflicts and protect civilian lives after the conflict.
3-48
MCM System: The MCM system described has been referred to as PRORYV.
The system has a footprint of 5m x 10m and weighs about 20t. The launchers consist of centered
stabilization rail and four rockets in a cluster. A laser designator (of unknown type and
frequency) is used to designate the lane. The fire-control system commences fire based on two
measurements. The first measurement used is the relative position of the launch platform behind
the target lane. The second measurement is that the craft is in a horizontal position and pointed
toward the target. The fire-control system fires front to rear. Using a speed of advance of 10
knots, the system fires a row of weapons at the target at a rate of 1 per second.
The launchers are set at 45 degree angles and are in fixed positions to provide a 40m-wide spread
of weapons. The claimed 3-sigma accuracy of the launch system is plus or minus Im in waves of
1.5m or less.
The ordnance is a rocket which is electrically ignited and uses a powder propellant. The rocket is
Im in length and carries a 35kg warhead. The explosive portion of the warhead is 25kg of
enhanced explosive with a TNT equivalent of 1.6 (40kg or 88 lb. of TNT). Fusing of the
warheads is accomplished by an inertial switch and a pmr of timers. When the rocket is fired, a
5-second timer begins to run. The 5-second timer is a sterilization or fail-safe timer. When the
warhead contacts the water surface or land, a second timer starts to fire &e warhead when it falls
2.5m. Should the inertial switch or the second timer fail, the 5-second timer will foe the
warhead. The claimed minimum kill radius for the warhead is 5m. A 5m kill radius has been
verified against TM-46 series A/T mines and L’DINA bottom influence mines.
Other factors: According to the Deputy Director, the fire-control system and ordnance are
adaptations of the very mature ASW systems currently produced and under development. During
a visit to MODI, the deputy director of MODI asked the Director of the development lab if the
fact that the system has been developed to operate firom the Russian conventional landing craft
would impact development of an LCAC-based system. The Director replied that it would not
because the system was designed to be autonomous from the host platform, the LCAC has plenty
of space and weight capacity, and that the LCAC would probably be more stable than the landing
craft as a launch platform.
A further conversation between the Director and Deputy Director of die Lab revealed that their
main placement error for the system would likely not be as a result of the sea state or other
environmental factors. Rather, it would be caused by the steering error of the Coxswam.
At the presentation of the DET/SABRE system at MODI, the Deputy Director and a lead scientist
' at the lab began a conversation of their review of DET-like and SABRE-like systems. There was
a great deal of noise in the room so that comprehension of the conversation was fir^ent^.
However, the lead scientist indicated that the net-system has been too heavy (mentioned the
number of 3t) and that it had been impossible to control the linear charge with sufficient
accuracy. Therefore they had decided to use individual weapons in the configuration shown to
SESSION ON THE LITTORAL ENVIRONMENT
MONTEREY BAY AQUARIUM
The physical and biological environments are dominating factors in the harnessing of
technology to deal with the "Mne Problem" — including the offensive/defensive use of mines for our
own purposes. The Symposium took advantage of the proximity of the Naval Postgraduate School
to the Monterey Bay Aquarium to raise the awareness of participants about the complexity of the
littoral environment.
There were three components to this Session. At the beginning were the formal presentations
by the Oceanographer of the Navy, RADM Paul E. Tobin, USN, and by members of a specially
constituted Very Shallow Water Mine Countermeasures Unit led by Capt. Thomas R. Bemitt, USN,
Commander of Explosive Ordnance Disposal (EOD) Group One. Bemitt’s team consists of specially
trained EOD, Navy Special Warfare (SEAL), and Marine Corps Combat Reconnaissance specialists.
Among other duties, this group has operational cognizance over the marine mammals used in Mine
Countermeasures. Admiral Tobin’s paper and the briefing by Capt. Bemitt and his colleagues are
reproduced here.
The initiative by th? Navy's Mine Warfare Command to establish a Very Shallow Water Mine
Countermeasures Unit that includes Marine Mammals (dolphins) is another example of "combat
oceanography.” Capt. Bemitt and his associates described their program, in which they have begun
to explore the operational aspects of VSW MCM with an emphasis on being able to conduct
clandestine reconnaissance of Amphibious Objective Areas (AOAs). They intend to be able to map
the locations of mines and obstacles, and also want to establish the nature of the bottom in terms of
traflBcability for tracked amphibious vehicles.
These speakers emphasized the nature of the VSW MCM environments. Even a seemingly
calm water surface can mask current and tidal surges, and those surges can involuntarily move a
swimmer as much as twelve feet. A swimmer immersed in such a water environment cannot control
his movements. In addition, there are difficult conditions of visibility, variations in the nature of sea
life and sea growth, and the possible use of "stealth" shapes for mines.
The exhibits at the Monterey Bay Aquarium served to show attendees just what a kelp forest
looks like, and what some of the underwater habitats can contain. Also, one must look at sea life as
potential false, non-mine targets for acoustic underwater detection devices.
3-51
This event provided graphic illustrations of the very real, operational problem facing mine
hunting and mine reconnaissance — finding the desired target in the presence of a great many false
targets. The problem is further exacerbated by mine burial, which can occur by the action of currents
and tidal surges.
The formal presentations concluded with a brief introduction of the capabilities of the marine
mammal systems. These dolphins have shown themselves to be remarkable, and still the most
sensitive, tools for finding and classifying sea mines -- including buried sea mines. (As a parenthetical
note, it is significant that the best landmine detection systems are also biological systems — dogs and
pigs. Reserach interest in just how these land mammals are able to detect mines is growing. There
are those who believe that the odor sensing capabilities of animals contain the key to mine detection,
both on land and at sea.
At the conclusion of the formal presentations, attendees visited some mine-related technology
displays provided by our Corporate Sponsors for this Session, and also visited the various tanks at
the Aquarium containing living examples of just what the littoral environment is like. Symposium
organizers believe that the images of the flora and fauna of the littoral environments, together with
the descriptive narration of the problems of Mine Countermeasures in Shallow Water, will be of great
utility to researchers as they search for technological answers to the Mine Problem.
This unstructured period also provided a relaxed opportunity for attendees to discuss what
they had heard from our distinguished speakers earlier in the day, and to interact with those speakers
and other senior personages who were present.
The most humbling lesson of the Symposium was that, despite real and significant
technological strides, porpoises (dolphins) remain by far the most effective and fastest detectors of
seamines, especially buried ones, to date; and dogs remain by far the most effective detectors of
landmines. A recent scientific report stated that the dog’s nose/olfactory system is a million times
more sensitive than that of the human’s. In South Afnca, where trained dogs have long been used
for landmine detection, they have the highest success rate — over 90 percent — of any method yet
tested, and dogs were successfully, and safely, used for mine detection operations in Bosnia.
Similarly, a trained dolphin is able to use echolocation/identification (acoustic sensors) to distinguish
between otherwise identical metalic mine-like objects whose only difference is a fraction of an inch
in thickness; can differentiate size, structure, shape and material composition of submerged objects
at incredible distances; and can complete a seamine search of a 2-square-kilometer area of ocean in
only 4 hours, compared to 14 hours by a dedicated Navy vessel going at 3 kts. Indeed, the porpoise
consistently outperforms any available artificial system, especially in poor acoustic environments, in
high-clutter areas, and where objects are buried beneath the ocean bottom — all cases where detection
capability is often the most needed. These highly intelligent sea mammals can also be used to develop
systems which work in shallow water/surf zones — by reverse engineering the dolphin’s sensory
capabilities, future AUV systems may be designed and built which begin to match their natural
capabilities. Clearly, we still have much to learn from humankind’s two “best friends.”
3-52
Developments in
Rapid Environmental Assessment
I do not presume to be an expert on mine warfare, but like almost all naval
officers, I have had many experiences that have been affected by the use or threatened use
of mines. In my ciurent assignment as Oceanographer of the Navy, I have been
emphasizing closer ties to the tactical commander and the operating forces. Rapid
Environmental Assessment is one of our highest priorities, and the Mine Warfare threat is
always an important part of any assessment. This is particularly true in the littoral.
My involvement with Mine Warfare started in Viemam in 1971 and 1972, where
we lived with the threat in the rivers and the deltas and in late 1972 undertook our last
major offensive mining operation in Haiphong. While I was Commanding Officer of
SWOS, I developed a strong relationship with the USS SAMUEL B. ROBERTS and
spent several days at sea with Captain Paul Rinn and his fine crew. Obviously we viewed
the subsequent mining of the S. B. ROBERTS in the Gulf with great concern and interest.
While in Subic Bay as Commander Task Force 75,1 developed a similar relationship with
Captain Ted Hontz of the USS PRINCETON and saw first hand the major damage caused
by a mine to that great ship. In that same job I oversaw the operation of the Mobile Mine
3-53
Assembly Groups in the western Pacific and learned a great deal about our offensive
capabilities and the very talented specialists who work in that field. We also had the
opportunity to host the Japanese Self Defense Force Mine Warfare squadron that was the
first Japanese Naval Force to deploy from home waters. In this case they were on their
way to make a contribution to the Gulf War. It is interesting that this particular force was
chosen to introduce a new dimension to Japanese Maritime Self Defense Force
operations. Finally, in my last assignment with the Chief of Naval Education and
Training, I worked closely with the Mine Warfare Training Group in the planning and
execution of the move fi-om Charleston to Engleside. I have not yet had the opportunity
to see the new facilities but hope to visit RADM Conley early next year.
The impacts of Mine Warfare in World War I, World War II, Korea, Vietnam, and
now the Gulf War are well documented. Mines continue to be a low cost, very effective
way to drastically slow a show, if not stop it. Weapons that can be effectively delivered
by a range of platforms fi'om rowboats to ballistic missiles pose a very real threat, and
unfortunately, this technology is accessible to almost any nation.
Oceanography has come a long way in solving this data base problems. From
man’s first ability to calculate latitude by the Greek Pytheas, who ventured as far north as
Iceland in 325 BC, to the voyages of Captain Cook in 1768 that used the Harrison
chronometer and accurately calculated longitude. In 1872 the voyages of H.M.S.
CHALLENGER established modem oceanographic science with detailed collection of
water column data and bottom samples around the globe. The explorations of GLOMAR
CHALLENGER in 1968-1983 yielded detailed core samples fi^om the ocean sediments
and a clearer picture of the ocean bottom topography. Side scan sonars, multi-beam
swath bathymetry and now perhaps, the most dramatic improvement of all, the Global
Positioning System, have finally allowed us to obtain a view of the ocean bottom.
3-54
Bob Ballard states the issue well in noting that when Neil Armstrong set foot on
the moon we had yet to discover the true extent of the largest geographical feature of our
own planet — the mid Atlantic ridge. In fact, the first real global representation of the
ocean floor developed by Bruce Heezen and Marie Tharp at the Lament Doherty
Geophysical Laboratory were not published until 1972.
This brings us back to the all important matter of defining the ocean floor data
base in the littoral as a key factor in mine warfare. With the new tools I have described,
we can survey the ocean bottom and construct computer generated images on the fly and
make analyses in real time with precise navigational accuracy. How are we doing? Even
though the demise of the cold war was provided us with unprecedented access to the
world’s littoral waters, we still have a survey backlog of some 360 survey ship years.
Further, as RADM Williams often reminds me, our pictures are not quite accurate
enough. We really need to know whether that new metal object is a refiigerator, an
abandoned car, or an influence mine. The tools to make these determinations may soon
be at hand, but presently we still have a need for time consuming swimmer or ROV
searches.
GPS accuracy to less than 1 centimeter is possible today, and new bottom
scanning technology coupled with sophisticated data base comparison software will
dramatically speed the survey process. Airborne LASER Bathymetry is a new fast way
to acquire clear water gross survey data in very short periods of time. Will we ever have
a clear picture of the bottom and slightly below? It is not a question of if, but rather of
when, and this will depend on the seriousness of our commitment. Unfortunately, today
we are the last remaining nation to be mounting a world wide survey effort. Our eight
ship fleet will be sorely taxed to gather the data that technology and world politics are
making available.
Conferences such as this, coupled with new parmership initiatives, like the
National Oceanographic Parmership Act and active support of Oceanography education
programs, will move us in the right direction. Public awareness of the challenges of
Oceanography, our last great firontier on earth, is growing, but the U.S. Navy’s interest
has and will continue to be long-standing. It is a core competency of our profession and
can make the difference between victory and defeat in modem warfare. I assure you that
the Oceanographer’s office will continue to work closely with the mine warfare
community in what is very clearly in both of our best interests. I appreciated the
opportunity to be with you tonight and look forward to our mutual challenges ahead.
3-55
3-56
Developments
in the
Very Shallow Water-Mine Countermeasures
Test Detachment Program
3-57
ery Shallow Water-Mine
3-58
• USMC Recon
utline
xn
xn xn
ssues to be Resolved at a
3-61
■ Escape and Recovery
■ Neutralization
°8
Ps! ?
hain of Command
o>
D E
OS
W M
O O
u u
" II
So
Q
< O
urrent Manning
^ CD ^ ^
■O In ^
-t « B
<n te —
= > £ =
c c C > 0)
CO
c
® w § ®
o> o <2
c <»>
o 2?
CO ^ o ?
^ !t _ it O it
S o RS o S o
S ^ 3 :-
CQ
oncept of Operational Employment
QJ
2-
Z QJ N
O o
HH O
H h-l H N ^
£>
C/)
cu N CD ^.S CA
CD
>
□o
CA St
H
O)
QJ o I-H
N(
o (/)
X
CN
u
cs
pa P^ CA
O H (/) QJ
U
^
O
CA
■S J
W>
c
0 ri
Z Z Ui
CA
•»-i tK) d
C wH ffi C
s
^£
« C/!)
JZ>
CO
w u C
O
cn
•+-» CO'
Co Cu
:z;
HH Z HH
u
QJ^
O)
' HH
^
CA QJ
d
s §
C/!! C/D in 73
-. ^
CA
CA
H-i
^ LT)
o (/) H
ftf -c CD
• fH
(D 5:i rQ
CA CD
O HH M-H d
d ^ Si
O
;-i
HH
HH
H-l
^5. H.
CA
o
CJ t«
Pin Q PQ S § o
QJ
u
QJ
>o QJ
•4-a d
ri
CD CS
-itt
Ou U
St CA
•+-» «
cd"
o w O I-H
cs ’t-ss
H Ph pj O)
U
C
C2
5^
tq
•4-»
CA
tq
n
CA
CA ^ tq
o
ri
I
> QJ
H
'O
Ph
3-64
CO
QO
^ >>
• 1-H
I' Q
quipment
<f ^
^ (U
(N
m
>
n
f 1
cy w
Ph 00
iBHIBIBIIHHIliBBIB
'.: ■.. ^
iver Concept of Operation
ver Tactics
ssassstsiB
St
•S
c/T
^
s
-h S
o
o
Q^
n
^
^
.
Prosecute tgts, return to lane and continue
:-v ••
.
ntegrated Diver/Mammal
3-70
(D
4h
O
c/3
capability with DPD)
tegrated Diver/Mammal
$=! >
i a tb
• fH
CD C/5
’S
-tH
bD *S
QW >
CD
• fH 4->
(D o o
"S "S
CD cd
m 05 H-»
CCS
s
5
.sbX) Ph
O
ti
P
o
4b (D • fH
c/i
(D
> > (D s
CD
td c/5
CD
CD
> cd bb
O
^ Q A
CD
o § 5 P
Ph ^ 05 CD O
Cl ^ Lh
Ph
GO O^ i < > CD
CO
g§ sfin IW wX ffl l3 q O
(D
Vh
bX)
Apr/May
May 96
O C^ Q
VO
OV
VO
ov
u
VO
ov
;-H
so Ll]
cd
u
On
Si
o
O WX
w
.cd ^ %-^
(U <i->
cS
bD cd
IB DOc sM
<D ;-i
O
'S
<N <'
O <D
<D
(D CN
X
S
o^
a S a
o
<D
bX)
> <u
o H h-l
o <D
c/5
b£)
ao Q
(U
• rH
•
C/D
o
'S
(U
(D
(D c/5 (D (D c/5
» C/5 ^73
§ <1 O
Vh
* rH
O
X ^PQ wX
s—>
(/) (D (D
CO
CO
H W
X
(U
W <L)
dJ
VO VO
bD o^ o^ o^ o^
O bX) s-> >
b3 CDCIh o
0
v < in :z;
■ ■ ■ ■
3-74
o\
o^ o^
W) >
1
c3
1 o
xercise Schedule
mI
G^
'-O
o\
N
(N1
O 'sO
< o^
Ph X o g
|Jh w) sdJ
13
2 H Q ^
■ ■
l> I-' i> o^ o (N
o^ o^ a\ o^ o^ o
^ hP H H H HU
O
Ph
<
1
IL
O
eg
1 1
U
O
1
O
o
1
U
O
1
O
o
IW Test Detachment
y Events Ahead
s:
J-76
CHAPTER 4: LANDMINES
AND HUMANITARIAN DEMINING
This Chapter presents material on the problem of proliferation of landmine technology, the
elements of the debate on the proposed ban or restriction of the use of landmines, and some of the
technological approaches to meeting the difficult demands of Humanitarian Demining.
HUMANITARIAN DEMINING
Collected for this Chapter are the papers and letters that the authors, themselves, felt were
most appropriate to the Humanitarian Demining application. The reader is cautioned that technology
applications discussed under Humanitarian Demining may, in fact, also have utility in military
operational scenarios of countering mines on land or at sea.
It is useful to keep in mind the operational distinctions between Humanitarian Demining and
the military scenarios;
* Technologies (and systems) destined for the humanitarian demining applications should be
readily obtainable and capable of being operated by personnel indigenous to the areas in which the
demining operations will be conducted. In general, such personnel will not be used to dealing with
complex tools and instruments. Ease in instructing such personnel in both training and maintenance
is therefore as important a consideration as are the costs (financial and personnel) associated with
acquisition, staffing and maintenance. Please see below for additional comments regarding cost.
* Systems and approaches developed for the Humanitarian Demining applications will not,
in general, be used to conduct operations under fire or under the tight time constraints imposed by
military land and sea operations. Thus, in humanitarian demining, the user has the lu^ry of picking
the time of day and the weather conditions most conducive to success of the operation.
A comment on cost. It appears that this whole subject of cost associated with humanitarian
demining approaches needs further discussion. While it is possibly relevant that the choice between
or amongst two or more approaches that promise the same operational capability can be influenced
by cost, that is not the situation that normally occurs. Instead, the choice may be between kboriously
crawling over the entire contaminated area using sticks as probes, versus the use of a million-dollar
4-1
bulldozer that can rapidly, and relatively safely, strip and sift the soil to remove landmines.
In the cost equation, it is also necessary to include opportunity costs. What are the true costs
associated with the continuing denial of productive farmlands or other areas of commerce and
economic viability? There has been a tendency to use such figures as $1,000 per mine recovered
without considering the alternatives. This is an example of what economists call "harmful sub-
optimization.”
The importance of economic measures of effectiveness is so great that the 1998 Symposium
on Technology and the Mine Problem will include this subject in its call for papers.
The Hon. H. Allen Holmes, Assistant Secretary of Defense for Special Operations and Low
Intensity Conflict (SOLIC), led off the subject of Humanitarian Demining with his luncheon address
Tuesday Nov. 19, on the history and direction of U.S. government policy in this critical area.
Following Ambassador Holmes’ address, the first plenary session established and delimited the terms
of the debate on important policy aspects of the issue. It also set the tone for subsequent discussions
of the more techmcal demining developmental issues. Session co-chairs focused on the different
organizational and cultural challenges which render policy choices and decisions on demining so
difficult to make.
Among the three speakers in the plenary session on Humanitarian Demining and Demining
Policy there was consensus on the need to effectively meet the humanitarian demining challenge, but
views differed when it came to interpreting U.S. administration policy, timing of implementation, and
whether its long-range effect would become more restrictive on the U.S. than on its potential
enemies. Mr. Steve Goose of Human Rights Watch was particularly effective in arguing that a total
ban on the use of anti-personnel landmines, “smart” or dumb, was the defacto outcome of U.S. policy
initiatives to counter what he saw as “WMD in slow motion,” Arms Control and Disarmament
Agency representative Mr. Robert Sherman tended to emphasize incremental progress through the
Convention on Conventional Weapons (CCW), whose membership included both Russia and China,
each large-scale manufacturers of anti-personnel landmines (APLs) with stocks for resale throughout
the globe. And Lt. Gen. Robert Gard, USA (Ret) presented a moving testimony detailing why the
time to outlaw anti-personnel landmines is long past. The Open Letter to President Clinton signed
by him and fourteen other high-ranking former U.S. military officers attesting to the fact that such
a ban is not only humane, but militarily responsible, is reproduced in this Chapter.
Despite differences of opinion by speakers over the relative merits of pursuing mine self-
destruction, self-deactivation, or self-neutralization technological approaches, however, the
conference as a whole seemed impressed by the gravity of how some form of universal ban on
landmines might help alter the grim statistics of 100,000 civilian casualities occuring per year from
previously unrestricted use of anti-personnel landmines.
The related plenary Session on Technology and Humanitarian Demining featured four
speakers who each detailed specific aspects of the problem in the field and their personal experiences
in tiyong to deal with humanitarian demining challenges in various countries. Major Colin King’s slide
presentation was utterly compelling in its detail concerning the anti-personnel landmine threat and its
4-2
impact worldwide (his paper is in Chapter 3 of this Proceedings). Sam Samuels of Essex Corporation
and Lt. Col. Garth Barrett explained their approach to training indigenous demining teams, and the
challenge of determining how effectiveness can be measured and proven to peoples of different
cultures.
Professor Nicoud reinforced this aspect of the problem through his detailed presentation of
the ongoing, massive demining efforts in Cambodia. The differential impact of culture and confidence
was vividly portrayed in his example, whereby demining teams would play a game of soccer on the
very grounds on which they had recently completed their demining efforts. Richard Walden of
Operation USA/Operation Land Mine explained how the new roles of non-governmental
organizations (NGOs) and PVOs are presenting challenges for governmental agencies to develop both
planning and resource methods to deal with these worldwide problems. IBs bottom line. We need
to know more than we do to effectively work together, but we must, and it will be done.
The final Session in this series, parallel Session XXn on Humanitarian Demining, featured ten
speakers who largely provided technical details reinforcing the insights from previous sessions.
Scheduled near the end of the conference, it covered new project initiatives, recent Bosnian
experiences with the M-60 A-3 turretless tank, a discussion of a new combined sensor package to
help discriminate non-mines from mines and largely eliminate the metal detection clutter problem,
and a presentation on the use of LEXFOAM to rapidly and effectively deal with designated anti¬
personnel lanemine threats with a minimum of danger to demining teams.
4-3
The History and Future of
U.S. Policy on a Universal
Anti-Personnel Landmine Ban
Good afternoon. It is a pleasure to be with you this afternoon and to be part of the 1996
favorite things - talk about how the U.S. government has responded to the worldwide anti¬
The anti-personnel landmine crisis has developed principally because of the way the mines have
been used over the last 15 years or so - mostly in internal conflicts and predominantly in
paramilitaries and insurgents. In some cases, they have been employed specifically as weapons
of terror against the civilian population. The result is a humanitarian problem of epidemic
proportions. It’s estimated that as many as 100 million anti-personnel landmines are scattered
over 60 countries, killing or maiming an estimated 1,200 people per month... or one every
thirty minutes. Last year, 80,000 mines were removed, but many more were planted. At our
current clearing rate of 100,000 anti-personnel landmines per year, it will take over 1,000 years
to clear the landmines in the ground today. The countries most severely afflicted by these
“hidden killers” are Afghanistan, Angola, Croatia, Iraq, Somalia, Mozambique, Bosnia and
Cambodia - where one of every 236 people is an amputee - about the highest rate in the world.
4-5
An anti-personnel landmine - a simple $3.00 weapon - doesn't know when conflict ends. And
it cannot distinguish between the steps of a soldier and that of a child. Long after a conflict is
over and the warring troops have gone, anti-personnel landmines remain ... often for 30 years
or more.
The anti-personnel landmine crisis has taken an enormous toll on populations and governments
around the world. And their cost goes far beyond the initial tragic toll in human suffering. The
failure of a country to address the proliferation of anti-personnel landmines, beyond the obvious
personal suffering, denies farmers use of their fields which stymies the resumption of
agricultural production, denies access to markets, reduces public confidence in fledgling new
governments and creates many other hurdles for a nation trying to heal the wounds of war.
The exorbitant cost of mine-clearing operations siphons off already scarce fimds. Anti¬
personnel landmines make the reconstruction of rail and road networks, of power lines and of
waterways nearly impossible. In Mozambique, where civil war was waged for almost 20 years,
over 2 million landmines have been laid by the warring parties. The United Nations reports that
all 28 major road systems are blocked by uncleared mines. And because many anti-personnel
landmines were designed to maim and not to kill, mine injuries cause tremendous trauma,
require extensive medical treatment and follow-on care, and overburden existing
health care systems, raising health care costs beyond what developing countries can
handle.
4-6
Perhaps most tragically, anti-personnel landmines block access to vast stretches of otherwise
habitable and usable land. The loss of agricultural land takes away the only means that many of
these poor agrarian people have to earn a living. So, beyond the injuries inflicted and the
medical expenses incurred, anti-personnel landmine fields drive whole societies into helpless,
destitute poverty with no way out. In northern Iraq, for example, children of farmers now
harvest anti-personnel landmines from their fields instead of crops. They risk life and limb to
The return of refugees is fraught with danger and can be delayed because of anti-personnel
landmines. Often, refugees who fled their country during war are forced to remain in a foreign
country, dependent on international relief. But when mines don t prevent relief organizations
from delivering food and emergency supplies, they make already difficult relief operations more
hazardous. If overland transport is too dangerous, air transport must be used - a more costly
alternative. This widens the ever-increasing gap between the growing humanitarian needs and
It is against this backdrop that the extent of the landmine problem became widely understood
and the call for a total ban on landmines developed and intensified. Based on what I’ve told you
about the socio-economic costs of mines — not only for afflicted countries but for the entire
international community - this might seem the appropriate solution. But the issue is really quite
complex. And there are no simple answers because at the present anti-personnel landmines are
an integral defensive element in our military doctrine - how we fight war - and even though the
U.S. has been responsible in its use of mines, we recognize that we must look for other ways to
do our business.
4-7
4
Landmines are important defensive battlefield weapons used to counter enemy mobility, help
shape the battlefield, protect exposed flanks from counterattacks, and create defensive positions.
Most militaries use minefields tactically as an integral part of many phases of war fighting. In
Operation DESERT STORM, for example, the coalition forces used air-delivered mines to
protect the right flank of U.S. and British forces while they swung around Iraqi troops in
Kuwait. These mines held two Iraqi divisions essentially immobile, preventing their counter¬
Minefields are an inexpensive and vital force multiplier. Used in this manner, mines can
successfully defend a small force against a larger attacking force or until reinforcement arrives.
Barrier minefields are laid in demilitarized zones and between hostile nations or opposing forces
to deter and raise the cost of aggression, to delay enemy forces in the event of attack, and to
counter the possibility of surprise, such as in Korea. Landmines save American soldiers’ lives by
Proponents of an immediate total anti-personnel landmine ban assert that while the mines serve an
important military function, this pales against the human tragedy resulting from the use of mines.
The leaders of this constituency within the U.S. Congress are Senator Leahy and Congressman
Lane Evans. They are joined in their views by such prominent people as former Secretary of
State Cyrus Vance and by well-respected organizations like Human Rights Watch and Vietnam
Veterans of America Foundation. Internationally, they are supported by such leading figures as
the Secretary General of the United Nations and by the President of the International Committee
4-8
5
policymakers in the U.S. and in many other countries have turned to technology in their search
for meaningful solutions. The militaries of most industrialized countries increasingly use
sophisticated mines that self-destruct after a certain period of time. Nations developed such
devices to protect their troops, which are often required to maneuver over areas where they have
laid mines. The commander would know that the mines had self-destructed after, say four
hours, and that he could safely permit his troops to traverse the minefield they had earlier laid.
The self-destruct feature ensures that the anti-personnel mine will not only be disabled, but also
that it cannot be re-used. The fact that a so-called smart mine is powered by a battery means
that a natural back-up feature exists that will ensure self-deactivation of the mine, even if the
self-destruct mechanism fails to work. As batteries deplete naturally, such landmines are
guaranteed to become in-active within 90 days (at a 99.999 % reliability) ... vice 30 or more
landmines pose virtually no threat to civilian life once a battle is over. But under the
comprehensive international ban we seek, use of even these smart mines would also be ended.
personnel landmines. And if we ask them to give up their dumb anti-personnel landmines, we
have asked them to give up all they have, while we continue to use the more responsible self-
destructing/self-deactivating model. In agreeing to give up our smart mines, the United States
has taken the moral highground and has set the example for other countries regardless of their
4-9
6
The anti-personnel landmine crisis has commanded this administration’s attention from the very
start. President Clinton has aggressively pursued a comprehensive worldwide ban on the
production, stockpiling, transfer and use of all anti-personnel landmines. And we provide
training, equipment and funds to help those nations most threatened by mines. Thanks to the
vigilance and hard work of Senator Leahy and others like him. Congress passed a unilateral
moratorium on the transfer of anti-personnel landmines in 1992. That moratorium has now been
extended to the year 2000. In 1994, the United States spearheaded a successful resolution at the
United Nations to ban exports of the most dangerous kinds of landmines. Also in 1994, the
president dedicated the United States to eventually eliminating all anti-personnel landmines.
In May of this year, the president announced a series of actions the United States would take to
pursue that goal. He ordered an immediate ban on the use of so-called dumb antipersonnel
landmines - those which remain active until detonated or cleared. The only exception will be
for those mines required to defend our troops and our allies from aggression on the Korean
Peninsula and those needed for countermine and humanitarian demining training purposes.
The remaining stockpile of nearly 4 million anti-persormel landmines will be removed from our
The president’s anti-personnel landmine policy strikes an important balance between military
and humanitarian imperatives by carefully ensuring that essential U.S. military requirements and
commitments to our allies will be protected. Until an international ban takes effect, the United
because there may be battlefield situations in which they are necessary to protect American lives.
4-10
7
The Administration is determined to end our reliance on landmines completely and has directed
the Defense Department to find alternatives that will not pose new dangers to civilians.
In compliance with that tasking, we are examining a wide variety of concepts, technologies and
systems. But so far, we have found no analytical basis for recommending any particular
But any replacement is likely to involve a combination of three elements: surveillance - a sensor
mechanism like JSTARS, UAV-mounted sensors or ground sensors; the overfire - mortars,
artillery or aircraft, for example; and the man-in-the-loop — the command and control element
which ties the sensor to the precision lethal fire. Let me assure you that although we don’t yet
know what the solution will look like, technologies clearly exist and we are confident that we
The military has demonstrated a strong show of support for the President’s anti-personnel
landmine ban. Secretary Perry has said he looks forward to the day when anti-personnel
landmines will not be used in Korea. Gen Colin Powell has said that he abhors landmines.
Gen Shalikashvili, chairman of the Joint Chiefs of Staff, said the president’s policy set a
“prudent and responsible course that will lead to the elimination of all anti-personnel landmines,
while continuing to protect American lives.” He added that as practical solutions are pursued,
our priorities must be to maintain warfighting superiority while concurrently protecting the
safety of U.S. service men and women. In April, more than a dozen retired U.S. generals signed
an open letter to the president which called the anti-personnel landmine prohibition “not only
4-11
8
The President’s anti-personnel landmine policy also directs DoD to expand our humanitarian
demining program. Demining is one of the most fundamental humanitarian missions that the
United States can be involved in. The goal of our demining effort is to help countries establish
from landmines, eliminating the hazards posed by landmines and returning mined land to its
previous condition. The program assists the host country in development of all aspects of mine
awareness and mine clearance procedures, with the caveat that no U.S. personnel will clear
landmines or enter active minefields. Under the auspices of my office, DoD is pursuing a vital
role in humanitarian demining while improving the readiness of U.S. forces through the unique
Special operations forces are the primary U.S. military resource for the training programs.
Civil Affairs units play a key role in developing indigenous demining entities and helping them
awareness programs which educate populations in affected areas regarding the dangers of land
mines, what they look like, and what to do if a landmine is located. Special Forces units train
host country nationals to train others in their country to locate landmines, to mark fields and to
destroy the mines strewn indiscriminately on key roads, in villages and in fields.
4-12
One of the most heavily mined nations in the world is a also developing success story.
Honolulu. In a quarter-century of warfare, the Cambodian civil war has become infamous for
its unrestrained violence, with an estimated one million deaths resulting from the takeover by
the Khmer Rouge in the 1970s. Now out of power, the Khmer Rouge continue to fight. Both
sides in the conflict have resorted to wholesale mining of the countryside to deny territory to
their adversaries and to control and terrorize local people. As a result, Cambodia is now riddled
Cambodia. Special operations forces trainers have conducted mine awareness, mine clearance,
and medical and professional training for the Royal Cambodian Armed Forces and the
Cambodian Mine Action Center. Our effort has helped reduce the rate of mine-related injuries
This year, our new program in Laos follows the example set in Cambodia. Over 20 years have
passed since the end of the conflict in Laos, yet a significant amount of land is still infested with
mines. In concert with the Lao National Steering Committee and the United Nations
and training assistance are now being expanded. In Vientiane, mine awareness and clearance
programs for both anti-personnel landmine and unexploded ordnance awareness and
clearance programs and training schools. This is being followed by the establishment
4-13
10
To support full implementation of the Dayton Accords, we are currently leading an international
effort to begin clearing millions of landmines scattered throughout Bosnia and Herzegovina.
We provided $3.5 million to establish the Bosnian Mine Action Center and have pledged up to
$15 million to continue demining operations this year. The Bosnia Mine Action Center operates
under a UN mandate, coordinating all mine awareness, data gathering and mine clearance
activities through three regional offices — one in each ethnic region of the country. It will
eventually become an entity of the Bosnian government. A U.S. Special Forces team recently
completed training of 155 Bosnian deminers representing all three ethnic communities. This
In response to the President’s policy, our demining operation has expanded significantly.
The number of countries eligible for assistance has increased from 9 in FY 1996 to 14 in FY
1997, with 10 additional countries being considered. During the same period, the number of
U.S. personnel deployed for humanitarian demining operations has increased by 77 percent;
the number of indigenous forces trained has increased by 133 percent; and the dollar value of
equipment transferred has increased by 32 percent. Further expansion and initiatives are
ongoing.
The President’s anti-personnel policy also directs the Defense Department to undertake a
substantial program for the development of mine-detection and mine-clearing technology and to
share this improved technology with the broader international community. In Fiscal Year 1995,
DoD began a $10 million dollar research and development program to develop simple,
hopefully inexpensive, equipment that can assist countries in detecting and clearing landmines.
During that year, 30 new demining technologies and equipment were developed.
4-14
11
In FY 1996, 13 items were evaluated. Congress authorized $14.7 million for the program in
Fiscal Year 1997, in an effort to expand research and development. The humanitarian demining
R&D program uses expertise from government, industry, academia, foreign countries, the
United Nations and NGOs to produce practical solutions for locating and clearing minefields,
and for detecting, marking, recording, reporting and destroying individual landmines. The
ultimate goal is to place demining equipment in the hands of indigenous deminers, non¬
In Bosnia, where it will take nearly 30 years to clear the region’s 3 million landmines with
current technology, we are field testing numerous equipment and techniques in support of the
Division at the Army’s Night Vision and Electronic Sensors Directorate, that is identifying and
evaluating technologies for mine detection and clearance. It is a highly successful program, as
evidenced by the technology we have made available in Bosnia, as well as in other countries.
For example, the first two mine-sniffing dogs employed in Bosnia are from our program ... as
are the barrels of liquid explosive foam (LEXFOAM) and backpack dispensers that are being
shipped to the area as we speak. We also developed the mini-mine detectors and mini-flails
now in Bosnia. The mini-flails, by the way, have received high marks from a number of general
4-15
12
My office is working to develop and deploy additional equipment to Bosnia for use in
humanitarian demining. I know that Mr. Hap Hambric, DoD’s Project Leader for Humanitarian
Demining Technology Development at the Army’s Night Vision Laboratory at Ft. Belvoir, VA,
has already talked to you in more detail about this equipment. Let me just say that we’re also
working to optimize the use of commercial construction equipment to dig up mines more easily
The Administration’s anti-personnel landmine policy puts the U.S. squarely on a path
to eliminating landmines within our own military while leading international efforts to ban anti¬
personnel landmines. Largely as a result of our leadership, more than 30 countries have either
declared formal moratoria on anti-personnel landmine exports or have other export controls in
place. However, the most effective means available to the international community to control
law. In the case of landmines, this law is embodied principally in Protocol II on Landmines in
The CCW was negotiated by the international community in the aftermath of the Vietnam War
to limit the use of conventional weapons which can cause unnecessary suffering or
indiscriminate effects. But because it was universally recognized as weak on the issue of
landmines, nations convened in Vienna last fall to negotiate ways to strengthen it. At that
session and at subsequent review conferences, considerable progress was made on several points
that will, when entered into force, go a long way toward reducing the humanitarian crisis by
4-16
13
Protocol II requires that dumb mines be used only in marked and monitored areas and that the
state laying the minefield be held responsible for the field until it is removed. Also, there
was substantial agreement on technical specifications for reliability of self-destruct and self¬
deactivation for anti-personnel landmines used outside marked and monitored fields.
On detectability, all states agreed that anti-personnel landmines should have a relatively high
metallic content to assist in mine clearing operations, Unfortunately, some states may require up to
nine years after entry into force to comply with these important provisions. Protocol II also
places the responsibility for maintenance or clearance of minefields on the party that laid the
mines and requires that this responsibility be carried out at the end of active hostilities.
The revised Protocol does not include all the improvements proposed by the United States.
Nevertheless, it is a remarkable achievement that will, if widely observed, save many civilian
lives and return killing fields to planting fields. And, although the Protocol stops short of a ban
on anti-persoimel landmines, it is a critical step on the road to our ultimate objective - the
At the United Nations General Assembly earlier this month, the United States introduced a
resolution calling for an eventual ban on the production, stockpile, transfer and use of anti¬
personnel landmines. It is the fourth, and strongest, UN resolution that the United States has
spearheaded to eliminate all anti-personnel landmines. We’ve already begun to consult with our
4-17
The U.S. has taken several other steps to address the APL problem. My office
initiated a novel approach to basic mine awareness and mine avoidance lessons.
the Superman character that graphically teaches children about the dangers
of anti-personnel landmines. The comic book shows the superhero rescuing two Bosnian boys
who are about to walk into a field of landmines. Superman also uses his x-ray vision to protect
the boys from a booby-trapped house. Half a million comic books were printed — in both the
Latin script used by Bosnian Muslims and Croats and the Cyrillic used by Serbs — ^re being
distributed to children in the region by the NATO-led peacekeeping force and the Mine Action
Center in Sarajevo. We hope to design different versions for other countries where children are
killed or injured daily by mines and ammunition.
management. The center will provide single point access to a full spectrum of information,
training, research, analysis and services in support of our humanitarian demining program.
Last year, we created and released the first ever worldwide mine database. MineFacts is a
compact disk containing over 275 megabytes of information on 675 types of mines throughout
the world. It presents illustrations of each type of mine, accompanied by a text describing the
mine, the various names by which it is known around the world, and its country of manufacture.
4-18
15
My office also created a landmine database specifically for use by soldiers in Operation JOINT
ENDEAVOR. The three disk set, called BosniaFile, contains critical information on the 36
landmines most commonly found in Bosnia. Data includes pictures, general information on the
size and weight of mines, the metal content, country of origin and emplacement methods.
And we developed a Demining Web Site on the Internet to provide information on all aspects of
anti-personnel landmines and their removal. This already popular web site is assisting the
demining community in two important ways: by providing easy access to detailed technical and
background information, and by facilitating communication among the participants in the fight
against landmines.
CONCLUSION
Let me close by noting that in the time that I’ve been speaking to you, anti-personnel landmines
have claimed another life. It is a complex technical, political and military problem that
nevertheless requires immediate solutions. Our children deserve to walk the earth in safety.
The Clinton Administration is committed to real solutions as quickly as possible. The president
has put the United States firmly behind a responsible program to rid the world of these hidden
killers. And he recently repeated his call to the United Nations General Assembly to negotiate
top arms- control priorities. Finally, I want to thank all of you for your dedication and ask for
We must find a responsible answer to this dilemma. Remember, we are only limited by our
4-19
4-20
Opening Remarks for Session X on
Humanitarian Demining and Mine Policy
4-21
4-22
Submitted by LTGEN Robert C. Gard, USA (Ret)
An Open Letter to
President Clinton
Dear Mr. President,
We understand that you have announced a United States goal of the eventual elimination of
antipersonnel landmines. We take this to mean that you support a permanent and total international
ban on the production, stockpiling, sale and use of this weapon.
We view such a ban as not only humane, but also militarily responsible.
The rationale for opposing antipersonnel landmines is that they are in a category' similar to poison gas;
they are hard to control and often have unintended harmful consequences (sometimes even for those who
employ them). In addition, they are insidious in that their indiscriminate effects persist long after hostilities
have ceased, continuing to cause casualties among innocent people, especially farmers and children.
We understand that: there are 100 million landmines deployed in the world. Their presence makes
normal life impossible in scores of nations. It will take decades of slow, dangerous and painstaking
work to remove these mines. The cost in dollars and human lives will be immense. Seventy people
wUl be killed or maimed today, 500 this week, more than 2,000 this month, and more than 26,000
this year, because of landmines.
Given the wide range of weaponry available to military forces today, antipersonnel landmines are not
essential. Thus, banning them would not undermine the military effectiveness or safety of our forces,
nor those of other nations.
The proposed ban on antipersonnel landmines does not affect antitank mines, nor does it ban such
normally command-detonated weapons as Claymore “mines,” leaving unimpaired the use of those
undeniably militarily useful weapons.
Nor is the ban on antipersonnel landmines a slippery slope that would open the way to efforts to
ban additional categories of weapons, since these mines are unique in their indiscriminate, harmful
residual potential.
We agree with and endorse these views, and conclude that you as Commander-in-Chief could responsibl}'
take the lead in efforts to achieve a total and permanent international ban on the production, stockpiling,
sale and use of antipersonnel landmines. We strongly urge that you do so.
4-23
Seeking Real Solutions
to the Landmine Problem
The anti-personnel landmine (APL) problem isn’t a matter of philosophy or ideology. It’s real civilians maimed or
killed by real mines every day, and denied use of their land because of the threat of uncle^ed live r^es. In landmine
negotiations, the only measure of our success is the real landmine casualties and land denial we ultimately prevent.
The vast majority of APL casualties are caused by mines produced, exported, and/or used by Russia and China. These
countries are also the most determined holdouts against mine limitations. I say this not to criticize these governments
but to define the problem we face. At the end of the day, the issue will not be the purity of positions taken by the many
nations who are not the problem. The issue will be the future humanitarian practices accepted and observed by the few
nations who have been the problem.
The root of the APL problem is the fact that most mines, by design, function for decades after emplacement. AnoAer
very serious problem is that many mines are non-metallic, and therefore not easily detected by humanitarian de-miners.
Yet neither long duration nor nondectability is usually necessary for the military function of the mine. The focus of
CCW, then, became to reduce the danger to civilians fi'om APL, while at the same time allowing their effective military
use. Within this context, CCW achieved some remaikable successes, neither trivial nor easily accomplished. These
include
SHORT MINE LffE. Unmarked APL must self-destruct with 90% reliability within 30 days of emplacement, and they
must self-deactivate (exhaustion of a battery without which the mine cannot operate) within 120 days of emplacement
with 99.9% overall reliability. A country may claim a 9-year transition period before implementing about half of these
restrictions; the other half are effective upon entry into force. Once this restriction is implemented, lethal duration of
unmarked i^L will be days, where now it is years; casualties will be reduced by several hundred times.
DETECTABILITY. All APL must have 8 or more grams of iron equivalent, to facilitate humanitarian demining. A
country may claim a 9-year transition period before halting use of nondetectable mines, but transfer of such mines is
prohibited immediately.
PROTECTING DEMINERS. Anti-detector mines, which are designed to ekplode when a magnetic mine detector
passes over them, are banned completely and immediately.
4-25
We would have liked to do still better. The United States sought to have no transition periods, 95% self-destruct
reliability, mandatory detectability for anti-tank as well as anti-personnel mines, and mandatory verification. While we
were unable to persuade the holdout states to go that far, these governments moved far beyond their positions of only
seven months before the end of the conference. At that point, any agreement accommodating the holdout positions
would have included no requirements for self-destruction or self-deactivation, and an unlimited transition period for
detectability.
The bottom line, of course, is not who “gave up” what. The purpose of the conference was to protect civilians from
mines. By that standard CCW will, if observed, succeed on a large scale. If it had been put into force and observed
thirty years ago, there would be no humanitarian landmine crisis today.
On May 16, 1996, President Clinton announced that “The United States will seek a worldwide agreement as soon as
possible to end the use of all antipersonnel landmines. The United States will lead a global effort to eliminate these
terrible weapons... and stop the enormous loss of human life.” He directed that US forces immediately and permanently
halt all use of long-duration mines except in Korea. He also announced that “Because of the continued threat of
aggression on the Korean Peninsula, I have decided that, in any negotiations on a ban, the United States will protect our
right to use mines there.”
Clearly, even if CCW were universally observed and prevented more than 99% of APL civilian casualties, 100% is
better. The only way to eliminate all APL casualties is to eliminate all APL. But the task before us is formidable.
In the near term, the U.S. must find an affordable and effective substitute for APL for the Korean situation. The
seriousness of the threat to Seoul, and the challenge of finding an effective substitute for APL in countering that threat,
should not be understated.
More enduring difficulties lie in the holdout states’ attachment to the military use of APL. The Chinese tell us they will
give up nuclear weapons before they give up APL; the Russians tell us the only popular concern they hear about APL is
fi*om mothers anxious that their sons in the Army have the means to defend themselves. We must have both these
countries on board if the final agreement is to be more than cosmetic.
4-26
Banning Anti-personnel Landmines
Stephen D. Goose,
Program Director, Human Rights Watch Arms Project
I would like to thank A1 Bottoms and the symposium organizers for inviting me to
participate, and also for including this particular panel on mine policy in the program. I
understand that the first NFS symposium did not have a policy dimension. I believe it is crucial
that the practitioners and planners of mine warfare and countermine warfare - the best and
brightest of which are assembled here today ~ understand the political environment in which they
are operating. I have concluded from listening to the speeches during the first day and one-half,
and from many private conversations, that most of those attending this symposium do not have
that understanding..
There seems to be a widespread lack of knowledge about U.S. policy with respect to
antipersonnel (AP) mines and about the rapid progress that has been made toward a
comprehensive international ban on AP mines. Though I had not planned on it, I feel compelled
to begin today with a clarification of the new mine policy announced by the President and
Secretary of State and Secretary of Defense on May 16th. It is now the ofiBcially stated U.S.
policy that all AP mines must be banned, and must be banned as soon as possible. It is not U.S.
policy that just “dumb” mines be banned. Yet, General Sheehan yesterday characterized it that
way. The brochure for this symposium characterizes it that way. General Gill yesterday went
considerably farther and complained that in some quarters dumb and smart mines were being
treated the same way, saying he was engaged in a battle of semantics and that a distinction must
be made between self-destruct and non-self-destruct mines. General Gill’s remarks strike me as
nothing short of an attack on official U.S. policy. As an NGO, I regularly attack official U.S.
policy, and will do so at some length in a few minutes, but I was quite surprised to hear an active
duty officer do so publicly.
“The United States will seek a worldwide agreement as soon as possible to end the use of
all anti-personnel landmines. The United States will lead a global effort to eliminate these terrible
weapons and to stop the enormous loss of human life.... Until an international ban takes effect, the
United States will reserve the right to use so-called ‘smart mines’ or self-destructing mines as
necessary.... But under the comprehensive international ban we seek, use of even these smart anti¬
personnel mines would also be ended.”
4-27
I urge conference participants to heed these words. The President repeated them in a
speech before the United Nations in September, and they are part of the U.S.-sponsored
landmines resolution now before the General Assembly. The U.S. government has recognized
that smart mines are ultimately part of the problem and that a total ban is necessary to solve the
mines crisis. You should be concentrating on ways to get rid of smart mines, on ways to fulfill
your military missions without smart mines, rather than ignoring or trying to reverse the policy.
Let me backtrack for a moment to discuss why the U.S. and so many other nations are
now calling for a ban. Simply put, we have a humanitarian disaster on our hands. This is a
symposium on “the Mine Problem,” but I have heard almost no discourse about the mine problem
as I know it, or as millions of people in Cambodia, Afghanistan, Angola, Bosnia and many other
nations know it .. I say this not to belittle the importance of countermine warfare, but to
emphasize the importance of the global context Of the landmines crisis. The State Department
estimates 26,000 people are killed or maimed by landmines each year, about one every twenty
minutes, one during the course of my talk. The victims are almost always civilians, often women
and children. Landmines cause more casualties after conflict has ended than during the fighting.
More than 100 million mines are planted in more than 60 nations; there are another 100 million in
stockpiles; the U.N. estimates some 2 million additional mines are laid each year. The horrific
human toll is compounded by the socio-economic impact, as mines prevent access to fields, roads,
rails, bridges. Nations like Mozambique of Bosnia achieve peace but cannot begin the process of
reconstruction and development until mines are cleared. Refugees and internally displaced persons
cannot return home. For these reasons, antipersonnel mines have been called, first by Human
Rights Watch and later by Secretary of State Christopher, “weapons of mass destruction in slow
motion.”
The International Campaign to Ban Landmines (ICBL) believes that any use of AP mines
is a violation of international humanitarian law. The weapon is inherently indiscriminate, and its
use clearly fails to meet the proportionality test of humanitarian law: the short-term military
benefits are far outweighed by the long-term human and socio-economic costs.
What is being done about the crisis? NGOs created the International Campaign to Ban
Landmines five years ago. It has grown into the most diverse and successful coalition ever. The
ICBL consists of more than 700 NGOs in more than 40 nations. It includes organizations involved
in demining, victim assistance, rehabilitation, human rights, arms control, humanitarian relief,
medical, veterans, religious issues and more, a senior UNICEF official recently said that the
ICBL is “the single most effective exercise of civil society since World War II.” The ICBL has
two calls: for a comprehensive ban on the use, production, Stockpiling and export of antipersonnel
mines, and for increased resources for humanitarian mine clearance and victim assistance
programs.
4-28
Though many participants here may be unaware, the successes of the ban movement are
stunning, especially over the course of the past year or year and one-half The movement has
quickly grown beyond just NGOs, and has been endorsed by the ICRC, UNICEF, UNHCR,
UNDHA, U.N. Secretary-General Boutros-Ghali, and the most influential media sources, such as
the New York Times and the Economist.
Under pressure, governments began coming on board. Belgium became the first nation
formally to legislate a total ban in March 1995, Norway followed suit in June 1995. Also under
pressure, nations agreed to review and revise the Landmines Protocol of the 1980 Convention on
Conventional Weapons, a process that took two and one-half years. Negotiations were supposed
to end in October 1995, but deadlocked with governments wanting to protect their own mine
stocks and methods of using them. At that time, the ICBL could count only 14 governments that
had publicly stated support for a complete ban. The negotiations finally concluded on May 3,
1996 and the results were sharply criticized by the ICBL as unlikely to make an significant
difference in the humanitarian crisis.
Still, it was clear on May 3 that an ever-growing number of governments recognized the
insufficiency of an approach based on complicated restrictions and technical fixes, and that a
comprehensive ban was the only answer. Many governments are now out in front of the U.S. on
this issue.
* We now have more than 50 pro-ban governments. The U.N. General Assembly
resolution calling on nations to “pursue vigorously” an international ban and to conclude a ban
agreement “as soon as possible” was passed by the First Committee on October 31 by a vote of
141-0, with 10 abstentions. Clearly, a new international norm is emerging.
* More than 20 nations have prohibited the production of AP mines, and begun destroying
stockpiles, including Germany, France and Italy.
* The Organization of African Unity has endorsed a total ban. The Organization of
American States adopted a resolution in June 1996 calling for the establishment of a hemispheric
mine free zone. The six Central American states declared themselves the first mine free zone in
September 1996.
4-29
Perhaps the key event was the Canadian government sponsored conference held in Ottawa
October 3-5, 1996. This brought together 50 pro-ban governments which agreed to a Final
Declaration calling for a comprehensive ban and, more importantly, a Chairman’s Agenda for
Action, which laid out concrete steps for achieving a ban rapidly. And in a dramatic
announcement at the end of the conference, Canada’s Foreign Minister Lloyd Axworthy stated
that Canada would host a ban treaty signing conference in December 1997. The international
community now has a deadline for agreeing to an international ban, and a roadmap for getting
there. The conference also featured perhaps unprecedented cooperation between governments
and NGOs, which has continued in the wake of the Ottawa conference. There has been great
enthusiasm for what might be called the Ottawa process, with various nations offering to host
preparatory meetings leading up to a treaty signing in December 1997. It is too early to tell how
many governments will come to Ottawa to sign a ban treaty; I believe it will be more than 50,
with significant numbers fi-om the developing world, where mines have been used most
extensively. It is doubtfiil if Russia or China will attend, but I believe they will be most effectively
brought in as international pressure builds and they can be stigmatized for operating outside the
international norm. The real “problem” states are not Russia and China, but the user states and
the affected countries, and it appears many of them will participate in the Ottawa process.
The U.S. has been decidedly cool to the Ottawa process. The U.S.-sponsored UNGA
resolution calls for conclusion of a ban treaty “as soon as possible,” but December 1997 appears
to be too soon. Instead, the U.S. is considering half-measures and a step-by-step approach that
will slow down momentum toward a ban and possibly undermine the Ottawa process. The U.S.
has refused to take steps at home that would turn its words into actions: changing its temporary
export moratorium into a permanent ban; adopting a production moratorium or ban; removing the
exceptions for use of dumb mines in Korea and smart mines anywhere.
I believe the U.S. is taking a go slow approach because of concerns expressed by the Joint
Chiefs of Staff and the regional Commander-in-Chiefs that alternatives to AP mines have not yet
been developed. Yet, fourteen of our most distinguished retired generals told President Clinton in
a full-page open letter in the New York Times, “Given the wide range of weaponry available to
military forces today, antipersonnel landmines are not essential. Thus, banning them would not
undermine the military effectiveness or safety of our forces, nor those of other nations.” They
also said, “We view such a ban as not only humane, but also militarily responsible.” Those who
signed include General Jones, former Chairman of the JCS, General Schwarzkopf, Commander
Operation Desert Storm, General Galvin, former Supreme Allied Commander Europe, and
General Hollingsworth, former I Corps in Korea.
We should not forget the dangers posed by mines to U.S. soldiers and peacekeepers.
General Sheehan yesterday told us that mines are a force equalizer that negate the U.S.
technological advantages and can inflict unacceptable casualties. He said they are the “war we are
not prepared to fight.” In Bosnia, landmines have claimed 224 UNPROFOR victims and 64 IFOR
victims.
Garth Barrett
Mechem International
We, in the commercial de-mining companies, see de-mining, when the conflict is over and
some sort of peace agreement is in place, as a three phase concept, as follows :
PHASE I : Urgent immediate de-mining by the military forces deployed to assist the
peace process. This will mean route, observation points and base
clearance. This process only eliminates about 5-10% of a country s
land mine problem.
PHASE in : General de-mining takes place with local personnel (military and civilian)
assisted by commercial companies.
Foreign military personnel, such as US Special Forces may train local
government forces in de-mining.
This process must deal with the bulk of the mine problem, with up to 90%
still to be eliminated.
The conflicts, which have created this land mine pollution, are normally found in poor
countries, which, if not impoverished prior to the conflict, have become so through the
fighting. Therefor, their ability to pay for and absorb technology can be very limited. The
military forces monitoring the peace do not transfer technology in Phase I or II and
commercial de-miners will not normally do so in Phase III, unless specifically contracted
to perform this service. However, the US, through the Special Forces and other countries,
do assist with training and the transfer of acceptable technology in this phase. Phase HI.
4-31
Some AID organizations require that only local civilian and ex military personnel from the
mine polluted countiy should be employed in de-mining, in order that they benefit
financially from the de-mining. This is very laudable, in that it assists in the financial well
being of the local community, but it is unfortunately extremely inefficient in getting the job
done. It has been stated that approximately 80,000 mines are being cleared each year in
de-mining operations and up to 1,000,000 mines are being laid, therefor we are seriously
losing ground under the present system. With the estimated 80-100Million mines that are
already in the ground, there is a major need to address this issue.
The need for the military to maintain speed and momentum in military mine clearance has
been stated many times, but this needs to be also a principle in humanitarian de-mining.
As an example in Somalia, the commercial de-miners speed across the ground was at best
1km per day, whereas in Mozambique we averaged just under 20kms per day, which has
also been maintained in Angola. This speed over the ground was made possible by using
advanced technology in the form of mine and ballistic protected vehicles and our explosive
vapor detection system. This vapor detection system eliminates the need to check areas
with no mines, which besets most de-miners. As another example, using four personnel
and three vehicles it was possible to clear 22,000 anti personnel mines in Mozambique in
three months of work.
The vision of de-miners conducting de-mining with almost the same equipment used in
World War II, has to change, if the task of clearing the world land mine problem is to be
accomplished, within the next two decades. Otherwise, we will still be de-mining into the
22nd century. Laboratories that produce equipment for the de-miner need to have de-
mining experienced personnel assisting them with the practical issues of humanitarian de-
mining, which not only deals with land mines, but UXO, as well as improvised explosive
devices. The mined areas of the world are "dirty", with all forms of explosive
complications which can pose all sorts of problems for high tech equipment. These areas
are, in most cases, battlefields We need to get equipment out of the "sand box" and into
the hands of operational de-miners hands.
We, in the US, need to find a means of combining commercial company personnel with
our Special Forces operating in foreign countries. Our Special Forces, when they are
training local government de-mining forces, are prohibited from entering a mine field,
which limits their effectiveness. It has also been an unfortunate problem that, after our
special forces personnel leave the country with the mine problem, many of the de-mining
efforts slowly wither and die. However, with a commercial partner with stay behind,
specialy tailored training materials in the program, their personnel could carry out this final
essential part of the program.
4-32
During this conference, we have heard about the vast sums of money being spent in the
political arena in connection with de-mining, but unfortunately very little of these funds are
reaching the rural areas that have the mines in the ground, where the problems actually
exist. We, as commercial de-miners, accept that, without the political awareness and
motivation, very little will be achieved. However, it is also fmstrating to see the
manipulation of the mine problem by many governments, whose countries are infested
with land mines. The powers that be need to find a formula, that can be translated into
meaningful action on the ground, where the real problems exist.
Whilst the ban on anti personnel mines and the destruction of these mines in certain stock
piles are greatly welcomed, we need to remember that there are still 80-100 Million mines
in the ground, which need our urgent attention, if we are to reduce the world financial
drain and economically uplift the populations suffering from this form of pollution !
4-33
4-34
Bringing New Technology to Bear
on Landmine Detection:
The Role of NGOs as Catalysts and Liaisons
Between Technology Providers
and the Mine-Affected Countries
Richard M. Walden
President, Operation USA
Operation USA and its new subs«dtar> "OperaUon Landmine: A Project To Rid
The World of Anti-Personnel Landmines" have been working on various
aspects of the landmine problem since 1979-tnitially through the wovision ol
prosdietics programs in Cambodia, El Salvador and Nicaragua and more
recently (1994-the present) as representative of 160 U.S. NGOs at a series of
landmine conferences in Geneva, Copenhagen, Ottawa, Washington,
Cambridge (MA.) and Tokyo,
The above activity fills a huge void. There are at present hundreds of signatoiy'
NGOs to the International Campaign To Ban Landmines and dozens (^' NGOs
active in prosthetics and mine awareness programs in mine-affected countries.
Until now, there were virtually no NGOs working on marrying advanced
technology to the detection, mapping and destructi(Mi problems. The crawl and
prod method, teams of dogs and largely ineffective metal detectors are the most
common methods in use. Statistical outcomes from the global $60-80 million
annual humanitanan demining budgets are paltry in terms of land cleared and
mines deactivated; mine laying still outetrips mine detection by 25 to 1.
4-35
NASA r«;ognized this and-at our urging and in conformity with its technology
transfer responsiHlity-set up the "Robotics Roundtable on Demining" at its
Western Center on technology Transfer m Los Angeles.[see attached
background information]. The Roundtable has attracted participation of the
Lawrence Livermore National Lab, the Jet FVopul$i<wi Lab, Lockheed-Martin,
Hughes, Lear Astronics, USC and the Defense Department's Humanitarian
Demining Unit.
That the science exists to find and destroy implanted AP mines is becoming
clear. That its re-engineering in affordable, maintainaWe, and mass producible
quantities will happen anytime soon is the challenge at hand.
NGOs have an emerging role in dus area, which had heretofore been considered
a military domain. The major NGOs have long-term relationships both with
mine-affected countries and with corporate donors to their programs. Major
technology comjpanies have yet to set up or sell any of their products to the
pwrest of the mine-affected countnes. They also do not directly pnovide de-
mining services of any kind, in order for them to become involved, NGOs
ha\'e to make the case that de-mming makes good business sense and is, in fact,
a multi-billion dollar opportunity to the company or companies that can convert
defense, space or other related technologies and bnng them to bear on the
landmine problem. Their path to ihe mine-affected countnes w ill be through
NGOs already field operational.
4-36
Proqrams In Development
Land Mine Clearance: New Technolopy
The roads, paths and fields of 64 countries are littered with .t 10 million land mmes. which kill or injure
more than 10,000 civilians a year Two to five million riew inmes are planted every year — twenty to
fifty times more than are cleared. Policy makers are discussing banpmg land lAit the prospect of
this occurring soon is minimal.
Operation USA is promoting the development of efficient high tech aliemattves to decades-old
technology apd substantial human labor presently used for demining Operation USA is working with
NASA, whose "off the shelf space technology Involving robotics, satellites and sophisticated sensors, is
well suited to mine clearance.
Our goaf is to hasten the re-engineering and deployment of roboircs-based demining technology,
and to.arrange it$ first field test in 1996 in the mine fields of Cambodia, which currently claim the lives
and hmbs of 4,000 people yearly
HealthCorps Vietnam
Operation USA’s 16-ycar cofnmifment to the people of Vietnam has led to our most ambitious health
development program. HealthCorps Vietnam will combine the knowledge, experience and dedication
of doctors and public health personnel in both the United Sutes and Vietnam to improve the health
Status of impoverished Vietnamese. It will balance public health programs, medical training, technology
transfer and health care systems development to foster greater accessibility and self-sufficiency m
Vietnam's provision of health care
By training local health professionals and introducing suitable medical technology, HealthCorps
Vietnam will help Vietnam match its rapid economic development with advances m caring for the basic
health needs of its people.
4-37
Organizations Calling for a Ban
Partial listing of over 450 NGOs in over 30 countries
Afghanistan
Afghan Coordinating Agency for Afghan Relief
Afghan NGO Coordination Bureau
Italy
AIFO-Associazione Itatiana Amici di
Raoul Follereau
What You Can Do
Afghan Technical Consultants CIES-Centre for Development
OAFA Information and Education
MDC EMERGENCY* I Endorse the Call for a Ban.
Mine Clearance Planning Agency FOCSIV-Federazione Organismi Cristiani di
Organization for Mine Clearance & Afghan
Rehabilitation
Servizio Internazionale Volontario
IRES Toscana
► Get your organization to join the
Radda Barnen, Peshawar Mani Tese
SWABAC Pax Christi Italy
Campaign.
Servizio Civile Internazionale
Austria
Greenpeace Austria Israel I Educate the public and media.
NGO Committee on Peace Association of israeli-Palestinian Physicians
Pax Christ! Austria for Human Rights I Urge your government to stop produc¬
World Peace and Relief Team
Kenya
Australia Maendeleo Ya Wanawake Organization tion, stockpiling, trade, and use of
Human Rights Council of Australia People for Peace in Africa
Medical Association for the Prevention of War
landmines.
Malaysia
Mercy Refugee Service of Australia
Asia-Pacific People's Environment Network
People for Nuclear Disarmament QLD
Just World Trust > Urge your government to support the
Belgium Mozambique
European Network Against the Arms Trade
Mozambican Association of the
United Nations voluntary trust fund and
Handicap International
Medecins sans Frontieres International*
Handicapped (ADEMO) other programs for mine clearance and
Oxfam Belgium Nepal
Pax Christi Flanders Women Development Society mine victim assistance.
Cambodia The Netherlands
Church World Service AMOK I Stigmatize the producers, exporters,
Coalition for Peace and International Fellowship of
Reconciliation Reconciliation and users of landmines.
Khmer Women’s Voice Johannes Wier Foundation for Health
and Human Rights
NGO Forum on Cambodia
Pax Christi
I Contact organizations on the back of
Canada
Canadian Council for Refugees New Zealand this brochure for more information and
Lawyers for Social Responsibility CALM-NZ Campaign Against
Physicians for Global Survival Landmines for a complete list of participating
Denmark Norway organizations.
DanChurchAid Norwegian People’s Aid
Handicap International Philippines
France Coalition for Peace
ACAT (Action Catholique pour I’Abolition de Gaston Z. Ortigas Peace Institute
la Torture) Jesuit Conference of East Asia
4-39
Meeting the Landmine Challenge - Research and Development in Support of Humanitarian Demining
world’s arsenal of military weapons. The end of the cold war missions show that mines are the primary cause of personnel
has not eliminated the potential for full scale warfare between and vehicle casualties. The political impact of landmine
military forces that would include the use of mines. They figure casualties among US military forces can jeopardize the
prominently in the war fighting doctrine of potential US successful completion of peace operations. Casualties suffered
adversaries such as North Korea and Iraq. This fact prevents by United Nations forces in Somalia demonstrated not only the
the United States from completely rejecting the use of effectiveness of mines as a weapon, but also how they can create
landmines by her military forces. It also illustrates the politically unacceptable losses. The few injuries and deaths of
unfortunate fact that landmines may remain a serious worldwide Interim Forces (IFOR) personnel in Bosnia have already
problem for the foreseeable future. sensitized populations and national leaders to the effects of
landmines.
The proliferation of landmines in the underdeveloped world
is the most significant cause of the high number of civilian The United States and other countries are working to
casualties. They are a prominent weapon in these regions eliminate the landmine problem. The plight of the many lesser
because they are so effective, yet so inexpensive and easy to developed nations suffering from severe landmine problems and
make. Landmines are frightening residual weapons of war that the threat to US forces engaged in peace operations has led to an
retard resettlement and economic renewal. This menace denies emphasis by the White House, Congress and the Department of
access to roadways and other lines of communication, villages Defense on the development of new technologies and equipment
and urban areas, agricultural fields and other rural areas long for mine detection and clearance. The development of these
after the declaration of peace. Their numbers and the new technologies will improve the efficiency, safety and
devastation they extract are staggering. When released in early effectiveness of the demining process. Research and
1995, the Department of State publication Hidden Killers, the development programs to meet the countermine needs of tactical
Global Landmine Crisis estimated that some 85-110 million military forces and of peacetime humanitarian demining
mines in 62 countries maim and kill approximately 26,000 operations are now underway at the Night Vision and Electronic
people a year. The problem is most acute in underdeveloped Sensors Directorate. Before describing the NVESD demining
nations already ravaged by conflict and that lack the resources technology development program in detail, an understanding of
and the infrastructure needed to deal with their landmine the process used by the United States military to assist other
problems. The removal and destruction of all forms of nations with demining is of value. Knowledge of this process
dangerous battlefield debris, particularly land mines and other will provide a better understanding of the types of technologies
UXO, are vital pre-requisites for a country to recover from the that the demining community needs.
aftermath of a war.
C. US Participation in Humanitarian Demining
B. The Landmine Problem is Tough to Solve
Military forces, non-governmental organizations and
The development of new demining technologies is a difficult contracted commercial enterprises have all been involved in
task because of the tremendous diversity of environmental demining. US military forces participate in the demining effort
conditions in which mines are employed and because of the within limits established by US government policy. American
wide variety of landmines. Mines range in size and type from forces may only perform demining for self-protection. American
anti-personnel models small enough to fit into the palm of a military forces involved in humanitarian demining will not enter
child’s hand to large anti-tank mines. There are different an active minefield. Rather than perform actual demining, the
activation mechanisms such as pressure, electronic and US theater commands establish and support demining and mine
command detonation. Mines use the blast effect from the awareness programs, and conduct demining training for
explosion or flying fragments to injure or kill their victims. indigenous personnel. Another important policy requirement is
Manufacturers make mines from metallic and non-metallic that deminers must destroy all mines where they find them. This
materials. Fusing, lethality, and emplacement methodologies is to prevent anyone from removing, stockpiling and re-using
have evolved significantly since WW II. Full width attack, them in the future. These policy requirements have an
stand-off “side attack” and “top attack” mines are either in important bearing on how DoD supports the demining effort,
development or already in the inventories of several armies. and therefore on the technology that it requires.
This tremendous diversity makes the demining mission very
complex and dangerous. Improvements in demining technology Planning, conducting and sustaining humanitarian demining
are critical to the success of any effort to reduce this threat to operations requires coordinated participation from US military
soldiers conducting peacetime contingency operations as well as and other national intelligence gathering and mine warfare
to the civilian population. analysis assets. Several military organizations with expertise in
mine warfare exist at commands responsible for training and
Reports from US forces deployed in support of peace doctrine and for material development. The expertise of the
enforcement, peace keeping and post conflict humanitarian
4-40
Meeting the Landmine Challenge - Research and Development in Support of Humanitarian Demining
Special Operations Forces (SOF) makes them the leading US forces continue to receive updates of mine employment
authority in this arena. prior to and during deployment. In the near future, detection
platforms carrying multi-spectral, integrated sensor systems
In any given regional area, it is important that all participants capable of detecting mined areas and areas containing no mines
be integrated into a command and control structure to support will provide these updates. Initially, the identification of mine
the theater command’s demining mission. The Special free areas is more important than plotting known mine fields.
Operations Commands in each theater provide this demining Identification of where mines are not will allow the population
command and control function for the CINC. to begin food production and other important tasks to rebuild
the economy in these areas while they mobilize the demining
Prior to US involvement in a demining program, intelligence effort.
gathering from national systems and on-site human intelligence
provide initial reports of mine warfare activities in the affected Mine awareness training is one of the most important parts of
nation. During this phase intelligence assets look for positive a humanitarian demining program, especially for returning
indications of the presence of mines such as stockpiles, mine refugees. Execution of this part of the demining plan is a
laying equipment, actual mine operations and casualty reports. combined effort between special operations, civil affairs units
More important, they search for areas not infested by mines to and the US embassy in that country. This mine awareness
identify safe areas. education requires effective training support equipment that
would include a comprehensive mine database, and multi-
The above process determines the scope of the landmine medial assets to disseminate mine awareness programs of
problem in the affected nation. With this as a beginning, instruction, warnings and photos.
intelligence and other fact finding assets survey and analyze
how and where the warring factions employed the mines. In addition to mine awareness training, the SOF component,
Electronic and in-place human intelligence assets will piece with civil affairs and psychological operations participation,
together the scope of mine use and the type of devices known or establish a training program for host nation deminers. The
projected to be in the warring parties’ inventory. training and mine awareness program will vary from country to
Simultaneously, assessment teams led by the responsible theater country based on the level of education and industrial capability.
special operations command will validate the threat. These These characteristics of the nation involved are important
teams may include representatives from the Joint Staff, the factors to consider when introducing demining equipment.
Department of State and various intelligence elements. This Establishment of a training program, and the types of equipment
knowledge is important to theater command planners as they needed are also highly dependent on diversity of the mine threat
tailor the personnel, training and equipment requirements to the in the host country, and the geographic and environmental
specific landmine threat in that nation. make-up of the land. When considering the need for
humanitarian demining around the globe, from desert to
The product from the above analysis is a list of known and temporal to jungle climates, there is a huge challenge to
suspected mines and booby traps, and their projected locations. optimize technology to make a meaningful difference in the
Analysts then enter this data into a comprehensive elimination of landmines.
mine/countermine database. The National Ground Intelligence
Center maintains this database. It provides technical and II. Progress to date
tactical information on all known mines in the area. This is
important information because the process to find and destroy A. Program Description
a mine depends on its physical properties. The team
simultaneously develops a database of known or suspected Traditionally, countermine/mine requirements have addressed
locations of mined areas. This tool is critical to planning and battlefield operations to support the pace of maneuver.
prioritization of humanitarian demining operations. Technology solutions for rapid surveillance, reconnaissance,
detection, and neutralization portend significant countermine
When a theater command completes its demining plan for a capability for maneuver forces needs to achieve requisite tempo,
given country, the Commander-in-Chief forwards the plan to the survivability, and battlespace management of
Joint Chiefs of Staff for approval. Once approved by the Joint countermine/mines operations. Humanitarian demining focuses
Staff and by the IWG, the theater command performs the on developing, testing, and evaluating the best available
mission. The Special Operations Forces (SOF) components technologies that might be applied throughout the full range of
assigned to the command execute the humanitarian demining demining requirements: locate minefields (or confirm their
program for the theater Commander-In-Chief (CINC). Policy absence); detect individual mines; clear and destroy a large
and funding for the operations is provided by Deputy Assistant number of mines rapidly and safely; enhance the safety of
Secretary of Defense for Special Operations and Low Intensify deminers; and tools to facilitate mine awareness and deminer
Conflict (DASD(SO/LIC)). training. Humanitarian demining efforts leverage, where
4-41
Meeting the Landmine Challenge - Research and Development in Support of Humanitarian Demining
applicable, the technology investments made for combat The demining staffs of the regional Commanders-in-Chief,
countermine as well as those investigated for remediation of the National Security Council’s Interagency Working Group
defense sites, Explosive Ordnance Disposal (EOD), and the (IWG) and Special Operations Forces representatives identified
clearance of our training and test ranges. and prioritized demining needs and sustainment issues. The
Research and Development sub-group to the IWG provided
Existing technology and equipment used for demining are scope and focus to the developmental activities.
slow, dangerous and man-intensive. During the past two years
the US Department of Defense has engaged in a substantial In compliance with Congressional direction, the NVESD
effort to increase the efficiency and safety of demining with designed, developed and evaluated some thirty items of
technology. As part of the international humanitarian demining equipment for mine detection and clearance that are applicable
effort, Congress provided the Army with $13M of RDT&E to demining and peacekeeping type environments. Several of
funds over FY95 and FY96 with direction to develop and these prototype items performed so well that the United States
demonstrate technologies applicable to humanitarian demining deployed them to support American forces now engaged in
and other Military Operations Other Than War (OOTW) peacekeeping and in demining operations. A by-item
situations. Congress tasked the Army and its countermine description follows this brief list of each humanitarian demining
scientists and engineers to solve unique humanitarian demining prototype technology developed to date:
equipment requirements by leveraging new, proven and
promising technologies that are capable of being used for • On/Off Route Mine Detection
demining and to share them in an international environment.
- Infra-Red (IR)/Ground Penetrating Radar (GPR) Mine
The diversity of the mine threat pointed to the need for Detector
different types of equipment to neutralize them. The short time - Mini-mine Detector
frame of this program dictated a development effort that - Hand Held Trip-wire Detector
maximized the use of existing technology. The requirement to - Ground Based Quality Assurance System
develop equipment for use by host nation deminers with very - Vehicle Mounted Mine Detector
different language, cultures and education levels added to the - K-9 Program
challenge.
• In-Situ Neutralization
The FY1995-96 Humanitarian Demining Technology
Program focused on training initiatives to assist other countries
- Liquid Explosive Foam
in developing effective mine awareness programs and on the
- Chemical Neutralization
development of improved demining technologies. Areas of
- Mine Marking and Neutralization System
interest for technology development were:
- Shaped Charges
- Explosive Demining Device
• Detection of metallic and non-metallic anti-tank and anti¬
personnel mines.
• Mine Clearance
The goal of this program was to rapidly provide suitable - Extended Length Weedeater
technology to detect all land mines, achieve near perfect - Extended Length Probe
removal and neutralization and operator safety and provide - Command Communications Video and Light System
special purpose hand and small power tools optimized for (CCVLS)
demining. This technology will allow the United States to - Sonar Imaging
expand her contributions to assist other countries in developing - Vehicle Protective Kit
effective demining programs. - Demining Kit
- Mobile Training System
- Mine Locating Marker
4-42
Meeting the Landmine Challenge - Research and Development in Support of Humanitarian Demining
- Blast and Fragment Container the vehicle path, vehicle coordinates, the IR and UV targets
- Blast Protected Vehicle received from the target recognition software, and the GPR
detections.
• Other
MINI MINE DETECTOR
Besides the sensor suite, the VMMD consists of video f. An 8mm or standard VHS recorder.
cameras, a Global Positioning System to determine mine
GROUND BASED QUALITY ASSURANCE
locations, remote controlled paint sprayers for marking, an
operator’s command station for operator controls, visual
The Ground Platform Mounted QA Sensor Suite uses IR (3-5
displays and control of sensor functions and parameters, and a
micron and 8-12 micron), UV and video cameras to find surface
skid steer loader vehicle.
and buried mines, trip wires and anti-handling devices.. The
purpose of this system is to confirm that an area believed to be
A portable controller at the operator’s command station
clear of mines is indeed mine-free. A covered, mast mounted
allows access to the various remote sensor functions. The
platform houses the cameras. The camera assembly can mount
camera select capability permits the operator to select the video
to a vehicle or to a static ground position. Signals from the
display source from the visual driving camera, the IR camera or
cameras transmit to a computerized control station. An operator
the UV camera. The portable controller is small, lightweight
at the control station can remotely operate the cameras. The
and has its own self-contained power. A personal computer at
system permits an operator to view an area on a computer screen
the control center runs a geographical information system (GIS)
from any one of the four cameras, and capture an image onto the
display. The GIS is very easy to operate. It accurately displays
computer’s hard disk at any time. The operator can then import
4-43
Meeting the Landmine Challenge - Research and Development in Support of Humanitarian Demining
the image into a program to enhance it in various ways to material in the air if a mine is close to the box. Three methods
highlight a mine. By performing this technique on images from exist for placing the collector boxes; by hand, from a vehicle
any combination of the cameras and comparing the results mounted platform and from the air. When put in place, the
simultaneously on screen, a trained operator can distinguish a collector boxes have markings showing their exact location.
possible buried or surface mine. The software also includes After a period of time, deminers retrieve the boxes and transport
Automatic Target Recognition (ATR) capability to designate them to the dogs. When a dog alerts to a collector box,
probable mines for the operator. deminers can then perform a detailed search, using a free
running dog, in the area from where they retrieved the box.
VEHICLE MOUNTED DETECTION (VMD) SYSTEM Collector boxes that the dogs do not alert to indicate mine free
areas. The Checkmate concept thus allows deminers to limit
This on-road and off-route system detects buried or surface their effort to the areas indicated by the dogs.
emplaced metal or plastic mines. Operators can rapidly switch
the system between on-road and off-route configurations. The C. In-Situ Neutralization
two configurations have a remote controlled vehicle with a mast
mounted camera suite in common. The system includes a LEXFOAM
control station for the operator that permits control of the
vehicle and sensor systems, and provides real-time output. The LEXFOAM is a nitro-methane based liquid explosive foam
station also displays sensor data and video. The control station used for military and commercial blasting agents. It is effective
is both man-portable and able to fit in vehicle-mounted at clearing or breaching mine fields, including those with
equipment racks. sophisticated anti-tank and anti-personnel mines. The closed¬
cell structure of LEXFOAM gives this technology a greater
Two interchangeable detection modules, each containing a shattering effect than devices using the same weight of high
metal detection array and a Thermal Neutron Analysis (TNA) density explosive. A disposable initiation device permits safe
sensor, give the system its on-road and off-route capability. The initiation and detonation of both foam and mines. There are two
purpose of the TNA is to confirm that an object found by the configurations of delivery systems. A man-portable backpack
metal detection array is a mine. The sensor uses a Californium configuration is for small or difficult to reach areas in a
252 radiation source which emits neutrons that penetrate the minefield, A palletized version is for large open areas of a
ground. These neutrons cause the high nitrogen content of land minefield that are accessible by a commercial pickup truck or
mines to emit gamma rays that the sensor head analyzes. The equivalent vehicle.
TNA thus discriminates between metal objects with no
explosive content and land mines. This greatly increases the CHEMICAL NEUTRALIZATION
efficiency of the demining process when compared to metal
detection alone. With metal detection only, deminers must treat This effort involves the use of chemical approaches to
every object found as a mine until they uncover it and establish neutralize mines in-situ. The chemicals change mine’s main
its identity. The system marks mine locations with a paint charge to an inactive state by burning or by detonation.
sprayer. To record and report mine location information, the Alternatives to be explored are:
system uses a combination of Global Positioning System (GPS)
and wheel encoders. a. Autocatalytic decomposition reaction with amines or metal
alkyls in the absence of air (buried mines).
K9 PROGRAM
b. Heterogeneous chemical reaction with amines or metal
Explosive materials in mines emit vapors that trained dogs alkyls in the presence of air (surface mines).
can detect. This program demonstrated the effectiveness of
dogs as mine detectors. The NVESD investigated two c. Detonation upon contact with interhalogens.
alternative K9 techniques:
The program evaluated two versions of the delivery system,
- Free Leash. Under handler control, dogs operate in designated as Gun 1 and Gun 2, to get the chemical into the
suspected mined areas and alert when they encounter a mine. mine and in contact with the explosive. They both operate by
The dogs alert by sitting down next to a mine when they detect firing a bullet into the mine to deliver the chemicals. Both
it. Besides mine detection capability, the test also revealed that systems are positioned above the target mine with a tripod.
dogs are able to detect trip wires. They differ as follows:
- “Checkmate” System. With this method, the dogs do not Gun 1: A small plastic bottle, approximately 1.5” in diameter
initially enter the mined area. Deminers place collector boxes and 3” high, contains the chemical. The capsule sits at the
at appropriate locations in suspected mined areas. The collector lower end of the tripod, just above the surface of the mine. A
boxes are vacuum filters. They trap the scent of explosive rifle caliber bullet, fired from above the capsule, goes through
4-44
Meeting the Landmine Challenge - Research and Development in Support of Humanitarian Demining
the chemical filled bottle and continues into the mine. The • Use of a new 3375 skid steer chassis from John Deer.
chemical falls into the hole in the mine created by the bullet and
neutralizes the explosive. • Remote reversal of the rotation of the flailing head.
Gun 2: In this version, the neutralization chemical is inside • Lighter armor in the flail cover and flail.
the cartridge so there is no capsule. When fired, the bullet
penetrates the mine casing, then releases the chemical to • Improved integration and protection of electronic
neutralize the explosive. controls and circuits.
MINE MARKING AND NEUTRALIZATION • Improved tires to withstand blasts from AP mines and to
spread weight on ground.
This product consists of a polyurethane foam that hardens
after being dispensed. The foam impregnates the exposed parts GRAPNELS
of a mine and then hardens, which renders the fuze inoperative.
The bright color of the hardened material clearly marks the A grapnel is a tethered device used to clear trip wires and
location of the mine. A man-portable dispenser applies the electrically fired mines. A spring loaded launching device
foam. The hardened foam does not destroy mines, but it does propels the grapnel a to given distance depending on the length
make mines safer to handle for subsequent destruction. It also of the tether and on the launch force. As deminers reel the
allows the capability to attach a rope to any kind of mine so that grapnel back towards the launch point, it activates trip wires to
deminers can pull it out of the ground from a safe distance. detonate mines a safe distance away.
SHAPED CHARGES The grapnel and launcher configuration will fit onto the
demining cart (see below) to support demining operations in
Current mine neutralization shaped charges are too large to small or confined areas. A casting device throws the grapnel
use on small anti-personnel mines. Another problem is that attached to a line from a modified deep sea fishing reel. An
hazardous fragments from shaped charge detonations remain electric powered reel recovers the grapnel, which snags
after the explosion. This program demonstrated the tripwires as it returns. The grapnel has some ability to extract
effectiveness of commercially available shaped charges. The oil itself from obstacles, but it is simple and inexpensive enough to
industry uses varying sizes of shaped charges to create oil well be a throw-away item. A heavier grapnel, designed to be
bore holes. A selection of charge sizes allows the use of the launched from a vehicle in less confined areas, is also under
optimum charge against a given size mine and reduces fragment investigation.
waste. Shaped charges are also much less usable as ammunition
compared to standard military charges. TELE-OPERATED ORDNANCE DISPOSAL SYSTEM (TODS)
EXPLOSIVE DEMINING DEVICE (EDD) The TODS adds mechanical mine clearance capability to an
off-the-shelf skid loader. TODS safely removes mines from
This device is a specially designed shaped charge mine sensitive or critical areas such as schools, hospitals and power
neutralization munition integrated into a fixed time delay fuze stations. In addition, it also exposes (unearths) and places
assembly. It produces a penetrating jet stream, which demolition charges onto mines that are too dangerous for people
neutralizes the mine. This design provides the mine to approach such as deeply buried, highly sensitive or booby-
neutralization capability of much larger charges. The EDD trapped mines.
neutralizes anti-personnel (AP) and anti-tank (AT) mines, both
buried and surface emplaced. Modifications to the skid loader are a teleoperation (remote
control) kit, detection capability and clearance attachments.
D. Mine Clearance Individual items on the vehicle include video cameras, a
manipulating arm with a shovel and a gripping attachment, an
IMPROVED MINI-FLAIL air knife, a metal detector, a GPS subsystem, a vegetation cutter
and blast deflectors. The modular command and control system
The Office of Science and Technology (OST) has already allows remote control of each electronic, electromechanical or
performed research with this small remote controlled clearer. hydraulic device. The manipulator allows mechanical pick-up
The Mini-Flail is a small utility vehicle (based on a commercial and placement of in-situ neutralization devices. The TODS will
Bobcat chassis) modified with a remote control kit, a rotating demonstrate the integration of a variety of sensors and clearing
flail mechanism and armor protection. Its purpose is to clear devices under remote control into an effective mine removal
anti-personnel mines from unimproved lines of communication system.
and from off-road areas that are not accessible to large area
mine clearers. Improvements to the original design are:
4-45
Meeting the Landmine Challenge - Research and Development in Support of Humanitarian Demining
The TODS depends on advanced knowledge of approximate BPA returns the processed soil back to the ground. With the
mine locations. An on-board metal detector or video cameras AT and AP mines in plain view behind the path of the berm
pinpoint the exact location. processor, deminers can neutralize them with much greater ease
and safety than manually removing them from the berm.
MINE CLEARING BLADES (MCB)
E. Individual Components
The program demonstrated the effectiveness of mechanical
demining blades designed for attachment to commercial EXTENDED LENGTH WEEDEATER
construction equipment. The purpose of these blades is to
remove anti-tank mines from the path of the host vehicle, and The NVESD evaluated two prototype extended length
collect or expose them for subsequent neutralization. To be weedeaters to increase deminer safety in case of a detonation.
most effective, MCBs could be used in conjunction with the One is a handheld model and the other is wheeled. Both are
Berm Processing Assembly and with anti-personnel mine commercial off-the-shelf (COTS) weedeaters modified for use
detonating rollers. The NVESD tested two configurations of as a humanitarian demining tool. The modifications involved
mine clearing blades. lengthening the shaft of the hand held version and extending the
handle of the wheeled version. The purpose of these systems is
The bucket design not only surfaces mines buried up to 10 to increase the safety of deminers operating in areas where
inches deep, it does so without destroying a cultivated area’s vegetation conceals mines. It is also necessary to remove
ability to grow crops. The shape and arrangement of the tines vegetation for ground coupling of detectors and visual or IR
are similar to a field cultivator. The bucket scarifies the soil and sensors.
leaves it in place Just as a farmer would cultivate a field. At the
same time, it brings buried mines to the surface for subsequent EXTENDED LENGTH PROBE
4-46
Meeting the Landmine Challenge - Research and Development in Support of Humanitarian Demining
positioned outside the safe area, to the command post via RF k. Spade.
link or coaxial link.
l. Mine probe and accessories.*
SIDE SCAN SONAR
m. Explosion container.*
The Side Scan Sonar detects and provides photographic
quality images of very small targets, such as mines and ordnance n. Chemical and/or explosive mine neutralization devices.*
in water with zero visibility. The operator can determine any
variation to the normal environment and allow for pinpoint Note: Items marked "*" are humanitarian demining
accuracy in marking target objects. The system uses a personal technologies under development as part of this program.
computer for control, display, and data storage functions. It also
incorporates a fully integrated navigational plotter and software MOBILE TRAINING SYSTEM
for image enhancement. The complete system consists of a
towfish (600kHz), a coaxial cable, and an IBM compatible PC The Mobile Training System is a suite of multi-media audio¬
incorporating an interface board, cables and the software visual and computer equipment that provides mine awareness
package. training. Effective training on mine recognition and what to do
when encountering mines is a significant means to reduce
MODULAR VEHICLE PROTECTION (MVP) KIT casualties due to landmines. This mobile and multi-lingual mine
awareness training facility trains indigenous personnel on mine
Modular Vehicle Protection is an add-on kit for commercial awareness, safety procedures and what to do in certain
vehicles to shield its occupants from a mine detonation. It situations. There are two versions of mine awareness trainers.
consists of a molded Glass Reinforced Plastic (GRP) ballistic A man-portable system fits into suitcases that trainers can hand
liner module and Aluminum Oxide Ceramic armor fastened to carry to difficult to reach locations. There is a vehicle mounted
the vehicle interior floor, doors, firewall, rear wheel wells and version for more accessible areas.
rear cargo compartment divider. In addition, transparent armor
attaches to the windshield, door windows and cargo PSS/12 MINE LOCATION MARKER
a. The hand cart or small vehicle to carry equipment. The BLAST AND FRAGMENT CONTAINERS
j. Pick-mattock.
4-47
Meeting the Landmine Challenge - Research and Development in Support of Humanitarian Demining
BLAST PROTECTED VEHICLE (BPV) automated multi-lingual capabilities into a system that can be
shared in an international environment.
The Blast Protected Vehicle system uses inexpensive off-the-
shelf material to add anti-personnel mine protection to vehicles
Demining is far more comprehensive than combat mine
being used for humanitarian demining. This program evaluated
clearance. Demining requires as close to 100 per cent
flexible blast blankets mounted under the chassis and
destruction as possible. A description of equipment needs
transparent armor to protect the vehicle from small AP blast and
appears below. There are four major categories:
fragmentation mines. Another modification is the addition of an
internal roll bar to protect occupants should a mine blast roll the Detection Systems, Mine Clearance Systems, Multi-media
vehicle. The roll bar also serves as an attachment point for a Support Systems and Individual Deminer Items.
safety harness and a seat anchor. The blast blanket consists of
Kevlar. Rigid glass fiber structures are also used. Chemically A. Detection Systems
bonded ceramic cement with steel wire for reinforcement is in
the front wheel wells and in the floor of the cab. The The capability to detect surface, buried and shallow water
transparent armor is Lexan protective shield. mines is critical to determine where mines are and are not.
Deminers must be able to locate minefields and individual
F. Other Items mines in all terrain. This will require a high degree of fusion of
the output signals from various sensor types and configurations
VEHICLE TOWED ROLLER specialized for the terrain, weather and environment of the area.
The following list describes the types of detection technologies
Anti-personnel mine detonating rollers towable by of interest:
commercial vehicles provide large area clearance.
- Modular, sensor fused detection systems. Such systems
III. Continuing TECHNOLOGY NEEDS could be ground based or mounted on fixed wing, lighter than
air or rotary winged aircraft. Output must be viewable in real
The Government’s increased understanding of the serious time, and as processed data for rapid analysis. The need for
economic and political implications to any nation with a severe ground based detection systems includes vehicle or fixed-mast
landmine problem resulted in the establishment of a DoD led mounted quality assurance systems to locate minefields and
and funded research and development program for humanitarian mine-free terrain. Along with new ideas for aerial based sensor
demining technologies beginning in FY97. In addition, DoD systems, the NVESD will examine the utility of the Airborne
recently created a UXO Clearance Executive Committee. This Standoff Minefield Detection System (ASTAMIDS) for
Committee is to ensure that there is a well-structured overall demining. The ASTAMIDS is an ongoing Army countermine
UXO clearance effort, and to act as a funnel to provide common R&D program to detect mined areas from sensors mounted on
policy guidance to all DoD activities working in this arena. The an aerial platform.
NVESD, in coordination with the UXO Clearance Executive
Committee, with assistance from SOF components now - Systems that can provide visual image data from cameras
involved in demining programs and with guidance from higher on manned or unmanned platforms to search shallow water
headquarters, will spearhead the new multi-year R&D program covered areas. These systems should find mines along
for humanitarian demining technologies. riverbanks, shallow ponds, rice paddies and other areas where
people would normally wade while carrying out their daily
There are a number of promising technologies that can activities.
enhance demining capabilities. For individual mine detection,
the major technical challenge is discriminating landmines from - Mine and ordnance detectors combined with precise
metal debris - future efforts to improve detection will focus on position locators and transmitters for reporting, recording, and
providing a discrimination capability that includes the fusion of electronically marking mines, minefields and unexploded
multi-sensor information and the incorporation of advanced ordnance. With precise data on the location and composition of
signal processing techniques. In the area of mine clearance, a mined area, clearing teams can proceed directly to a suspected
cost-effective and efficient clearance techniques will be needed location. These systems could use available data links to
to clear landmines in all types of terrain. For neutralization, the transmit new mine data directly to the support element
challenge is to develop safe, reliable, and effective methods to responsible for data analysis. The hand-held marking system
eliminate the threat of individual mines without moving them - would allow predetermined messages and codes representing
new technologies will be needed to economically and safely specific mission situations to be included in the transmission.
neutralize the latest mine threats. For mine awareness and This capability facilitates planning and it defines future
demining training systems, the challenge is integration of the clearance missions and near term mine awareness information
latest computer and training technologies, database links, and for the local population, including where there are no mines.
4-48
Meeting the Landmine Challenge - Research and Development in Support of Humanitarian Demining
Bio-sensors, vapor collectors and element analysis - Special purpose grapnels launched from vehicles to activate
systems to confirm the presence or absence of explosives. A trip wire fuses, expose electronic mine activation links and cut
monitoring system for hazardous and traceable chemicals from tactical or electrical wire used to command detonate mines.
explosive compounds will be valuable in discovering small These devices must be capable of operating in heavy grass and
amounts of explosives. Examples of such systems are highly underbrush. They will be critical to establishing safe lanes in
trained dog and handler teams and bio-chemical, optical or areas where deminers suspect the presence of trip wires or
mechanical devices. These systems should be able to locate the command detonated devices. They will also be the initial thrust
approximate position of landmines, and more importantly to areas where deminers suspect the presence of influence fuse
confirm the existence of explosive compounds associated with activators. Their purpose is to dig up and cut the command
other sensor alerts. wires between mine firing command modules and the mines.
Employment of seismic sensors has already occurred in
Physical marking and fencing of minefields and operational mine fields.
unexploded ordnance to warn soldier and civilian populations
of potentially hazardous areas. - Mechanical landmine destroyers or removal devices, such
as plows, blades and rollers specialized for humanitarian
Infrared sensors, ground penetrating radar, pulsed induction demining. An important design consideration is that cleared
mine detectors, miniature handheld mine detectors and hand areas must still be able to support agriculture. These devices in
held trip wire detectors are examples of technologies that have areas of open terrain to separate mines and other ordnance from
applicability in more than one of the above areas. earth and smaller debris. An additional need is for similar
devices or kits for rugged terrain. These devices must be able
B, Mine Clearance Systems to expose and mark mines and unexploded heavy ordnance
without detonation. Follow-on personnel will destroy the
- A means to remove mines from berms is important for uncovered items. These devices must allow tandem operation
humanitarian demining. It is critical for decision makers to of two or more systems during breaching or demining
understand that breaching and clearing a minefield to military operations where the terrain allows. Removing mines from
standards, and demining are two very different actions even berms is a particularly difficult task and requires heavy, robust
though they use similar technologies. Current military systems able to meet the stress of daily use and of occasional
mechanical breaching equipment is not effective for detonations. These items should work with commercial
humanitarian demining. This equipment moves and deposits the horizontal construction equipment.
mines into a berm. Demining requires that there be no mine
laden berm following area clearance. Berms also cover ground - Remotely controlled clearers capable of safely neutralizing,
that may contain mines. Experience in the Kuwait clean-up digging up and removing buried mines and other explosive
effort proved that one of the more dangerous jobs in clearing devices equipped with anti-disturbance devices in close
mines is removing them from these berms. The design of new proximity to critical facilities. This machine would use foam or
innovative devices to accomplish this task is mandatory, since other suitable compounds to encapsulate sensitive explosive
the failure to do so will result in continued reliance on clearance devices and remove them from high value areas.
by hand.
- In-situ neutralization devices for easy and safe destruction
- Equipment that creates safe lanes through minefields to of mines where deminers find them. Of special interest are
facilitate the start of demining operations. This equipment explosive mine neutralization devices that are not practical for
would either destroy the mines that they encounter, position use as ammunition, and non-explosive chemical neutralization
them for manual destruction by follow-on deminers or a means.
combination of both.
C. Multi-media Support Systems
- Remotely controlled special purpose platforms such as
mini-flails or mechanical diggers for detecting and breaching Mine awareness is one of the most significant factors in the
on-road and off-route anti-personnel mines. These platforms reduction of landmine casualties. Multi-media hardware and
should be able to detect and activate simple pressure, trip wire software systems will greatly aid SOF demining trainers to
and sensor activated anti-personnel mines. This equipment will educate host nation people and to establish demining training
also expose or move anti-personnel mines not activated or programs. Such systems must be robust, and must support many
destroyed by the platform’s neutralization mechanism. Heavy different languages and cultures.
earth tilling machines will destroy any devices in its path by
tearing them apart. - Interactive Mine/Countermine Databases will play an
important part in mine awareness and in demining training. The
main purpose of such databases is to support rapid mine
4-49
Meeting the Landmine Challenge - Research and Development in Support of Humanitarian Demining
hardware analysis. This facilitates identification and training for demining process where the situation requires that it be
host country demining cadres on the types of mines they are performed manually. Specialized probes, light grapnels, long
facing. It also provides mine awareness media for training the reach weed eaters and high power air jets are a few examples.
local population. Demining mission elements will deploy with This one system will increase coverage by at least ten times the
computer, printing, and visual media projection capabilities. current capability.
Group training and mine awareness training media may be in the
form of mine database information displayed on a monitor, and IV. Conclusion
by distribution of printed media such as posters, booklets and
iron-on patterns for T-shirts. A database of worldwide Humanitarian demining requirements are vast and highly
countermine equipment will also provide for more effective varied. The full range of Land Countermine, UXO Remediation,
mission planning. This database must have real-time capability Battlefield UXO and explosive ordnance disposal activities
for two-way communication between the technical support includes detection, marking, reporting, recording, breaching,
assets who sustain and keep the database up-to-date, SOF and clearing, neutralization, destruction, training and mine
their host country demining cadres. awareness. These activities occur simultaneously and
continually throughout the operation. No individual item can
- Portable, mobile training systems with hardware and perform all of these functions. The demining community needs
software for the mine / countermine database and for the multi- wide variety of low and high technology solutions in the field as
media training support system described above. Such a system soon as possible. The FY95 - FY96 program is a significant
must have the ability to rapidly prepare mine awareness and beginning to achieve this goal. A large part of the effort
deminer training media and information as leaflets, radio and beginning in FY97 will build on the success of the progress
television presentations, movie film and posters printed in the made to date.
host country or in a common denominator language. In regions
of low literacy, such systems should use descriptive photos and Demining is a very high visibility international effort. The
line drawings, instead of words, to describe all devices in the Night Vision and Electronic Sensors Directorate engineers and
area and the dangers associated with them. scientists are working on new ideas for technological solutions,
and are continuing to improve on promising alternatives
D. Individual Deminer Items developed to date. The NVESD welcomes assistance from all
of our colleagues.
An important need is a “tool box” of specialized individual
hand and power tools to make demining safer, faster and easier.
A cart stocked with such items will greatly improve the
4-50
Command Communications Video & Light System (CCVLS)
1. Introduction
Figure 1. Command Conununications Video and Light
Antipersonnel (AP) and antitank (AT) landmines are a threat System (CCVLS)
to US personnel both in combat and in Operations Other Than
War (OOTW). The United States Department of State estimates
that 80-110 million mines litter the world, the majority of which That number should be decreasing thanks to numerous efforts
were deployed during the last 15 years. Various reports by the United States and the United Nations (UN). The U.S. has
estimate that 150-500 people are killed or wounded every week formed the Interagency Working Group (IWG) on Demining
throughout the world, mostly innocent civilians. Landmines and Landmine Control to coordinate and administer efforts in
prevent growth and development in emerging or rebuilding this area and has declared a moratorium on the export of
countries, impede repairs to infrastructure, disrupt humanitarian antipersonnel landmines. A Demining Assistance Program has
aid shipments and destroy the moral of the civilians living close been established to initiate research and development into cost-
to the minefields. effective demining techniques.
Landmines also effect the world’s economy. It has been The Humanitarian Demining Team at the Night Vision and
estimated that mines sometimes cost as much as $1,000 each to Electronic Sensors Directorate (NVESD) at Fort Belvoir has
clear. This does not take into account the cost of treatment and developed numerous technologies to support individual
rehabilitation for mine incident survivors or the training that is deminers and demining operations in Operations Other Than
required for demining operations. War (OOTW) scenarios. This paper describes a development
effort that aids demining personnel in mine clearing operations
The world landmine problem is still getting worse. More and training procedures using a video and audio
mines are being laid than are cleared each year. The January communications system.
1994 issue of the New York Times Magazine stated that 340
types of mines are manufactured in 48 nations. Some of the This paper will also briefly discuss the U.S. policy that has
makers are state owned and others are private manufacturers led to the requirement and need for such a system. The
who traffic specifically in government contracts. It estimated Command Communications Video and Light System (CCVLS)
that 10 -30 million mines are produced each year.
4-51
Figure 2. CCVLS provides audio and video
conununications to personnel located at conunand post
4-52
Figure 5. Down range unit receives signal from the walk
around unit and restrnsmits signal to the command post
4-53
Figure 6. Perimeter link unit transmits video from a
second camera source directly to the command post Figure 7. Command post unit is used to receive video
from the mineHeld and transmit and receive audio to
and from the deminer
As previously explained, the purpose of the CCVLS is to
enable the deminer to transmit video and communicate with IV. Test & demonstration results
personnel located at a command post outside the active
minefield so that he/she may receive instructions, techniques The Humanitarian Demining Team conducted an initial
and procedures, and safety recommendations associated with Operational Capabilities Test & Demonstration (OCTD) in
the neutralization of specific landmines. In addition to the November 1995 at Fort A.P. Hill, Virginia. The Command
purpose of communication is the ability to record minefield Communications Video and Light System (CCVLS) was tested
activities that can later be used to train indigenous personnel in in various demining scenarios to determine; 1) it’s effectiveness
mine clearance operations. in assisting the deminer in locating and neutralizing landmines,
2) any human factors associated with the CCVLS equipment, 3)
CCVLS has been designed to be a low cost demining the effectiveness of the two-way radio communication, and 4)
support system that is easy to transport, easy to operate and easy the quality of the video at the command post.
to maintain. The purpose of if s low cost design is so that U.S.
military or possibly host nation countries could afford to The CCVLS was effective in assisting the deminer locate
purchase such items to support their demining operational mines in various environments and conditions using the helmet
needs. It must also be easy to operate as indigenous personnel mounted camera source and the telescopic pole.
will be required to operate the equipment once they have
received the required training. Finally, the system must be Two U.S. Special Operations Forces (SOF) personnel were
easily maintained. Preventive maintenance should be trained as trained on the CCVLS equipment and performed the entire test
well as simple repair procedures. The system must be able to be and demonstration. One played the role as the deminer in the
sustained in the field. field and the other performed the functions of the command post
operator. Throughout testing, the SOF deminer wore the walk
around unit and helmet mounted camera. The perimeter link
camera source was set up in a safe zone approximately 75-100
4-54
In fiscal year 1996, the Humanitarian Demining Team made
several modification improvements to the CCVLS equipment
and performed follow-up testing in August 1996 at Fort A.P.
Hill, Virginia.
4-55
V. Conclusions The Command Communications Video and Light System
(CCVLS) is a communications system that allows hands free
Humanitarian Demining requires new technologies in areas audio communication and video transmission to operators
of detection, clearance, neutralization and training. Because located at a command post outside the demining area. CCVLS
U.S. soldiers are not allowed to enter active minefields, is a low cost, easy to use, easy to maintain and easy to transport
technologies must be developed that the host nation’s deminer system that can support many different humanitarian demining
can use to allow hands free operation while enhancing the missions immediately.
overall efficiency and safety of the demining mission.
4-56
LABORATOIRE DE MICRO-INFORMATIQUE EPFL DeTeC Demining Technology Centre
Abstract
It is difficult to have a clear view of all Establishment) organized the following year
the activities in the field of demining that a technical workshop in Stockholm
take place in Europe. Conferences are [FOA94]. Several technologies were
quite visible, but military projects are not. analyzed. Another meeting was organized
Nationally funded, industrial and private in November 1994 in Ispra (Italy)
research are difficult to evaluate concerning [ISPRA94], including the initial proposal
their real support and hope for success. for a European project.
Nevertheless, it is clear that Europe is
making a significant effort in favour of In 1995, the year of the first Monterey
demining technologies, with several projects Conference [AV/MCM95], there were in
under way and a reasonable spirit for Switzerland two consecutive conferences.
cooperation. The technical meeting in Lausanne
[WAPM95] was a good complement to
1. Introduction the major political and technical UN
meeting in Geneva [UN95], attended by 3
Europe has developed a high sensitivity to to 10 delegates from each of 97
the problem of antipersonnel mines. countries, plus governmental and
Accidents in the former Yugoslavia non-governmental organizations.
involving journalists have been reported in
detail. During the Bosnia conflict, The UN conference was repeated at
3 millions mines have been laid close to Elsinore (Denmark) in July 1996 [UN96],
the heart of Europe, in a charming country without the political side (280 delegates
where many Europeans used to spend from 47 countries). The evolution in one
peaceful holidays. year showed positive and negative aspects:
deminers have improved their SOP
The action of the NGOs before the Vienna (Standard Operation Procedures) and are
Conference {Sept 95), which unhappily did more open to different solutions (toolbox
not register any significant progress, drew approach). But the projects for new sensor
also the attention of the political, systems, announced in Geneva, are still in
economical and scientific community to the their starting phase.
problem of land mines.
The ISMCR Conference on Measurement
2. Conferences and Control in Robotics [ISMCR96]
included two keynote speeches and a
The first humanitarian demining workshop special session on mine clearance. The
was organized in 1993 by the CICR in public was however not really receptive to
Montreux [CICR93]. The problem was humanitarian demining, since the
well stated, but only few technological conference topic concerned mainly control.
aspects were considered. The MD'96 conference [MD96], which
took place in Edinburgh early October
The FOA (Swedish Defense Research 1996, can be considered to be the first
4-57
scientific conference devoted to B. European Community
humanitarian demining; unfortunately it was
restricted to sensor technologies, with too Since 1994, Dr. Linkohr from the
many "academic" papers. There were 180 European Parliament influenced the German
participants from Europe and the rest of government and the EU community:
the world, including the very active
1) to execute a study (done at the
Australia.
Joint Research Center in Ispra)
[ISPRA94] on the state of knowledge
3. Projects
in Europe;
A. Military 2) to recommend a Ecu 50 Mio ($ 65
Mio) R&.D programme to be launched
Military projects are of course not well by the EC and managed by the JRC.
known, and never give a high priority to The solution must be achieved by
real humanitarian demining. France, UK, teams working exclusively in the
Germany, and Israel have the most civilian domain.
important programs. Swedish research is
the most visible, since the FOA (Swedish A call for a prestudy has been launched
Defense Research Establishment) has a in summer 96 and triggered a wide
clear interest in humanitarian demining. The interest. A dozen of groups of partners
FOA announces its projects and have answered the call, showing a
demonstrates in public its working considerable interest. The call for proposal
prototypes: at the Montreux Symposium should be followed, in early 1997, by the
[CICR93], original odor sensor approaches last set of calls of the 4th framework
were presented, which were later program. Due to restrictions in this
discontinued. In Geneva [UN95], a program, it is probable that only Ecu 20
man-portable combined metal detector - Mio will be available for a development to
ground penetrating radar - odour sensor be finished in 1998. Some details have
was announced. Its development, however, been announced by the JRC: a vehicle
did not start as soon as expected, and it carries an extended arm (10 metres) with
is underway now without the odor sensor. an array of sensors at its end; sensor
The FOA works together with Bofors, a fusion and interpretation of data is
private company producing military performed between 3 selected sensors
equipment, which demonstrated a motorized (GPR, induction gradiometer, polarimetric
roller that digs up to 50cm in the soil, infrared sensor). Geographical Information
and breaks antipersonnel and antitank mines Systems, soil parameters data bases,
into pieces if they do not detonate (the signature catalogs for relevant mines,
roller survives to 12 kg of explosives). A working groups with similar programs in
Leopard tanks equipped with one of these USA and Japan, and coordination with
rollers has been sent recently to Bosnia; NGOs are also mentionned.
results will be interesting to know. In
Elsinore [UN96] and Edinburgh [MD96], The role of the JRC is however not
Bofors presented its odor sensor project, very clear. It claims to be a center of
with still several unsolved issued. excellence, ready to help and coordinate
European projects, but JRC is also
ELTA Electronic, Israel, has been develo¬ answering to the call for project, the
ping for several years a vehicle-mounted definition of which it has strongly
array of GPR sensors. The application is influenced. The EC executives clearly
primarily for detecting anti-tank mines at a would like to see the JRC doing the
speed of 4 km/h, and it interests very project.
much some armies.
4-58
C. National projects
4-59
simulation algorithms have been developed, Nations as a forum or catalyst. The
as well as at Nice University. problem is that the UN has no money for
doing such a work.
Walter Krohn in Germany has privately
realized a mechanical engine similar to The major difficulty the researchers and
Bofors' motorized roller, developed initially the industry will be faced with is how to
to convert the forest soil into agricultural test the new equipment, given that any
lands. Two such machines have been error may cause an expensive accident and
tested in Mozambique in August 95, with will stop further financing. Proposed
the support of the German government. solutions have to be certified before being
6400 mines have been neutralized over 56 accepted by demining organization. Most
Ha, during a 5-month campaign with a products now do their final tests and fine
team of 16. tuning within the first customers. This is
not possible with demining equipments.
Industry is interested in carrying out
projects and investing on demining There is a clear need for an international
technology. But when the problem is center working in relation with test
explained to them in an honest way, and facilities at DRA, FOA, JRC, test fields
if there is no military background interest, outside Europe, and on real mine fields in
they are discouraged by the low probability order to:
for return on investment.
- make the information on updated
current researches more easily
Many researchers are interested in working
accessible;
on demining projects, but specific financing
mechanisms do not exist. Projects are - make the tests on simulated and real
started on the initiative of researchers with mine fields easier;
the support of their company, local NGOs - evaluate the proposed solutions,
and in some cases government funds. A claimed to be ready for use by
good level of cooperation exists between demining teams. This should be made
teams from different countries, but by a neutral organism having the
researchers only recently discovered their confidence of the UN, of the NGO
common interest, frequently through the supporting demining activities and of
Web. The Internet is a powerfull tool for the demining organizations (CMAC,
cooperation between scientists, both for MAG, NPA, etc.);
discovering partners of common interest, - support financially either the
and for exchanging recent publications and companies building high technology
results. demining equipments or the demining
teams, in order to obtain the initially
E. International Coordination too expensive equipments being used
for real demining.
It was emphasized at the Geneva and
Elsinore UN conferences that research into
demining technologies is taking place in 4. Conclusion
many countries, usually under army
contracts. Unfortunately, there is overlap, It is time for the political and scientific
duplication, lack of communication, and community to increase its commitment in
competition. Besides the few scientific developing technological solutions for
workshops, there is no mechanism for humanitarian demining. Besides the social
drawing together the researchers to achieve and scientific value, economical
collaboration. To get the best result from justifications can be found, for the benefit
the scattered resources available, research of the countries plagued with mines, and
needs to be coordinated, at national level for the selfish interest of developed
first, but international collaboration should countries providing equipment and getting
be promoted, possibly using the United control of new technologies applicable to
4-60
other fields.
References
4-61
LABORATOIRE DE MICRO-INFORMATIQUE EPFL DeTeC Demining Technology Centre
1 Introduction
this work as soon as possible, with of
Industry and research, previously supported course much more limited resources. This
by military funding are currently quite has happened in Cambodia 5 years ago.
interested in the developement of there and is happening now in Bosnia. It is
activities toward humanitarian demining. called humanitarian demining, but we will
Warfare money is decreasing, and the refer to it "post-conflict demining" in order
media have focused the public attention on to avoid the confusion with the third
the major problem created by antipersonnel
phase.
landmines. Research and development are,
to a certain extent, easier to obtain Phase 3: When the UN removes its
funding for, when the "humanitarian support after 2 or 3 years, e.g. as in
demining" label can be used. The nature Cambodia (Afganistan, Angola, Mozambique
of the research is however quite different are or will be similar), the economy is
when initiated by the army, by the UN, still completely down, the government has
or by a country which, plagued by mines, severe financial problems. Many people are
is trying to peacefully recover its killed or maimed everyday by antipersonnel
economy. mines without any compensation. Non
profit demining organizations, supported by
There are three phases of mine use, the UN and several NGOs (non
during and after a conflict, which will governmental organisations), start to train
remain as long as antipersonnel mines are the cheap local deminers in proding the
not banned. ground patiently in order to find all the
Phase one: During the war, the armies mines. In Cambodia, at the presend rate,
protect their strategic positions with it may take more than 100 years. This is
antitank and antipersonnel mines. The sustainable humanitarian demining.
opponent's activity is to breach into these
minefields, regardless of the material and 2 Post-conflict demining
human losses. We will not come back on The equipment used by the UN and the
these mine warfare aspects. armies are quite traditional: armored
Phase two: When the conflict is over, vehicles carry the expatriate personnel, and
Koweit being an exception, the country is teleoperated tanks pushing rollers trigger
economically ruined and the United Nation the explosion of anti-tank and
calls for the help of the armies of anti-personnel mines laid on roads [1].
goodwill nations, in order to re-establish The exception is the US army which has
the communications, remove the anti-tank a special requirement that the US soldiers
mines from the roads, delimit the are not allowed to walk on an active
minefields in which antipersonnel mines and minefield. The "CNN" effect (any accident
unexploded ordnances could be found. This will be amplified by the media), gives a
very usefull, but expensive activity, tremendous importance to the life of a
requires trained people and special soldier, and motivates the development and
equipments. In general, the objective is to use of very expensive equipments, which
get the local army trained for overtaking no other country could afford.
4-63
An important amount of research funding, deminers has significantly reduced the
worth about 40 Mio dollars last year, has number of accidents over the last few
been spent for the development of years. These so called Standard Operation
technological solutions related to Procedures (SOPs) specify how to organize
post-conflict demining, with the beleif that the work and security on the site. Groups
the ultimate solutions will soon be of 2 or 3 deminers progress in 1 meter
available, if sufficient funding for R&D is wide lanes distant by 20 meters (figure
given. Many projects for teleoperated or 1). Only one man at a time is in the
autonomous robots, cameras, sensors, dangerous area, wearing helmet and west
airborne detection systems which can only shield; the other deminers stay 20 meters
recognize anti-tank mines are developed by
engineers who have not even seen a real
mine-field. The objective appear to be to
make a demonstration on some military test
field, and get the project continued. From
humanitarian deminers' view point, this is
a complete waste of money.
4-64
into the ground, the maximum angle of find other equally well paid jobs ?
prodding being 30 degrees. Several
Let us however mention the four design
actions every 2 cm in the direction of
options for a demining vehicle or robot.
the spot allow to define the shape of
the object. With the aid of a towel, Existing vehicles used for war and
the earth is removed and placed in a post-conflict demining are teleoperated
sand bag (some metal debris may be heavy tanks pushing a roller or a flail. US
inside and have to be removed). The Army used 7 tele-operated M60 tanks
side of a large enough object is pushing rollers in Bosnia [1]. The Bofors
cleaned with a brush, waiting for roller belongs to that category. These
inspection by the section commander. If vehicles are designed to withstand
it is a mine or UXO, the team starts a anti-tank mines, that is 12 kg of
new line, and the object will be explosives. They cannot be brought easily
destroyed by a 200 g explosive charge to the field, due to the poor conditions of
at the next break. Some deminers roads and bridges, and will destroy most
prefer to unfuse the mine immediately, of the surviving irrigation structure. The
with the advantage the ground will not mini-flail [3] is the only known small size
be disturbed by the explosion, no new (1 ton) teleoperated vehicle designed to
additional pieces of metal being spread withstand only 80 grams of explosive of
onto the field. an anti-personnel mine. Several vehicles
have been proposed for moving an array of
5 Need for adequate technologies sensors (metal detector, GPR) over a road
that may contain anti-tank mines; the
Humanitarian demining teams are open to
VAMIDS project in US uses the Schiebel
new technologies, but they must fit their
metal detector array, the ELTA (Israel)
needs. CMAC is ready to do experiments
vehicle carries an array of GPRs.
on their training field near Phnom Penh,
and later on real mine fields. Thirteen The Pemex has been designed as a
metal detectors have been evaluated in lightweight autonomous robot for searching
summer 96, dogs and the Bofors demining antipersonnel mines [4]. The pressure on
vehicle [2] will be tested in winter the ground, 5 kg, should not trigger the
96/97 with the support of the Swedish mines. The sensor head oscillates under
government. the alternating movement of the wheels, in
order to scan a width of about 1.2
What deminers however need in priority is
meters. The project is suspended until an
a better mine sensor, able to distinguish
adequate sensor, weigthing less than 4 kg,
between a metallic debris and a mine. Its
can be installed inside the head.
cost must be low enough. Let us assume
that the sensor is 5 times more expensive The Lemming [5] has also been proposed
(reasonable guess if the additional sensor for exploring minefields. Its smaller size
is a ground penetrating radar, the only brings more constraints to the sensor, but
available technology now). Looking at the it may navigate correctly in some dense
figures given in section 3, this means that vegetation, with however the difficulty to
the cost of the campaign will double. explore systematically every square inch of
Hence, the demining speed should at least the area. Robots with a snake-like shape
double to make the solution acceptable. have been proposed [6]. They would be
But since more than 50% of the time is better in dense vegetation, but where to
spent in removing vegetation, it is just put the sensors?
impossible. The vegetation problem needs substantially
The idea to use a robot for replacing the more attention of the researchers and the
men doing this dangerous work brings the funding agencies. HALO Trust has
same cost/effectiveness issues. A deminer developed and tested in Cambodia and
is paid $100 a month and his family gets Afganistan a $40,000 vegetation cutter
$5000 in case of an accident. Replacing which they claim to be cost-effective. It
these deminers by robots brings also a requires a free safe space next to the
social and economic problem: will they field. For insertion within the above
4-65
described SOP, a light-weight, of an adequate neutral organization will be
man-operated, seed cutter could be of required.
interest. It should not exercise pressure on
the ground, and its maximum cost will 7 Conclusion
depend on its efficiency, bearing in mind
that the man-equivalent time saved is paid Important amounts of money are devoted
less than $1000 per year. to projects related to post-conflict
demining, but they do not bring what
The third major problem humanitarian deminers on the field expect. A better
demining is faced with is getting the understanding of their needs, obtained by
precise boundaries of suspected mine visiting these teams and making early tests
fields. Dogs, as successfully used by on real mine fields with the people able
Mechem in South Africa are one solution. to continue to operate the proposed
Proposals with expensive infrared and equipments is essential. Certification of the
microwave devices embarked on a plane or proposed products will be a long and
helicopter have been partly demonstrated serious process that should be considered
only for clear land (no vegetation) with from the beginning of the project.
big anti-tank mines recently buried. If
odor sensors develop correctly over the References
next years, there is some hope that an
[1] D.W.Parish, J.Moneyhun, " Minefield
autonomous small and reasonably cheap
Proofing and Route Clearing in Bosnia
robot could explore every square meter of using UGV", Symposium of Technology and
an area and come to the conclusion that the Mine Problem, Monterey, Nov 18-21,
there are no mines in it. This would be a 1996.
real major progress. [2] Bofors demining Vehicle, see
http: //WWW. bofors. se/press. htm^Demining
6 Financing scenarios and http://\AAVw.bofors.se/minr.htm
It is clear that demining operations are too [3] Mini-flail developed for the DOD, see
http: //w\A/w. demining, brtrc. com/solut/sminif I.
slow. Hundreds of years will be required
htm and
for removing existing mines at the present http: //vww. demining. brtrc. com/so lut/testpla
pace. We need more money for removing n/miniflai.htm
mines on the field, but just increasing the
[4] J.D. Nicoud, Robots for Antipersonnel Mine
amount given to demining organizations, Search, Control Engineering and Practice,
without working on the technology, will Elsevier, Vol 4 No 4, pp 493-498, 1996
reduce the number of innocent victims
[5] "BUGS: Basic UXO Gathering System,"
(10,000 civilians per year), but will not Presented at the Autonomous Vehicles in
reduce the number of maimed deminers (1 Mine Countermeasures Symposium,
for about 2000 mines). We need (Monterey, CA), April 4-7, 1995. See also
technology to both reduce the number of on the Web:
accidents and increase the productivity of http: //WWW. cs. nps. navy. mil/research/eod /
(Robots form Foster-Miller, IS-Robotics,
demining teams. If the amount of money
K“2T)
given by the international community
doubles every year up to a factor ten, [6] Piraia project at the Swedish Institute of
Computer Science, see
and if 25% of that money is devoted to a http://www.sics.se/piraia/
research centered on the needs of the
deminers, the present manual demining [7] J.D. Nicoud "A demining Technology
Project", Inti Conference on Abandoned
activities will benefit from an immediate
Landmines, MD'96, Edinburgh UK, October
increase, and will have access to improved 1996, Symposium on Technology and the
sensors and equipment, within 2-3 years Mine Problem, pp 37-41
[7]. Since any new technology is
expensive as long the quantities have not
ramped up, it will be necessary to
subsidize the products for some time.
Another part of the problem will be to
qualify and test the prototypes: the support
4-66
Minefield Proofing and Route Clearing in Bosnia Using Unmanned Ground Vehicles
and the Standardized Teleoperation System
David W. Parish
President
Omnitech Robotics, Inc.
2640 South Raritan Circle
Englewood, CO 80110 USA
Telephone: (303) 922-7773
I. Background
In October of 1995 the Office of the Deputy Chief of Staff
Operations, United States Army Europe requested that the Fig. 1: A “Panther” system consists of a turret-less M60 tank with
Unmanned Ground Vehicles/Systems Joint Project Office mine rollers and the STS for teleoperated control.
(UGV/S JPO) upgrade seven Panthers (turret-less M-60 tanks
with mine rollers) with the Standardized Teleoperation Sys¬ width mine clearing rollers. Fig. 1 illustrates a Panther with
tem (STS) for countermine operations in Bosnia Herzegovina mine rollers attached. It is used to proof suspected mined
in support of Operation Joint Endeavor. The STS is a state-of- fields that have been cleared by Combat Engineers and/or
the-art teleoperation system which allows operators to control former warring factions in Bosnia. The Panther was originally
vehicles from a safe distance during hazardous operations. An outfitted with a previous generation remote control system
Integrated Product Team - “Team Panther” was formed to through a contract administered by the UGV/S JPO as an
tackle this urgent and potentially dangerous problem. Team interim measure until the next generation STS teleoperation
members included representatives from the United States controls for the Panther could be designed, produced, tested
Army Engineer School (user and countermine instructors), and fielded.
Omnitech Robotics Incorporated (STS contractor), the United
States Army Missile Command (technical/logistical assis¬ III. The Standardized Teleoperation System
tance), 59th Ordinance Battalion (maintenance team mem¬
bers), and the UGV/S JPO (team leadership). The STS is a modular kit of components which can be added
to any vehicle to convert it to teleoperated control (remote
control with real-time video and audio feedback). The major
II. “The Panther”
components of the STS kit are shown in Fig. 2. They consist
The Panther is a turret-less M-60A3 tank equipped with track of the Operator Control Unit (OCU), Vehicle Control Unit
4-67
vcu
ocu
VTU
PTU
Fig. 2: The main STS components are modular to allow converting any vehicle to remote control, teleoperated control, or semi-auton-
omous control.
(VCU), High Integration Actuators (HIA), System Input/Out¬ The challenge for “Team Panther” was to leverage this experi¬
put (SIO), Video Transmitter Unit (VTU) and Pan/Tilt Unit ence to meet the needs of combat engineers in real world
(PTU). Additional components are also available for autono¬ operations in Bosnia Herzegovina for proofing mined fields
mous control (GPS/INS based navigation), safety radio sys¬ and roads. To assure user satisfaction, high system reliability
tems, and mission and payload specific interfaces like clear and rapid fielding were top priorities.
lane marking system control, Mine Clearing Line Charge
(MICLIC) control, mine detector control, etc. The STS uses a Leveraging success from a Small Business Innovation
serial control bus called Controller Area Network to provide Research (SBIR) contract with Omnitech Robotics, Inc. for
scaleability of the design, allowing as few or as many compo¬ the development of the original STS, and under the direction
nents (HIA, SIO, VTU, etc.) as desired to be controlled by the of the UGV/S JPO led Integrated Product Team, the teleoper¬
OCU and VCU. ated Panther went from concept to reality in less than 7
months. This resulted in the fielding of nine STS kits for con¬
One key feature offered by the STS is the ability to change verting the seven M60 Panthers with mine rollers already in
instantaneously from a manned mode to teleoperated mode Bosnia, including two spare kits. Subsequently additional kits
with the flip of a single switch, thus maintaining the conven¬ and spares have also been supplied.
tional manned capability while offering the advantages of
teleoperation when desired. This allows commanders to take The IPT team used their combined talents to define, plan, exe¬
the Soldier or Marine out of harms way during hazardous cute and manage the project while meeting user requirements.
operations using teleoperation, while not affecting mission This included incorporation of system and programmatic
convenience or reliability since the manned mode is always upgrades derived from suggestions and feedback obtained
available. The STS increases force survivability by reducing from various users while conducting previous ACTD and
loss of life and increasing system survivability during danger¬ AWE efforts. Some of the upgrades include:
ous operations. Additionally the STS enhances mission per¬ • High brightness (daylight viewable) video and status displays
formance by eliminating the operators stress caused by were incorporated
danger, allowing Soldiers and Marines to make clear and well • Minimal vehicle integration time was obtained by using pre¬
thought out decisions in a safe, protected environment. assembled STS “packs” that were dropped into place in the M60
tank turret
• High reliability of the system design was enhanced by perform¬
IV. From Concept to Reality
ing environmental testing of two operational systems at Aber¬
Prior to the “Team Panther” effort, the STS had been proven deen Proving Grounds, including EMI, temperature, humidity,
vibration, and rain testing of the systems. Lessons learned were
in Advanced Concept Technology Demonstrations (ACTD)
incorporated in the production units
and Advanced Warfighting Experiments (AWE) on multiple
• High reliability of the system design was enhanced by requiring
D7G dozers. Ml main battle tank chassis, and HMMWVs. 40 contiguous hours of operational testing using army test per-
sonnel in a realistic scenario at the US Army Engineering
School at Fort Leonard Wood, MO on their robotic vehicle test
track (RTIA)
• High reliability of each unit was enhanced by mandating 72
hour bum in testing of each completed system
• High reliability of each unit was enhanced by including Built In
Testing (BIT) in the system components to assist in fault detec¬
tion, isolation, and recovery
• The Operator Control Unit (OCU) was reduced in volume by
57%, and in weight by 33% from the previous version to
enhance user portability
• Arrangement and spacing of the OCU control inputs and display
feedback were optimized based on human factors testing and
reviews to meet military specifications
• Operator and Maintenance manuals were prepared to assist
users in the field
• A Mobile Training Team consisting of three specially trained Fig. 4: Transportation within Bosnia was via military convoys,
US Army Engineers, was formed to train users in Bosnia including supplies like these individually crated STS kits.
• A Forward Support Team, consisting of three specially trained transported by a material transport vehicle.
US Army personnel was formed to provide ongoing mainte¬
nance support of the Panther systems in Bosnia Fig. 5 shows a photograph of an STS pack being installed in
an M60 Panther (the STS pack is the group of white boxes
V. Supporting Operation Joint Endeavor located in the rear-center of the M60 turret opening). The STS
pack was removed from its individual crate, and set in place
In June 1996, the UGV/S JPO along with the US AES and
on the floor of the M60 tank using the transport vehicle’s
Omnitech Robotics, Inc. dispatched an 11 man Team to Bos¬
crane (or a Combat Engineering Vehicle’s crane). Completion
nia to install, train and maintain the STS on the Panther. This
of the installation consisted of securing the pack, attaching the
team is shown in Fig. 3. The team deployed to Bosnia and
pre-fabricated push/pull cables and mechanisms, mounting
traveled to seven different companies in the 16th, 23rd and
the video cameras and covers, and connecting the electrical
40th Combat Engineer Battalions located throughout the
cables and connectors to the tank’s electrical system. Finally
American sector in Bosnia.
calibration of the servo actuators was performed, and testing
In order to install the seven STS kits in the Panther vehicles as of the system was conducted to verify proper operation. This
soon as possible, the 11 man team would travel for one day, entire process took between seven and ten hours using four to
install an STS kit for one day (7 to 10 hours typical), and train six personnel.
user personnel for one day, then repeat the cycle for the next
A total of 34 soldiers from seven companies were trained to
company. Transportation within Bosnia was provided by US
operate the STS. Fig. 6 shows a photograph of a group of
Army transport convoys consisting of HMMWVs, 5 ton trans¬
combat engineers being trained in the field on operation of the
ports, and other material transport vehicles. Fig. 4 shows a
photograph of two STS kits in their individual crates being
Fig. 5: The STS pack is mounted in the turret of the M60 Pan¬
ther. It controls the vehicle with five push/pull cables and sev¬
Fig. 3: The Bosnia support team for “Team Panther.” eral electrical connectors.
4-69
Fig. 6: Field training of combat engineers was performed by US
Fig. 8: The Panther in action - proofing a suspected mine field
Army engineer personnel from the Engineering School at Ft.
with the track width mine roller
Leonard Wood, MO
Panther systems. Fig. 7 shows a photograph of a combat engi¬
neer performing a mine field proofing operation from the front
passenger’s seat of a HMMWV. The simple and accurate pro¬
portional controls of the STS, combined with bright, easy to
see video and status displays helped users to catch on quickly
to driving by teleoperated control. Although the OCUs were
originally intended to be mounted in an Ml 13 armored per¬
sonnel carrier, in Bosnia the users requested that the OCUs be
set up in the front seat of HMMWVs or inside the turret of
Combat Engineering Vehicles (CEV). Fig. 8 shows a photo¬
graph of the M60 Panther with mine rollers proofing a sus¬
pected mine field.
that the mine rollers failed to protect the tank treads and road
wheel due to the fact that the rollers caster side to side exces¬
sively when the tank is turned sharply. The Panther has been
proven on 3 other occasions so far, detonating anti-personnel
and anti-tank mines that would have otherwise injured or
killed U.S. soldiers or civilians. In all instances the Panther
has accomplished its mission - detonating land mines while
keeping our soldiers out of harms way.
Fig. 7: Operation of the Panther was performed by combat engi¬
neers using an Operator Control Unit mounted in a HMMWV, VL Other STS Applications
CEV, orM113 APC
A. Interim Vehicle Mounted Mine Detector
The soldier’s response to the new STS was overwhelmingly The Interim Vehicle Mounted Mine Detector (IVMMD) is a
positive. On 29 June 1996, the 23 engineering battalion, “A” joint initiative with the PM Mine/Countermine Office and
company, detonated an anti-tank mine during mine proofing UGV/S JPO. Omnitech was directed to prepare four teleoper¬
operations. While the tank sustained damage, the STS contin¬ ated High Mobility Multipurpose Wheeled Vehicles
ued to operate. Fig. 9 shows a photograph of the M60 tank on (HMMWV) equipped with mine detectors. Alternative vehi¬
a transport vehicle after hitting the mine. Notice that the first cles are also possible. Presently, the first two HMMWV vehi¬
road wheel and the track has been blown off It is suspected cles have been teleoperated and are awaiting payload
4-70
integration.
4-71
C Joint Amphibious Mine Countermeasures The mission for this experimental vehicle is “in-stride breach”
where ROCV will travel in a manned mode with maneuver
The Joint Amphibious Mine Countermeasures (JAMC) pro¬ forces until a mine field is encountered. Then ROCV person¬
gram is focused on clearing mines from the shallow water nel will be evacuated from the ROCV to a supporting
mark up to a cleared landing zone for LCAC landing craft. Armored Personnel Carrier (APC) or MBT, and teleoperation
The vehicle used for this application is the D7G dozer fitted will commence allowing unmanned mine field breaching and
with a mine rake, magnetic signature duplicator, explosive net clear lane marking using the ROCV while the accompanying
array, pathfinder clear lane marking system, and a towed MBTs return cover fire if necessary. This promises to reduce
chain array. engineering soldiers’ casualties by giving them a means for
in-stride breach while maintaining the protection of armored
Two systems were developed in 1994 and 1995 to support a
vehicles. By contrast, the current alternative is to have sol¬
Milestone 0 decision and Advanced Technology Demonstra¬
diers dismount and place and detonate explosive charges on
tion (ATD) in November of 1995. Subsequently, two addi¬
the individual mines to achieve mine field clearance, or to
tional systems are being developed, one operational in
wait for engineering vehicles to arrive to clear the mines,
November 1996 and one in January 1997. These second gen¬
potentially while under hostile fire. The ROCV concept prom¬
eration systems feature significant upgrades of the STS sys¬
ises to speed the deployment of maneuver forces by reducing
tem as well as the countermine payloads and integration
or eliminating the need to stop and wait for engineering vehi¬
resulting from evaluation of the first generation systems.
Notable upgrades include reduction of the STS components cles to arrive to breach the mine field, thereby improving mis¬
sion performance and reducing casualties. Fig. 14 shows the
size (OCU), addition of a simple hand held remote control
first ROCV as configured for testing at Ft. Knox, KY in July
only capability to allow driving the vehicles off the LCAC,
and addition of an integrated GPS based navigation and map¬ 1995. ROCV is being developed for the US Army Engineer
ping system to autonomously control the vehicle trajectory, School, Director of Combat Development, with support of the
and map the area cleared. Fig. 13 shows a photograph of the UGV/S JPO.
JAMC dozer as configure for the ATD in November of 1995.
Fig. 13: The JAMC system uses a Caterpillar D7G dozer and
numerous countermine payloads VII. Summary and Conclusion
The Standardized Teleoperation System has demonstrated the
JAMC is being developed by the U.S.M.C. MARCORSY-
ability to remote control, teleoperate and autonomously con¬
SCOM (AWT) with support of Wright Laboratories, Con¬
trol a variety of different vehicles for numerous missions.
struction Automation Group (WL/FIVC) at Tyndall AFB and
Over 30 systems have been developed so far. It has been suc¬
support of the UGV/S JPO.
cessfully operating seven Panther vehicles in different loca¬
tions in Bosnia for 6 months straight with minimal problems
D. Robotic Countermine Vehicle
or maintenance. The versatility, portability, and reliability of
The Robotic Countermine Vehicle (ROCV) is an experimen¬ the STS make it a valuable asset for any mine clearing, proof¬
tal concept vehicle consisting of a turret-less Ml main battle ing, detecting or similar application.
tank (MBT) equipped with a track width mine plow, dual
Mine Clearing Line Charges (MICLIC), and a Pathfinder
clear lane marking system. The first operational ROCV was
demonstrated in 1994, and two additional units with upgraded
capability including semi-autonomous driving capability are
being developed currently.
4-72
Multi Sensor Vehicular Mine Detection Testbed for Humanitarian Demining
Jason J. Regnier
US Army CECOM NVESD
Abstract - This paper presents research results on a tele-operated ultraviolet bands. All the sensors, either those employed or
multi-sensor vehicular mine detection testbed (VMDT) which was earlier models or prototypes, had been proven in mine detection
successfully demonstrated by Special Operations Forces at Fort A. field demonstrations. The visible, UV and IR sensors were
P. Hill in November 1995 in support of humanitarian demining.
forward-looking and were to be used to spot surface-laid or
The VMDT uses a combination of several sensors to detect buried
shallow-buried mines. The metal detector was the primary mine
anti-tank and anti-personnel land mines. The sensor system
included a metal detector array, a thermal neutron analysis (TNA) detector and the TNA sensor was used for confirmation. The
sensor, and commercial cameras to provide images in the visual, metal detector was set with a discrimination threshold consistent
infrared, and ultraviolet bands. The 2-meter metal detector array with the desired high probability of detection for minimum
was shown to be very sensitive to metal, both mines and metal metal mines and scanned the path in front of the robotic vehicle.
clutter on the test range and it performed well as a primary buried Suspicious spots that triggered the metal detector were
mine sensor. The TNA, as a secondary sensor, detected all anti¬ interrogated by the TNA sensor.
tank mines and half pound and larger anti-personnel mines.
In the on-route configuration, the sensors were mounted to
L Introduction allow for the maximum area coverage in the minimum amount
of time. Figure 1 shows the VMDT on-route configuration at
The Government Operational Capability Demonstration Test Ft. A. P. Hill during field testing. The Schiebel flexible 2 meter
(OCDT) was a Congressionally directed program to metal detector was mounted on a wear sheet that was dragged
demonstrate the present state of technologies applicable to along the ground. In combined mode operation, when the metal
humanitarian demining scenarios. Science Applications detector alarmed, the operator stopped the vehicle. The TNA
International Corporation (SAIC) was selected to provide a sensor is then positioned over the detection point by advancing
multi-sensor vehicular mine detection testbed (VMDT) for the the vehicle and positioning the sensor over the suspected mine.
OCDT. Most mine detection systems do not in fact detect The TNA image was analyzed to determine the size and position
mines, but rather anomalies such as dielectric differences in soil of the mine. A ground marking device was provided but not
for radar, induced magnetic fields in the case of pulse induction demonstrated in the field tests.
detectors, or thermal differences in the infra-red. The unique
feature of the VMDT was the potential ability of the system to
be more than an anomaly detector. The VMDT concept is to
use a combination of independent sensors to indicate not just
magnetic and thermal anomalies, but also the presence of
explosives. Targets can then be classified as mines or clutter.
This ability to reduce the false alarms associated with clutter
will be an invaluable part of future mine detection and clearance
because of the time saved in reducing excavation.
4-73
In the off-road configuration the sensor structure was visually by the operator were interrogated with both the metal
exchanged for a robotic manipulator arm. This manipulator was detector and the TNA. Other components include the
a modified Bobcat standard backhoe. The TNA sensor and a navigation system which is comprised of a wheel encoder, a
4x4 Schiebel metal detector were mounted on the end of this differential GPS, and a digital compass. The encoder is
robotic manipulator. The off-road configuration system was not intended to provide short range accuracy on the order of one
demonstrated due to time constraints during the tests. inch. It allows one to display images of the metal detector
signals and to control the motion of the vehicle in positioning
A. Schiebel Metal Detector Array the TNA over a detected target. The differential GPS is
intended to provide long range accuracy on the order of one
The primary sensor on the VMDT was a Schiebel metal meter or less. It allows one to display a symbol of the vehicle
detector called the vehicular array mine detection system in a navigation window and to record the position of detected
(VAMIDS). The flexible 2 meter VAMIDS was designed to targets.
operate from a vehicular platform and detect metallic objects,
including land mines with a very low metallic content. The III. Testing PROCEDURE
detector consisted of a modular electronics unit housed in a
standard military 19" enclosure and individual segments, each The Government Operational Capabilities Demonstration
with a width of one meter. Inside each 1 meter segment is an Test (OCDT) for humanitarian demining was conducted from
array of eight detector heads based on the US Army standard September through November 1995. The various mine lanes
AN-19/PSS-12. The normally audible tone in the presence of and sites included the following surfaces: concrete (with and
metallic targets is converted to a visual display of intensity. without steel rebar), asphalt, sealed and unsealed gravel, an
open grassy field, a plowed farm field, a patterned mine lane in
Although the system is designed to be used along roads, it an open area, a small urban area and an unimproved dirt road.
was also shown effective in off-road application. In the field
tests it was used as a stand alone sensor and as primary sensor The VMDT was operated and evaluated by Special
with the TNA as a confirmatory sensor. Operations Forces demining personnel. The test criteria for the
operators were:
B. Thermal Neutron Analysis Sensor
• To remotely find buried land mines with infra-red and
The vehicular mounted mine detection testbed incorporated ultra-violet cameras
a thermal neutron analysis (TNA) sensor as a confirmatory
sensor to detect the presence of explosives in buried objects that • To remotely find buried land mines with a metal detector
trigger the Schiebel metal detector. The TNA sensor array
incorporated gamma neutron analysis in a compact sensor with
a low intensity isotopic neutron source. These sensors detect • To operate as a combined system to remotely verify if a
ingredients specific to high explosives in the mine. The video or metal detection is an explosive device without digging
configuration of the nuclear sensor assembly was compact with by using the TNA
a 20 microgram isotopic ^^^Cf source, eight Nal(Tl) gamma-ray
detectors and a single thermal neutron detector contained in a To demonstrate these criteria, the operator conducted the
single metal housing. The sensor weight was about 350 lb. and following mission sequence. First, the operators determined
can be used in either the on-route or off-road configuration. what search pattern to run based on terrain. Second, the visual,
IR and UV video equipment is used via tele-operation to spot
In practice the operator positioned the TNA sensor over the surface anomalies. These are noted and interrogated with the
suspicious spot on the ground and activate the sensor. The TNA metal detector and the TNA. As the VMDT moves forward the
signals were processed by the on-board signal processing system operator scans the ground with the metal detector and searches
which formed a TNA image of a buried mine (or ground). The for metallic anomalies. Finally, all metal detections and visual
image used a thermal scale to represent the intensity of the mine anomalies are interrogated with the TNA sensor to determine if
signal. The image was sent to the VMDT operator console for a target is a mine or clutter. Positive targets are then marked for
display. excavation and avoided.
The visible, UV and IR cameras were incorporated into a A. Schiebel VAMIDS Array
single, camera/optics module which mounted to a pan/tilt
mechanism on the tele-robotic vehicle. These cameras were The Schiebel 2-meter VAMIDS array was tested on various
used for initial target identification. Any anomalies detected surfaces including paved roads, unimproved road, a grassy field
4-74
and patterned mine area, along with various calibration lanes. B, TNA SENSOR TESTING
The majority of the testing were done using the array in a stand¬
alone mode. The VAMIDS array proved very sensitive as it To expedite individual sensor testing, the TNA sensor was
detected small pieces of metal and fragments that cluttered the initially tested on various test areas independent of the metal
range. detector area because of the large quantity of metal clutter on
the range.
1) Field Calibration Lane
During the test period, the TNA sensor experienced
A field calibration lane had been setup between the Farm and calibration drifts and had to be frequently re-calibrated. These
the Field Grassy areas. This lane contained a total of 10 mines. drifts often caused distortions in the TNA mine image presented
The position and type of the mines was known, and the purpose to the operator. However, the majority of the TNA data was
of the lane was to allow the operators to experiment with the saved to files which allowed post test processing. Much of the
response signals before tackling the blind areas. data came from the calibrated reprocessed data. During the
actual test period, the TNA sensor demonstrated that it was able
Fig. 2 shows the response taken with the VAMIDS Manager to image buried anti-tank and large anti-personnel mines in real
software on the field calibration lane. The scan was taken at a time. In most cases the TNA sensor was allowed a time budget
vehicle speed of about 3.2 km/hr (2 mph). The large metal anti¬ of 5 minutes to form an image. During tests on the paved road,
tank mines are easily and unmistakably seen. This data was tests were conducted to establish a minimum detection time,
acquired with no detonators in the M14 anti-personnel mine. In which was less than I minute for anti tank mines. These results
this scan, the VS2.2 did not have a detonator and was not seen. are discussed below.
The VAMIDS did however see the TS50 which also did not
have a detonator, showing its high sensitivity. It should be 1) TNA Images of Mines
noted that during calibration and trial scans on this calibration
lane, much of the metal clutter was detected and removed. The The TNA data was recorded as a set of eight detector
clutter seen next to the M15 and the PMD6 remained. These responses from which the TNA image was produced. Based on
data were the clearest recorded showing the effectiveness of the observations in the field, the individual detector response was
VAMIDS in detecting the range of mines of interest to the added to the TNA image screen for the post processed data. In
humanitarian demining program. addition, two features indicative of the presence of a mine were
added, namely the average signal in all detectors and the
2) Grassy Field average signal in the three detectors with the highest counts.
Figs. 4-7 show the TNA images of buried mines of increasing
The first set of data was collected by Schiebel using the mass. Fig. 7 is the Ml5 which is the largest of the mines. The
VAMIDS windows software. Fig. 3 shows the recorded real time response is shown in the bar graphs one through eight.
VAMIDS response data from the runs on the grassy area. As Note that in this image detectors 3, 4 and 5 are off the display
shown in the figure, the large metal anti-tank mines, M15 and
TM46, are easily detected. The figure shows a total of 9 mines
detected. The signal response for the anti-personnel mines, both
VS50 and PMD6, is very similar to much of the metallic clutter.
4-75
Figure 4: TNA image of TS 50 mine
Figure 6: TNA image of PMD6 mine
TNA
scale. The ninth bar graph is the average of the highest three
detectors which are also off scale. The tenth bar graph is the
average of all detectors. The image on the right is a spatial
representation of the mine position under the TNA sensor. The
top of the image is the front of the TNA sensor.
4-76
Table i: VMDT MINE DETECTION RESULTS
Mine Mine Type Explosive/ VAMIDS TNA
Quantity
M15 Anti-Tank Mines -Metal RDX/15.4 lb. Detectable Detectable
TMD44 Anti-Tank Mines -Metal TNT Dynamite Detectable Detectable
TM62 Anti-Tank Mines -Metal HE/15.4 lb. Detectable Detectable
M19 Anti-Tank Mines -Plastic Detectable* Detectable
VS2.2 Anti-Tank Mines -Plastic (data not recorded)* Detectable
VS 1.6 Anti-Tank Mines -Plastic HE/4.1 lb. Detectable* Detectable
PMD6 Anti Personnel Mines - TNT/0.44 lb. Detectable Detectable
M14 Anti Personnel mines Tetryl/0.06 lb. Detectable* Below demonstrated
TS50 Anti Personnel mines T4/0.11 lb. Detectable* marginal
(detected once)
*when detonators present
4) VMDT Operational Mode Testing These tests on the unimproved road were the only true
operational tests of the combined VAMIDS/TNA sensor system.
VMDT was tested in an operational mode in which the They were successful in demonstrating that the two sensors
VAMIDS and TNA worked together with the metal array could be used to identify suspect locations and confirm the
triggering and the TNA immediately interrogating detected presence of mines. The VAMIDS is a very sensitive, rapid
targets. In this way it was demonstrated that the TNA detector of metal but metal is not specific to mines. The TNA
performed well as a confirmatory detector - clearing false signals are specific to mines with an interrogation time much
alarms and confirming the presence of buried mines. longer than for the metal detector. Thus the TNA is well suited
as a confirmatory sensor.
The unimproved road consisted of a relatively flat section and
an adjacent vehicle tire rutted area. During the tests, the 5) Special Note on Limitations of TNA
VMDT advanced until the VAMIDS had an alarm. The VMDT
would then stop and move back and forth to maximize the The net nitrogen signal is to the first order proportional to the
VAMIDS response to determine the suspected mine position. nitrogen mass in a mine, but is modified by burial depth, stand
The ground positions of the VAMIDS triggers were then off and other effects. The nitrogen mass in the mines used in the
marked with spray paint. The VMDT vehicle was then moved various test areas is shown in Table 2.
forward under tele-robotic control, and the TNA sensor
positioned over and lowered onto the suspect location. Those Practical field experience indicates that at least a half pound
VAMIDS alarms that were declared to be mines by the TNA block of TNT (such as the PMN or PMD6 AP mines) is
operator were marked with a flag. required to activate the detector reliably for real time operations.
Potentially the system will detect the smaller mines such as the
On the rutted part of the unimproved road, the VAMIDS TS50 in post processed applications where a field is swept, the
sensor alarmed on the two metal anti-tank mines and a coffee data stored, then analyzed later.
can buried as clutter. The TNA sensor confirmed the two metal
mines and indicated the can did not contain explosives.
Table 2
Mine Mine Type Explosive Qty(gms) %N N-gms
M19 Anti-Tank Mines -Plastic CompB 9091 31% 2773
TMD44 Anti-Tank Mines -Metal 9545 18% 1718
M15 Anti-Tank Mines -Metal RDX 7000 38% 2660
TM62 Anti-Tank Mines -Metal HE(TNT) 7000 18% 1260
VS2.2 Anti-Tank Mines -Plastic CompB 2136 31% 652
VS 1.6 Anti-Tank Mines -Plastic HE(CompB) 1864 31% 568
PMD6 Anti Personnel Mines -large/metal TNT 200 18% 36
M14 Anti Personnel mines small minimum metal Tetryl 29 24% 7
TS50 Anti Personnel mines small minimum metal T4 38% 19
VS50 Anti Personnel mines small minimum metal RDX 43 38% 16
4-77
C. UV/IR CAMERAS can be made about the ability of the visual, IR and UV
sensors to satisfy the demonstration criteria.
The UV and IR cameras were demonstrated to be
functional during the early field integration period. The The TNA sensor was successfully able to confirm or deny
testing of these sensors was given a lower priority than the the presence of explosives in all anti-tank and some anti¬
testing of the VAMIDS and TNA sensor. No tests were personnel land mines in real time during the demonstration.
carried out in the optimal conditions for use of these cameras. In addition, the field data was stored and analyzed in the
post-test period. This analysis showed the TNA sensors were
IV. Conclusions able to detect all anti-tank mines and the anti-personnel mines
of a half pound or more. For the smallest anti-personnel
The tele-operated vehicle generally performed well and mines the TNA had marginal performance. The TNA was
satisfied the demonstration criteria in that all vehicle mine shown to be insensitive to road and field surfaces, and clutter
detection operations were conducted remotely. The Schiebel objects. Overall, the TNA demonstrated the ability to
2 meter VAMIDS array was shown to be very sensitive to function as a confirmatory sensor and met the demonstration
metal, both to mines and to metal clutter on the test range and criteria for large mines.
easily met the demonstration criteria. Field observation of
the VAMIDS performance on the paved road showed it could Authors:
detect the buried metal anti tank mines even with rebar in the
cement. Overall it performed well as a primary in the Douglas R. Brown*, Joseph Bendahan, Giancarlo Borgonovi,
combination of sensors. The video subsystems as detectors Delmar Haddock
were not adequately tested due to the poor environmental Science Applications International Corporation
conditions present during the demonstrations. No assessment
Jason J. Regnier
US Army CECOM NVESD
4-78
Tele-operated Ordnance Disposal System for Humanitarian Demining
Jason J. Regnier
US ARMY CECOM NVESD
Joseph Foley
OAO Corporation
I. Introduction
Figure 1: Manipulating arm with bucket attachment and metal detector
A. Background
B. Demonstrated Results
The United Nations estimates that there are approximately
110 million land mines laid in 62 countries in combat zones, A solution to the problem is the Tele-operated Ordnance
and civilian commercial and agricultural areas. At the Disposal System (TODS). The TODS is an off-the-shelf skid
current rate of clearance, time estimates range from hundreds steer loader modified for tele-operation with mechanical mine
to thousands of years to complete the cleanup. This rate is clearance capability (see Fig. 1). In 1995, it was selected to
clearly unacceptable for humanitarian and economic reasons. demonstrate mine clearance capability in the Congressional
In the humanitarian sense, the majority of the casualties are directed Operational Capability Demonstration Test (OCDT)
noncombatants and frequently children. Financially, at Fort A. P. Hill. The following paper describes the system,
agriculture based economies cannot withstand the denial of the test requirements, and discusses the results of TODS
vast tracts of farmland because of the mine threat for demonstration.
hundreds of years. The application of technology not only has
the potential to increase mine clearance to an acceptable rate, The TODS is composed of three main subsystems. First is
but will also provide safer methods. Deaths and injuries due the chassis which is a diesel powered commercial skid steer
to mines occur not only in the general population, but also to modified for full tele-operation capability. This includes two
personnel specifically trained to conduct humanitarian remote cameras, a portable base controller, and differential
demining operations. Specific instances include Kuwait in global positioning system (GPS) navigation system. All
which, 84 demining experts were killed or maimed during the other subsystems are attached to and controlled from the
cleanup, and of a local demining team of 49 in northern chassis. The second is the manipulator arm which is a
Somalia, 17 were killed or injured in accidents over the past commercial backhoe that can be used with either an
3 years[l]. A solution is needed to solve the mine clearance excavation bucket or a mechanical gripper. They are
problem by increasing the rate of clearance with the interchangeable and can be swapped in minutes. The bucket
constraint that casualties are unacceptable in humanitarian attachment is used to excavate mines or dig trenches. A
demining. commercial metal detector is used in conjunction with the
bucket when needed to pinpoint targets that are either
4-79
bucket when needed to pinpoint targets that are either
approximately or electronically marked. In place of the
bucket, the gripper attachment can pick up mines or place
demolition charges near mines too sensitive for human
approach. The arm also has an articulating air knife which is
used to clear soil with compressed air from targets to aid in
identification. The third subsystem is a commercial bushhog
which is installed for specific missions in place of the
manipulator arm. Frequently in humanitarian demining, the
war ended years before and footpaths and farms are
impassable by deminers because the abandoned mined areas
are overgrown. Deminers must cut grass by hand and search
for and clear mines on an inch by inch basis. With an added
commercial bush-hog, the TODS can be used from a standoff
to safely cut heavy brush suspected of being infested with
mines so that detection and clearance equipment can follow.
4-80
operator identifies the mine, uses the bucket to excavate it. material types as shown in Table 1. 87% of the mines are
The mine is then stored and the operator digs a disposal pit. “old,” meaning that they have been buried for 4 months or
The bucket is replaced with the gripper arm so that the more. The remaining 13% are “new,” meaning that they have
operator can pickup, transport, and place the mine with others been in the ground for 1 week or less. The significance of
in the disposal pit. using old mines is that sufficient time will have passed to
allow for rainfall and settling of the fine soil around the mine.
The test criteria specify that this operation was to be The result is that the soil is tightly packed around the mine
conducted in a time equal to or less than the current method case and they are more difficult to excavate than freshly
of manual mine clearance, and without detonating the mines. buried mines. These long buried mines represent the
This procedure was repeatedly conducted and timed so that majority of the threat faced in Humanitarian Demining in
a reliable assessment could be made as to the average time which the conflicts are long over. 71% are shallow buried,
these operations could be completed. Also, most of the which is defined as 1 inch or less. The remaining 28% are
mines had smoke fuzes or other mechanisms to indicate if a deep buried, which is 6” or more. This is an approximate
mine fuze was activated during clearance. The test operators mix of the depths that are encountered in real demining
and evaluators were a combination of Special Operations situations.
demining personnel as well as contractor and Army program
personnel. The following section details the numbers and The mines were all excavated in a random order, so that
types of mines used. the depth and type of mine were unknown to the operator.
His only indicator was a flag near the location of a target.
C. Threat Land Mines Used
Tablei
ANTI-TANK MINE TEST POPULATION
AT Mine Population OTY Breakout by depth Breakout by age in ground Breakout by Type
Size r ■6” : Old New Plastic Metal
Large (16 - 25 lb.)
M19 (square) 23 13 10 16 7 23
M15 (round) 34 25 9 30 4 34
Medium (8- 151b.)
TM-60 (round) 10 10 10 10
M-6 (round) 20 14 6 20 20
Totals 87 62 25 76 11 33 54
Tablei
ANTI-PERSONNEL MINE TEST POPULATION
AP Mine Population QTY Breakout by Age in Ground Breakout by Type
4-81
III. Discussion of Test Results
A. Navigation.
The test criteria specified that the GPS navigation system had
to be 20 meters accurate at a minimum, and potentially guide the
TODS to within two meters of a specified location. The system
Figure 3: Bushog Attachment
worked well enough to guide the vehicle in real time to
designated minefield boundary markers. The TODS easily
satisfied the minimum capability. It proved to be consistent The vegetation clearance operations were conducted in
through a series of seven test runs and multiple position several mission scenarios. None of these clearance scenarios
verifications tests throughout the demonstration period. The include time required for mine detection which was not part of
rate of advance using only the GPS was limited only by the this program. The first was to clear an area containing high
maximum forward speed of the vehicle which was grass, weeds and small trees. This 20 x 30 meter area was
approximately 6 mph. successfully cleared with no operator line of sight in about 30
minutes. The second was to clear the side of a hill covered with
For individually marked targets, the TODS GPS system had high grass that was suspected of having anti-personnel mines.
a precision of plus or minus 1.1 meters limited by the last This 2x8 meter area was more challenging because of the off¬
significant digit in the operator’s video overlay data display. road nature of the terrain but was also completed in 30 minutes.
When the TODS was parked over a surveyed GPS marker it The third was to clear a simulated off road area of heavy brush
took about four minutes for the system to hone in on the exact with varying terrain. This 4 x 50 meter area was successfully
location rounded to the precision of the system. From this cleared within 30 minutes so that off-road detection devices
range, all marked targets were easily identified using the visual could be brought in to search for mines. All of these areas
cameras of the tele-operation system. would have taken hours if done with the current method of hand
clippers because of the danger of accidentally detonating mines.
B. Vegetation Clearance.
The fourth scenario incorporated the ability to clear a pattern
The test criteria for the bushhog are a qualitative statement on around various obstacles simulating an urban area. The backhoe
the ability of the system to cut light and heavy vegetation mounted camera was also moved to the reach riser boom to
without endangering the deminer. This required that the cutting provide a fixed, side-angle view of the cutter. This operation
operations be completed via tele-operation with no operator line was also easily completed.
of sight. Over the course of the test period, the TODS proved
to be effective at clearing large areas rapidly, and tight areas The vegetation cutter is a standard commercial item that met
with little room for movement all via tele-operation. Also, the qualitative test criteria. Deminers can utilize the tele-operated
operators of the TODS had no trouble with any vegetation from bushhog attachment as an effective tool to safely clear brush in
high grass to densely vegetated areas that even contained small hazardous conditions.
trees. An important note for the reader is that the system only
clears as far into the minefield as the cutter can reach (about two C. Excavation.
meters) before the wheels of the chassis enter. Thus clearing
procedure requires that an edge strip of a minefield is cut, The first part of the criteria for excavation required that the
checked for mines, cleared of mines, and the process is repeated mines be precisely located and identified before they were
(see Fig. 3). The key point is that the operator operating the excavated with the bucket. This way, the bucket could start
TODS remotely and is not exposed to the effects of a sensitized behind the mine and scoop underneath so that no contact was
mine or booby trap detonation. made with the pressure plate on top. The air knife mechanism
4-82
“identify” is defined to be the instant that the operator
recognizes the mine type. This is the time period when the soil
is being cleared from a suspected target so that the entire top
surface can be viewed, “excavate” is the time elapsed until the
mine is in the in backhoe bucket. The total elapsed time is the
average entire event for each mine with the high/low
disregarded and this represents a conglomerate of five different
operators over a period of seven days. Transport time was
site/scenario specific and was not used in the calculations.
The key result is the time required for excavation. For a total
of 151 mines, almost all mines were excavated on average in
less than 10 minutes. This rate is comparable, and probably
better than the current manual method. The most important note
is that this was all completed via teleoperation and even in the
event of an accidental mine detonations or booby trapped mines,
there would have been no casualties.
4-83
Figure 5: Manipulating arm with bucket Figure 6: Gripper excavating M19 AT mine
placing mine in trench for destruction
Initially, the TODS was not meeting the criteria of avoiding individual mines and stack them in the trench for disposal. The
mine detonations. However, all of the fuze initiation incidents operators were easily able to use the grippers even in muddy
occurred early in the test program. The activation of the anti¬ conditions to complete the task without activating a single anti¬
tank mine fuzes was a function of operator training and tank or anti-personnel mine fuze.
familiarization. As the operators learned of the results of each
excavation that caused a detonation and discussed the causes, A final important note about the excavation of the mines
new excavation techniques were developed that precluded any concerns two special mines that simulated unexploded ordnance
further fuze activation. The newly developed techniques (UXO). The excavation of these two targets was not included
included using the air knife to not only find and identify each in the test criteria but was meant to establish a bound on the
target, but also to completely clear the top of each mine. This capability of the TODS. The two inert full weight M-15 mines
procedure insured that the operator could not place the digging were buried 36 and 38 inches respectively from the surface four
bucket in a position where it could accidentally contact the months before the demonstration. These represented UXO that
pressure plate of the mine. Also, the excavation process was had penetrated the ground on impact, or very long buried mines
adjusted by adding the following extra step. The operator that were in an area where soil washed over and built up as seen
would dig a small trench behind the mine so that the bucket in Kuwait and Southeast Asia. The operators were easily able
could scoop far below the mine to remove it from the soil. This to excavate these targets from these depths. This result
prevented the bucket from spearing the side of the mine or exceeded the expectations of the capabilities of the TODS.
skipping up over the top edge. These techniques prevented
further fuze activation, which is an important part of the test All of these tests and results led to the conclusions in the
criteria. These same techniques were also used to avoid following section.
activation of the small mines and were demonstrated
successfully in the later stages of anti-personnel mine IV. Conclusions
excavation.
The TODS operators were able to meet the test criteria in the
As the mines were excavated they were either transported in following areas: safety, ease of operation, vegetation clearance,
the bucket to a common staging area or placed along side the GPS and visual navigation to both minefields and individual
excavated hole. Once a sufficient amount of mines were mines. Also, criteria for tele-operation and remote control
unearthed, the TODS was used to dig a trench approximately 15 capability, mine manipulation with the gripper, mine
feet long, two feet deep and 18 inches wide. See Fig. 5. The identification and preliminary excavation with the air knife
TODS manipulator arm was reconfigured from the bucket without activating the fuzes were also met. The digging bucket
attachment to the gripper assembly so that the mines could me was capable of excavating all mines from the smallest plastic
precisely placed in the pit. The system was used to pick up the anti-personnel mine near the surface to the largest metal anti-
4-84
tank mines buried up to twelve inches. Operators also References
4-85
4-86
Mine Marking and Neutralization Foam
Abstract—This paper presents research results concerning a and booby traps cause more casualties in low-intensity con¬
system for effectively marking and neutralizing anti-personnel flict theaters than any other factor. Further, the mined and
land mines by employing a unique single-use kit containing a booby trapped area continues to be a dangerous threat to the
rapid-rise-and-cure, rigid-foam material* Because of its sim¬ area’s inhabitants long after a conflict has ended. The United
plicity and ease of use, this foam is highly suitable for inunediate Nations estimates that 110 million live land mines remain to
use in humanitarian demining. The foam and dispensing system be located and neutralized in 64 countries. Many of these
were developed by Hughes Aircraft Company, El Segnndo, CA, cheap, easily portable weapons can be detonated by the pres¬
as part of the Congressionally directed Humanitarian Demining sure of even a child’s footstep. About 100,000 mines are
Technology Program. found and disarmed each year, but meanwhile millions more
Each kit consists of liquid foam components packed in a twin
are planted. Current figures show that, for every mine neu¬
disposable cartridge together with a mixing nozzle, all sealed in
an aluminum foil bag. The foam expands to many times the tralized, seven more are being planted [1].
liquid volume, forming a bright orange-red, easy-to-see mound Widely diverse mines are available today. Mines used in
over a mine. It is applied over exposed mines using a manport- Third World countries range from clandestine, homemade
able, manual, double-caulking gun dispenser or by simply mix¬ devices fabricated from indigenous materials to sophisticated
ing the components in the foil packaging bag and pouring the military devices of American, NATO, Eastern Bloc, and
contents around the mine. Operationally, the foam impregnates Third World origin. These mines use military explosives
the exposed parts of a mine prior to curing and hardening, such as pentaerythrite tetranitrate (PETN), cyclonite (RDX),
rendering the fuse inoperative. The bright color of the and C4; commercial explosives such as dynamite, TNT,
hardened material clearly marks the location of the mine.
black power, and nitroglycerin; and homemade explosives
While the hardened foam does not destroy mines, it makes them
such as ammonium nitrate (fertilizer) and fuel oil (ANF),
safer to handle for subsequent destruction. The cured foam
distributes loads applied (l.e., foot pressure) over a much larger potassium perchlorate and aluminum powder, and sodium
surface than a mine’s pressure trigger area, substantially chlorate and petroleum jelly.
reducing the likelihood of detonation. It also enables the Triggering devices that detonate the explosives can be
attaching of a rope to any anti-personnel mine so that the mine divided into two basic types: pull and pressure activated.
can be pulled from the ground at a safe distance. The rope is Military-designed pressure devices such as the MlAl and
placed next to the mine, and the chemicals are spread over it M5 (mousetrap) and Ml pull firing devices (and their Eastern
and the mine before the foam hardens. Bloc counterparts), utilized for their reliability and resist^ce
In the 40 field tests conducted by Army Special Operations to the environment, are fairly common. Clandestine devices
Forces on ten different mine types, flie foam marked, neutralized,
made of indigenous materials are, however, also encountered
and aided in the removal and destruction of anti-personnel land
frequently. All of these devices can be silently and quickly
mines. It functioned in cold and warm weather, and under wet
and dry conditions. The foam neutralized boA pressure-fused deactivated by a rigid, foam-in-place material interfering with
and tripwire-fused mines. The foam will not impede the effec¬ the operation of triggering devices (Fig. 1). The hardened
tiveness of conventional explosive charges in destroying the mine. foam can prevent the firing pin on an Ml pull firing device
The foam consists of a water-blown, two-part 50:50 mix ratio from striking the percussion cap or prevent the metallic con¬
polyurethane foam to which a dye is added. The nonflammable, tacts on a clothespin device from closing (completing the
environmentally benign foam material is dispensed as a liquid circuit and initiating an explosion).
and cures to a hard, smooth surface. This foam material was
selected because its rise and gel time are fast enough for field use
even at low temperatures, but do not cause the foam to lift the
mine as it expands, possibly triggering anti-tamper devices. At
room temperature, the foam rises and is tack free within
5 minutes. At near-freezing temperatures, approximately dou¬
ble this time is required. The foam material components and
dispenser were intentionally chosen to be low-cost, commercially
available items. Commercial sources for packaging the foam
kits are being developed.
I. Background
The planting of land mines and booby traps has become Fig. 1. Mine encapsulated, marked, and disabled by foam. An Ml pull
triggering device and tripwire are fully encapsulated in foam that
one of the chief ways of providing effective, low-cost barriers
prevent the pin from dislodging and detonating the mine.
against land forces. Analysis of past conflicts shows mines
4-87
11. Introduction TABLE I
Mine marking Foam Formulation—A Combination of Two Commercially
The purpose of this project was to develop a simple, hand¬ Available Materials
held, portable, single-use system for marking all types of land
mines and disabling selected types of anti-personnel land Isocyanate Part A Polyol Part B
mine triggering devices. This task included fabricating sev¬ Material designation (pbw)a (pbw)
eral kits of a selected mine marking and encapsulating mate¬ PDL 328-4 FAST 100 73
rial and associated dispensing equipment.
R6-OR9014 13
The objective was to develop a rigid polyurethane foam
with suitable colorant to mark land mines and disable them apbw = parts by weight
A. Foam Formulation
The foam formulation chosen consisted of a two-part MDI
(polymeric diphenylmethane 4,4 diisocyanate)-cured, water-
blown polyurethane foam mixed with a bright fluorescent
colorant. The final formulation (Table I) consisted of a modi¬
fied commercial foam made by Urethane Technologies, Inc.
(Polymer Development Laboratories Division, Orange, CA), Fig. 2. Side view of foam dispensing gun showing added splash shields.
PDL No. 328-4 FAST, combined with a Magruder Color
Company (Radiant Color Division, Richmond, CA) orange- TABLE II.
red pigment, No. R6-OR9014. Other f oam Materials Evaluated
The foam was dispensed using a Techcon Systems
(Carson, CA) No. TS 529S double-cylinder manual caulking- Material
type gun. Polycarbonate splash shields (6 X 8 X 1/8 in.) designation Manufacturer Reason not chosen
attached to the front and rear of the gun (Fig. 2) also provided
a convenient method for resting the gun in an upright posi¬ PDL 707-2 Urethane Technologies, Inc. Lifts mines during
tion. (These shields prevent foam chemicals from contacting rise and cure;
the operator. They do not provide sufficient protection, how¬ surface appearance
ever, to prevent injury from concussion or debris should a PDL 707-3.5 Urethane Technologies, Inc. Lifts mines during
mine detonate.) The foam was packaged in a pair of Techcon rise and cure
System II No. SII300S polyethylene 300 cc cartridges and PDL 328-4 Urethane Technologies, Inc. Too slow to rise
dispensed through a Techcon No. TSD 160-830 static mixing (regular)
head nozzle that swirls and mixes the two urethane chemical Stathane 4802W Expanded Rubber and Too slow to rise;
components as they are extruded. Plastics Co. pigment
Six other foam materials were evaluated before the mate¬ compatibility
rial described above was selected (Table II). Stathane 4804 NF Expanded Rubber and Too slow to rise;
Tests with simulated mines showed that the rise and set Plastics Co. pigment
times of the PDL 707 foams were too rapid. When applied, compatibility
they tended to lift mines out of loose soil, possibly causing an Stathane 6603 WF Expanded Rubber and Too slow to rise;
anti-disturbance device attached to an actual mine to trigger. Plastics Co. pigment
Also, the cured surface of this family of foams was rough and
compatibility
uneven.
4-88
Foam materials from Expanded Rubber and Plastics, Inc. nearly 3 feet the mixing distance achieved using the static
were too slow to rise and showed compatibility problems mixing process. Here, an intermediate compressive strength
with the colorants studied. This situation resulted in streak¬ value was achieved. This experimental extension is not prac¬
ing and a roughened surface, indicating incomplete nuclea- tical due to increased back pressure, resulting in excessive
operator effort to dispense the foam. (The single-nozzle
tion of the foam.
system is preferred because the operator is not exposed to the
B. Foam Properties unreacted foam chemical components when using the dis¬
The average compressive strength of the foam formulation pensing gun.)
selected varied from 59.1 to 83.1 psi (Table III), depending It was also noted that the interior areas of hand-mixed
on the mixing method used and whether the foam was pig¬ foam appeared to be less friable and more homogeneous than
mented. These values, considered more than adequate for the foam samples made using the static mixing head. Neverthe¬
intended application, met the program requirement of 25 psi. less, all specimens made in the laboratory and in field tests
Field tests also showed that as little as 1/2 in. of foam cov¬ with any of the mixing methods evaluated provided more
ering a pressure triggering pin on an M-16 mine was than adequate strength and durability for the foaming
sufficient to prevent tripping the trigger when stepped on. application.
Only after foam-encapsulated mines were removed from the
ground and jumped on with both feet was it then possible to C. Colorant Selection
Table IV shows the dye and pigment colorants studied for
trip some triggering devices.
Table III shows that density values of the pigmented foam this program. Dyes, being liquid in form, tend to be more
were slightly less than for the unpigmented material. The compatible with urethane foam chemicals than powder pig¬
density of the unpigmented foam did not, however, vary sig¬ ments. In general, however, the dye colors are not as bright,
nificantly with the mixing method used. Compressive and they are not ultraviolet (UV) fluorescent.
strengths of the pigmented foam were approximately In all cases, colorants were added first to the polyol com¬
10 percent less than values obtained with unpigmented foam ponent (Part B) by mechanical mixing. Colorants were not
when both were dispensed using the static mixing nozzle. added to the isocyanate portion in advance to avoid intro¬
While density was not significantly affected by the mixing ducing moisture into this component (which reacts with water
method used, the compressive strength of unpigmented as part of the overall foam reaction). Some settling of the
machine mixed samples was more than 20 percent greater colorants occurred in all formulations with the various foam
than samples made using a single static mixing head nozzle materials tested. The R6-OR9014 pigment selected showed
to combine the foam components. This difference was due to the least settling while providing the material compatibility
better nucleation of the foam-generated carbon dioxide bub¬ and bright orange-red color desired. This material also pro¬
bles that create the foaming action and was aided when air vided UV fluorescence for nighttime observation.
was stirred in with the foam chemicals using the hand-mix High concentrations of colorant increased the viscosity of
and machine-mix methods. The static mixing head does not the foam chemicals beyond that acceptable for proper
allow air into the chemicals as they travel down the nozzle. dispensing. A 7.0 percent concentration of the R6-OR9014
The hand and machine mixing processes allow for better pigment, selected as optimum, was added to the polyol as a
mixing because the nozzle permits only a limited mixing time 15 percent concentration. When the polyol plus pigment and
as material is swirled through it. This behavior was demon¬ isocyanate components of the foam were mixed, the effective
strated by attaching three nozzles end to end, extending to concentration was 7.0 percent by weight.
TABLE III
MECHANICAL PROPERTIES OF CURED PDL 328-4 FAST F0AM
-----
Mixing method
Density (Ib/in ^
3.9 _c
Pigmented
4.1 4.3 4.3
Unpigmented 4.3
4-89
TABLE IV TABLEV
Selected C olorant (R6-or90I4), s howing B est C ompatibility Mine Marking Foam Used Successfully by S pecial Forces
AND leasts ETTLING. OTHER COLORANTS EVALUATED ARE SHOWN. Personnel ON Various Mines at Ft. A P. Hill
Colorant type Material designation Manufacturer Mine type Number of mines tested with foam
Pigment R6-OR9014 Magruder Color Co. (Radiant M-16 8
Div.)
PMD-6 4
Pigment P7-OR0624 Magruder Color Co.
VsMk2 6
Pigment TLOR6714 Magruder Color Co.
Vs-50 3
Pigment GF-OROOI4 Magruder Color Co.
Ts-50 1
Pigment P7-OG0623 Magruder Color Co.
M-I4 3
Pigment P7-OR9013 Magruder Color Co.
PMN 1
Dye Reactint x52 Milliken
M-3 1
Dye Reactint x38 Milliken
Valmira 69 1
TMD44 1
Color stability was tested by exposing cured foam samples
to bright summertime Southern California sunlight for
6 weeks. At the end of that time period, the foam had turned Ambient temperature during the demonstration was
a darker reddish-brown color. This color, combined with the between 30 and 50°F. Foam cartridges were stored overnight
foam shape, was still readily distinguishable in the field. at temperatures between 20 and 30°F prior to the demonstra¬
Field tests at Ft. A.P. Hill under overcast conditions showed tion. The cartridges stayed quite cold because the plastic
no color change after 3 weeks of outdoor exposure. It was cartridges and the foil pouch containers insulated the foam
noted that UV fluorescence was no longer apparent after chemicals. Under these conditions, the following observa¬
1 week’s exposure to sunlight. tions and conclusions were made:
D. Wet Surface Test Results 1) The foam was successfully mixed and dispensed from
Tests of the selected foam in wet conditions were per¬ the gun during periods ranging from 1 minute 50 seconds to
formed to examine the effects of rain and high humidity on 3 minutes 45 seconds, and began to rise from 1 minute
foam application. 30 seconds to 3 minutes thereafter (Fig. 3). The foam was
hard after 4 to 11 minutes total time following dispensing. It
1) Spray mist test: Foam was dispensed while spray mist is difficult to extrude foam chemicals using the dispensing
from a water bottle simulated rain. No adverse effects on gun when temperatures are below 50°F.
foaming action, hardening, or curing were noted.
4-90
2) The foam adhered well to most metal and plastic 3) The foam color was excellent; the brilliant day-glo
surfaces, even when the surfaces were cold, wet, and dirty orange-red (Fig. 6) was easily visible from over a 100 yards
(Fig. 4). While the foam did stick to some wooden box away. It was demonstrated that the foam fluoresced in a
mines (Fig. 5), it appeared that the amount of mine surface darkened room when exposed to UV light.
exposed prior to applying the foam was more critical for this
type of mine. (The foam pulled away from one of four PMD-
6 and one large anti-tank mine when attempts were made to
pull these mines from hard, wet earth.) Mines having
exposed tripwire mechanisms required minimal unearthing.
Tripwire mechanisms provide a good “handle” for the foam
to grab onto. They prevent the mechanism from setting off
the fuse and allow the mine to be pulled from the ground by a
length of rope embedded in the foam which acts as a lanyard.
While it would never be done in practice, it was shown that
tripwires could be used to pull out form-encapsulated M-16
mines buried in the ground without tripping the mine’s fuse.
I ' W:
r-
’■''a"**”*‘iti
4-92
Fig. 13. Only colored residue and rope remain following mine destruction.
4-93
method provided the necessary visual cues, and the robot
operator said it was an excellent aid in guiding him to dig up
and capture the mine properly (Fig. 17).
Fig. 15. Degree of adhesion of cured foam to mine varies with that portion
of the mine encapsulated.
References
4-94
The Development of a Multimedia Electronic Performance Support System for
Humanitarian Demining for the Proceedings of Technology and the Mine Problem
Symposium
William C. Schneck
Humanitarian Demining Project
Night Vision Electronic Sensor Directorate
Ft. Belvoir, Virginia 22060-5608 USA
Amos L. Samuel
ESSEX Corporation
1430 Springhill Road, Suite 510
McLean, Virginia 22102 USA
ssamuel@essexcorp,com
Abstract - Any effective humanitarian demining training and these initiatives is the Demining Support System (DSS). The
operational support system must provide on-demand access purpose of this paper is to discuss the rationale behind the
to all the resources that are needed to perform a task, solve development of the system, to discuss design considerations, and
a problem, train and inform indigenous personnel and to provide an overview of the systems content.
support the overall management of the demining program.
REQUIREMENTS
The use of a multimedia electronic performance support
system (designed for in-country field use) can be an
As part of the humanitarian demining program there are
extremely cost effective tool in being able to generate
provisions to leave behind developed solutions for continued in-
performance and training support at the moment of need.
countiy use. The Office of the Deputy Assistant Secretary of
Today’s technologies allow for compact storage of vast
Defense for Humanitarian and Refugee Affairs determined the
amounts of information, full motion video, graphics and
most successfiil program in terms of cost feasibility is one that
multi-language audio, all available in a variety of output
trains to promote an indigenous capability in countries affected by
formats, rugged enough for use in the field. Properly
landmines so that they can eventually solve their own landmine
designed and packaged, these systems can significantly
problems. The skills that must be learned include: minefield
improve the efficiency and safety of humanitarian demining
reconnaissance, locating mines, reporting mines, mapping
operations.
minefields, destroying mines, managing the mine-removal
INTRODUCTION operation, launching a mine-awareness campaign, addressing the
consequences of landmines on public health, implementing first
Currently there are an estimated 110 million landmines aid and follow-up treatment, and addressing the psychological
scattered throughout 64 countries. Landmines maim or kill an impact of mine-related injuries.
estimated 500 people per week [1 ]. The United Nations projects
DEFINITION OF ELECTRONIC PERFORMANCE SUPPORT SYSTEM
that if the use of landmines were stopped immediately it would
take 1,100 years and $33 billion dollars, at the current rate to
To accomplish these diverse tasks and meet the needs of the
clear, those already in place [2].
trainer, a different performance technology and media approach
The statistics associated with reported landmine casualties are
are required. The technique capable of supporting the
staggering. The landmine problem has resulted in arrested
requirements is the Electronic Performance Support System
economic development that, if not effectively mitigated, will
(EPSS). The goal of an EPSS is to provide whatever is necessary
result in continued economic devastation and migration to
to generate performance and learning at the moment of need [3].
neighboring countries with already fragile infrastructures.
This can also be referred to as on~demand, just-enough, or just-
Currently there is a major international thrust to develop
demining equipment and techniques capable of augmenting the in-time training.
There are several working definitions for an EPSS that range
demining effort. To support this effort, the Night Vision and
from “...the electronic infrastructure that captures, stores and
Electronic Sensors Directorate at Ft. Belvoir has developed over
distributes individual and corporate knowledge assets throughout
30 items to assist in the mitigation of the landmine crisis. One of
4-95
the organization, to enable individuals to achieve required levels simultaneously provides a synergistic approach
of performance in the fastest possible time and with a minimum of required to support multiple training options.
support from other people.” [4] to “...[being] universally and
consistently available on demand any time, any place, and Supportable - the use and support of technology must
regardless of situation, without unnecessaiy intermediaries not hinder mission deployment.
involved in the process.” [3, p. 34] For our purposes, we will
define EPSS as the integration of available technologies (human To accomplish these tasks requires that a system possess
performance and computer) to facilitate the accomplishment of a certain attributes. First, it must have a large and flexible data
desired outcome. In this case, the desired outcome is for countries storage capability and play multiple CD-ROMs. Second, the
affected by landmines to conduct successful demining operations. ability to use full motion video is needed. Video is the only media
that permits the introduction of training and informational
WHY MULTIMEDIA?
materials in aural, verbal, visual, and kinesthetic imagery [5].
This has direct implications for the selection of informational
An immediate reaction to the system regarding a multimedia-
materials to be included on the system and instructional design
based platform is normally, “Multimedia for demining? The
considerations. Third, it must have the ability to incorporate
people who developed this can’t possibly be in touch with reality.
hardware into the system that permits modifying and adding
Demining is low tech and requires a low tech approach.”
materials to the base system instructional materials. This permits
Based upon an analysis of the ARSOF training requirements,
localized customization of training and mine awareness materials.
target audience (deploying trainers to host nation deminer), and
Fourth, it must be expandable without changing the base
currently utilized training methodologies, multimedia was selected components.
as the platform to deliver demining training materials. The
A multimedia system can support these characteristic by
utilization of a multimedia platform provides the trainer the ability
providing flexible software and hardware configurations that
to give customized and just-in-time training for diverse target
facilitate the integration of voice, sound, image, and motion. The
audiences and locations.
enhanced technology of multimedia enables presentations to take
The design of the DSS was derived from the data obtained
on a more realistic appearance than systems Aat provide only still
through interviews of mission planners, trainers, and medics, and
images without voice, sound, and motion [6].
a review of field manuals and after action reports. The analysis
yielded the following desired system characteristics. MODULE DEVELOPMENT
Visual based - training based on visuals addresses the An instructional systems design (ISD) methodology was used
literacy level of the indigenous populations. to develop the content of the systems module. The ISD process
consists of five phases: analysis, design, development,
Language support - audio narrations provide an implementation, and analysis. The analysis methodology used
immediate, consistent approach to the transfer three distinct phases. First, relevant literature was examined to
of information in the native language of the identify mission planning formats and programs of instruction
trainee. (POI). Second, we interviewed ARSOF subject-matter experts
who have experience in mission planning, demining operations,
Content based - preselected and organized visual and mine awareness and medical procedures. The interview formats
audio materials provide a baseline for training were designed to elicit experience and knowledge through the
in demining techniques, medical procedures, retelling of specific incidents to provide an idea of how the
and mine awareness. mission specialist performs the tasks. Third, we identified the
POIs employed by Special Operation personnel on missions
Tool based - an option to modify or add demining, similar to what can be expected to be found during demining
medical or mine awareness material addresses operations.
the unique circumstances encountered in the The analysis results were presented in a survey format to
field. subject-matter experts at Ft. Bragg and Ft. Campbell. The survey
participants reviewed the list, verifying the topics suitability and
Flexible- The user is in the best position to decide what identifying their importance to a humanitarian demining mission.
form of output (poster, printout, video, The result of this analysis determined that the systems modules
presentation, audio) most suits the situation. A should consist of demining techniques, medical training, and
program that supports all five media options mine- awareness.
4-96
SYSTEMS FUNCTIONALITY AND EQUIPMENT Mission Planning & Management Guide
Provides immediate access to materials required for the
The DSS design does not preclude a low-tech implementation planning and execution of a demining mission to
of training materials. It enhances the instructional capability of include applicable references, equipment manuals (text
the ARSOF trainer by providing the ability to present instruction and electronic manuals), forms and formats for reports
using different media, depending on the situation and student and briefing.
audience. Although the system is capable of delivering
instructional materials in audio and video modes, the primary Training
delivery media is print. The courses in this module provide advanced training
The system can display, print and edit lesson plans, field for individual skills topics from selected demining
evaluation cards, graphics for heat press cloth transfers, handouts, sources. The current available courses are the Combat
labels, and posters, in support of training and mine awareness Life Savers Course (CLS), Demining Course, and
activities. The inclusion of a scanner and digital camera greatly Communication Course.
enhances the capabilities of the system by providing a means for
the SF trainer to rapidly and easily develop complex training Medical
materials produced for a specific purpose which was not foreseen Provides instruction on the treatment of landmine related
during predeployment planning. injuries. The module contains information about Buddy
Aid, provides a copy of the current Combat Life Savers
Composition of the Demining Support System Course, and current literature on topics related to
landmines.
17" Mitsubishi rackmount, touch screen monitor
Mine Awareness
Speakers/Amplifier - Fostex 630LB Mine awareness information printable on items such as
scarves, T-shirts, ponchos, tote bags, or mine awareness
Fieldworks 766P Laptop - Pentium 166, 4X CD-ROM, 32 posters. Cultural considerations have been included in
MB RAM, 1 gig hard drive the development of mine awareness materials that may
be distributed as part of an in-country mine-awareness
CD-ROM Jukebox (7 Disk) program.
4-97
training materials. The DSS is modular and transportable, which [4] B. Raybould, “Performance support engineering: An
permits rapid worldwide deployment to support the demining emerging development methodology for enabling
mission. organizational learning,”
http ://www. cct .fsu. edu/sv2000/PIQ/Raybould. html,
REFERENCES October 1996.
4-98
LEXFOAM FOR HUMANITARIAN DEMINING
C. John Anderson
Mining Resource Engineering Limited
1555 Sydenham Rd, R.R. #8
Kingston, Ontario K7L-4V4
Canada
and
Joseph L. Trocino
Golden West Products International
15233 Ventura Blvd., P-8
Sherman Oaks, CA 91403
USA
4-99
C.^Landmine Countermeasures - Neutralization foam from a liquid explosive, or from a liquid wliich would
have explosive characteristics in foam form.
Most techniques in humanitarian demining arc borrowed Several researchers have investigated the use of explosix'c-
from equipment, materials, and removal procedures established impregnated polyurethane foams [3-8]. Tlie detonation
by militaty doctrine intended primarily to ‘'breaclv’ minefields properties reported for each type of foam included delonability
for or during combat. A Quality Assurance clearance level of limits, and (in some cases) thcorclical/expcrimenlal detonation
95% is generally acceptable in military operations, however, velocities and pressures as a function of density. Howex^er. in
this level approaches 100% for humanitarian demining. at least one case, the investigators found that the viscosity of
The countermine materials and equipment used for the urethanc/e.xplosive mix precluded efficient dispersal of the
militaiy breaching, and at times in humanitarian demining, foam as a mine neutralization technique [8],
include mechanical plows. Hails, rollers, line charges, and solid Pool inxestigated a foamed liquid explosixe invoh ing a
cxplosix'c charges such as C4 and TNT. Each of Ihcsc methods mixlurc of nitromethane and metal stearate surfactants |9|.
has its drawbacks for humanitarian demining, including This composition was whipped into a semi-slabic foam willi a
effectix eness and/or cost. foam density of 0.5 g/cc. Ii was reported that the foam
In humanitarian demining, it is generalIv accepted that drainagecharaclerislics were railicr j)oor. .As well. Alford has
upon detection of the landmine, ‘'bIo\\-in-placc“ techniques described a liquid foamed cxplosix c, produced using aerosol
xvill be used to neutralize tlie tlireat. This is generally done with teclmology [10]. The foam xvas based on an aqueous solution
liigh explosives (C-4, TNT block) or by directed energy (shaped containing inorganic nitrates and PETN. It should be noted
charge attack). The use of these techniques also has its that this composition, due to the PETN content, w^as classified
drawbacks, including: the costs associated with the logistics of as an c.xplosivc (UN LID). This foam w'as knowm by the trade
handling high explosives; the possible safely factors involved names FOAMEX or PRIMAFOAM (depending on the countiy'
with placing explosive directly on an exposed landmine (no of manufacture/salc).
standoH): and tlie dangers associated with sccurilx concerns in The preceding (dclrimcnta!) factors hax'C resulted in the
many of the third world countries, sucli as theft by terrorist development of a liquid explosixc foam - LEXFOAML This
organizations. The following sections present information nitroparaffin-based foam is based on the use of aerosol
regarding the use of a novel distributed cxplosix c technology technology and emulsion science. The cticrgclic componeni
based on tlie use of a liquid explosive foam - LEXFOAM*. for comprises approximately 90% of the foam material.
use in humanitarian demining. The use of a nitroparaffin-based explosix c foam offers
adxantages not found with other explosive technologies
II. Tin: Li:\f().am Soi.rnox including the adx antagcs of safety and cost. LEXFOAM Stock
Solution (from xvhich the foam is produced) is classified as a
A. History of LEXFOAM flammable liquid - Class 3, UN 1261. This offers significant
logistics cost savings, as the precautions necessary for the
The detonation properties of solid high density explosives shipping and storage of high e.xplosives are eliminated. As
(>1.0 g/cc) have been extensively investigated. These t\^cs of well, the BACKPACK and PALLETIZED foam dispersal
c.xplosives have detonation pressures of several hundred kbar systems arc designed to be easily used by trained personnel in
and detonation velocities up to 9 km/s. At the other end of the field applications: the training can be accomplished in a
spectrum are fuel-air (or o.xwgcn) explosives wliosc detonation minimal amount of time. Finally. LEXFOAM has consistently
pressures are less than 20 - 50 bar with detonation velocities neutralized a wide x'arietx' of mines in a \aricty of test and
less than 2 km/s. E.xplosix'es systems cox’cring the entire range ex’aluation situations, with success Icx'cls consistently
betxveen tliese two extremes are theoretically possible, however approaching 100%. The following sections outline: the
only a few such systems have been studied experimentally. detonation properties of LEXFOAM; the countermine
The most straightforward method of varying the successes and ongoing database development associated with
detonation pressure and velocity is to vaty the density of high this technology; a description of the dispersal systems designed
explosive loading. Tulis has shown that clouds of high to facilitate the use of the foam in humanitarian demining
explosive dusts dispersed into air can produce detonation applications: and a discussion of LEXFOAM countermine
pressures (100 bar) well in excess of those produced in standard techniques.
fuel-air configurations f2|. The dispersal of explosixc dust
clouds witli uniform cloud consistency, howcxcr. is fraught B. Lh'Xh'OAM Delonalion Bropernes
with experimental difficulties.
A potentially more effective method of producing This section summarizes (lie results of an inx estigation to
explosive systems with low loading densities and uniform experimentally determine the detonation pressures as a
detonation properties is to disperse the explosix'e in a porous function of foam density, and to correlate these results with
foam matrix. An alternative method would be to produce a
4-100
measured in-ground pressures and impulses as a function of iii) similar in-ground pressures for different foam densities
foam density and foam layer thickness. can be attained by adjusting tlie thickness of the foam
The velocity of detonation (VOD) of the LEXFOAM was layer. For example, a 5 cm tliick layer of foam, having a
determined using either the continuous resistance wire foam density of 0.25 g/cm^ will generate similar in-
technique, or piezoelectric pin (point-to-point) techniques. groimd pressures (at gauge burial depths of 20-30 cm) as
Piezoresistive (carbon) and piezoelectric (Polyvinylidene a 2.5 cm thick layer of foam having a foam density of 0.5
difluoride-PVDF) gauge elements were used to determine the g/cm^.
detonation pressures for both incident (sweeping) and
transmitted shock loadings. In-ground pressures and impulses It is clear that 5 cm thick layers of LEXFOAM, for a
were measured using either modified Kulite gauges and/or given foam density, result in higher pressures at increased
column based stress (CBS) gauges developed by Waterways gauge burial depths than 2.5 cm thick layers of foam. This is
Experiment Station (WES), in conjunction with “flat-pack” due to the increased impulse associated with the thicker layer
PVDF gauges developed by DYNASEN f 11-13]. of foam. Duvall has obserr'ed this behavior [14]. Measured in-
Incident and transmitted detonation pressures were ground impulses at different gauge burial depths remain
measured for foam layer thicknesses of 2.5 and 5 cm, with relatively constant for a 5 cm thick layer of foam. A 2.5 cm
associated foam densities of 0,2 and 0.4 g/cm^. As well, in- thick layer of foam exhibits decreased impulses at the deeper
ground pressures and impulses were determined for these foam gauge burial depths.
configurations. In-ground pressures were also determined for
2.5 cm and 5 cm thick layers of LEXFOAM having foam C. LEXFOAM Countermine Successes
densities of 0.25 and 0.5 g/cm^.
Figure 1 illustrates the measured detonation velocity as a LEXFOAM has consistently demonstrated that it is an
function of foam density. Figure 2 shows measured incident easily used, versatile, sprayable foam explosive material able to
and transmitted detonation pressures, and compares these neutralize anti-personnel (AP) mines, anti-tank (AT) mines,
pressures with calculated pressures from three different and unexploded ordnance. The method of neutralization
equations used to approximate detonation pressures. It is clear involves either: sympathetic detonation of the main explosive
that the relationship P = poDV4, where is the initial foam charge or fusing mechanism; function of the fuse and mine
density and D is the measured VOD, most closely approximates detonation; or mechanical destruction of the mine/fusing
the detonation pressures for the foam system, particularly at the mechanism. Using these criteria for success, LEXFOAM has
lower foam densities. Furthermore, it is apparent that the proven 100% effective as a mine neutralization technology in
transmitted detonation pressures are of the order of twice the several series of field trials for both commercial and
incident detonation pressures. Duvall has noted that, government clients. Table I outlines a list of land mines and
depending on die target material, this phenomenon can readily unexploded ordnance successfully destroyed in unclassified
occur for shock waves impacting normal to a target [141. This government (U.S. Army Night Vision Directorate -
increase in transmitted detonation pressures has ramifications Humanitarian Demining Program) trials and during
for mine neutralization; spccificall>' that the LEXFOAM should commercial field trials in Kuwait following the Gulf War. It
be located and initialed such that the detonation \va\'C impacts should be noted that the database of mines neutralized by
the mine with a transmitted shock for optimum effect. LE.XFOAM is continuously being updated.
It should be noted that there was minimal difference in
measui-ed incident detonation pressures for both 2.5 cm and 5 III. LEXFOAM Delivery Systems
cm thick layers of LEXFOAM, and that all transmitted
detonation pressures were measured using 5 cm thick layers of In a typical mine-clearing operation, LEXFOAM is
foam at the associated pressure gauges. deployed Uuough a hand-held spray gun. The spray gun comes
Figures 3 and 4 outline the measured in-ground pressures with a 2-foot long detonation trap assembly and two additional
and impulses, for various foam layer thicknesses and foam 2-foot long quick-disconnect extensions. This allows the user
densities. Each datum represents an average of 3-5 the option of selecting dispensing gun lengths of 2, 4 or 6 feet.
measurements. The followung trends are apparent; In most instances, rvhere a mine has been located, identified
and checked for trip rvires, a "close approach" has already been
i) for a given foam density, increasing the foam layer made and a short dispensing gun assembly may be acceptable.
thickness increases the in-ground pressures at a given In addition, LEXFOAM can be sprayed in large patches
gauge burial depth; or in more defined patches over individual mines. A number
of mines can then be explosively linked to each other by thin
ii) for a given foam layer thickness, increasing the foam strips of LEXFOAM, or with detonating cord, allowing all
density increases the in-ground pressure measured at a linked mines to be destroyed using a single detonator.
gix en gauge burial depth;
4-101
Two separate, yet complementaiy delivery systems have The Palletized Delivery^ System, wdth associated trailer, is
been developed for the dispersal of LEXFOAM. The systems designed to produce four 200 kg batches of LEXFOAM before
are designed to be loaded in the field with LEXFOAM STOCK the components need to be restocked. In addition, the palletized
SOLUTION: this solution is classified as a flammable liquid. uni! is also designed to load the previously described
Class 3, UN I26L and therefore offers considerable ad\'antagcs backpacks, three of which can be included w ith ilie palletized
regarding safety, shipping, and storage considerations. It unit as part of the complete LEXFOAM deli\ciy s\siem
follows that the stock solution is an unlikcl\‘ candidate for theft technolog}'. It should be noted that both the Backpack and
and terrorist use. In addition. LEXFOAM is en\ ironmentally Palletized Units can be immediately refilled and used to mix
friendly in that it is relatively non-toxic, biodegradable, and and disperse more foam, or tliey can be cleaned with ^vater and
easily disposed of by burning or washing away with water. stored for later use.
The two methods of foam dispersal utilize backpack and
palletized delivery^ systems. Details of these systems are IV. LEXFOAK4 Countermine Techniques
outlined in the following sections.
A. General Considerations
A. Backpack Delivery System
Based on experimental data, countermine successes and
The backpack s}^stem, when loaded, weighs field experience, a variety of different techniques have been
appro.ximately six1y pounds (27 kg). The backpack is ideal for developed to successflilly neutralize land mines. In most cases,
spot coverage, for small jobs and for reaching mines or the optimum foam configuration comprises a 5 cm (2") thick
ordnance in locations that are difficult to reach accurately with layer of foam, wdtli a foam density of 0.5 g/cnr\ In conjunction
the larger palletized system. The backpack is designed for with these parameters, the experimental data have
operation by low skill-level indigenous personnel with minimal demonstrated the importance of the following criteria;
training. In brief LEXFOAM STOCK solution is pumped into
the backpack, followed by addilion of a metered amount of i) If the mine is exposed, a layer of foam should be placed
liquid propane using pressurized nitrogen as the dri\ ing gas. against the mine, with the detonator inserted at the
The ^\^o components are mixed by inverting the backpack opposite end of the layer. This ensures proper ‘Tun-up’' to
se^'cral times, after \^Tich the delivery s>stcm is ready for use. full detonation and subsequent impingement of a
Figure 5 illustrates the backpack components and a fully transmitted shock wa\c for maximum pressure transfer.
assembled backpack.
The e.\plosi^’c foam can be sprayed directly on any ii) If the mine is buried, tlie foam la>cr should be dispersed
c.xpcnded ordnance to be neutralized. As an alternative, it can on tlic ground such that one end of the layer coxers tlic
be applied directly on the ground over a known or suspected suspect mine and Initiation is implcmcnlcd from the
mine. The total elapsed time for filling the s>stcm and mixing, opposite end of the layer. Again, this insures full run-up
dispersing and detonating 15 kilograms of foam can be as little and maximum detonation pressure imparted to the
as lO minutes. ground cover.
B. Palletized Delivery System Once the foam is deployed, a blasting cap or a detonation
charge can be placed at the appropriate location in the foam
The palletized delivery^ system is a 60-gallon (227 liter) and triggered to detonate the entire foam lax^cr. Detonation of
vessel mounted on a steel skid for case of transportation b}' a the foam will reliably induce sympathetic detonation or
small trailer. 3/4-ton pick-up truck or other similar \chiclc. destruction of surface ordnance and exposed mines. Buried
The system includes: prc-mcasurcd containers of ingredients: mines which arc fused and armed arc reliablx functioned,
a pumping system for transfer of the stock solution from a 55 including mines buried at depths down to lO inches.
gallon drum to the stainless steel pressure \csscl lank: an (Depending on the fuse l\pe. mines ha\c been initiated at c\cn
agitator to mix the ingredients in the tank; nitrogen for greater burial depths.)
pressurizing the system; and a hose, trigger and nozzle system It should be noted that most demining practices involve
including a detonation trap for safe dispersal of the pressurized partial e.xposure of the mine for threat identification purposes.
foam explosi\'C. The system also includes a po\^’er source, Any mines, ordnance items or other explosive devices
controls and \’arious safety systems and features. The system detonated by the foam may create craters in the ground and/or
is designed for operation by low skill-level indigenous produce shrapnel and other potentially dangerous side effects.
personnel with minimal training. The loading procedures are Therefore, practical safety^ precautions, including safe stand-off
similar to those employed with the backpack system, albeit on distances and personnel protection must be taken when
a larger scale. Figure 6 illustrates the Palletized Deli\ery detonating the foam.
System, with attached trailer and added Backpack units.
4-102
A dispersed layer of foam, in the absence of other be to attack the PROM-1 using LEXFOAM. This situation is
explosive devices, will not create craters and will generally illustrated in Figure 7(C). Only the shoulder of the mine must
have a trivial effect on the ground surface. Typical effects are be e.xposed. reducing the risk of accidental initiation. The foam
limited to an approximately 5 cm (2") compression of the soil follows the contours of the shoulder to achieve intimate
surface. The amount of compression which occurs depends on contact. The detonator can be placed to give the optimum
the soil composition and moisture content. The compression direction of initiation.
effect on paved surfaces such as roadways or ninways of the The PMR-2A fragmentation stake mine is generally
order of fifty percent less than that induced in soil. mounted on a stake above the ground to optimize the mine's
In addition, if it is decided not to detonate the foam for range and effectiveness. Initiation is almost alwa}'S by tripwire.
any reason, it can be neutralized with water or left to The main explosive charge is housed in the top tw^o thirds of a
decompose into harmless and environmentally friendly by¬ thick steel body. To achieve successfiil in-place destruction, a
products. This process will occur naturally over a period of conventional demolition charge must be placed adjacent to the
hours, depending on ambient temperature and moisture mine body. Trying to position a charge off the ground and
conditions. against the mine is clumsy, time-consuming and dangerous. In
addition, acliieving close contact of the demolition charge with
B. Specific Examples the rounded mine body is also difficult, increasing the chance
of only partial destniction. With LEXFOAM, there is no need
Although the preceding criteria for use of LEXFOAM to build a stnicture to support the demolition charge at the
should generally be adhered to, as with all field applications correct height. The foam adheres to the mine body, filling the
actual demining scenerios are developed based on the grooves and contouring around the body to achieve intimate
experience of the deminer. This section outlines several contact. A trail of the foam is then dispersed down the slake
e.xamples of the use of LEXFOAM against selected AP mines, and on the ground to a convenient and safe initiation point,
some of which are notoriously difficult to neutralize in-situ. well away from the tripwire. The mine, and the explosive
Most mine clearance organizations now insist on in-placc neutralization configuration, arc illustrated in Figure
destniction of land mines to simplify training and create clear, Directional fragmentation mines of the Claymore t>qoe
unambiguous drills. However, successful in-place destruction are almost always mounted above the ground to maximize the
is not as simple as it may seem, and there are many situations fragmentation range. Although generally supplied with legs
were conventional explosives (such as TNT and C-4) meet their mounts, the mines are often mounted (depending on the
limitations. The unique properties of LEXFOAM offer practical terrain) on trees up to on meter above the ground. The mine is
solutions to these problems. Tlie following examples are chosen usually initiated by either trip wire or remote command.
to represent circumstances which generally occur in minefields Positioning a demolition charge adjacent to the mine is
throughout the world. awkward, while working close to the tripwire represents a
The PROM-1 bounding fragmentation mine, illustrated constant hazard. As demonstrated in Figure 9. LEXFOAM can
in Figure 7(A), is designed to be buried with only the pronged be dispersed against any surface of the mine body and a trail
fiize above the ground. It can be initiated either by pressure or run to a convenient and safe initiation point. If a trail of foam
tripwire, and utilizes a small charge to deploy to a height of is neither possible nor desirable, a patch of foam may be
approximately 0.7 meters, followed by detonation of the main dispersed against the top of the mine. In either case, detonation
charge. The side w^all is made from thick steel to enhance of the LEXFOAM wall result in sympathetic detonation of the
fragmentation, but the thinnest and most Milnerabic part of the mine.
casing, and therefore the ideal place for explosive attack, is the The PMA-3 pressure operated blast mine represents a
rounded shoulder. Attacking the PROM-1 using a demolition class of plastic mines which arc difficult to delect and arc
block of high explosive presents problems in that the block designed to be blast-resistant. This mine is show n in Figure
cannot easily be placed on the shoulder of the mine due to the 10(A). The main e.xplosive charge is housed in a small ca\'ity
proximity of any tripwires. This situation is shown in Figure at the top center of the mine body, and is surrounded by air-
7(B). As well, the block docs not have good contact with the gaps and a resilient plastic casing as illustrated in Figure
rounded shoulder of the mine. This results in the block having 10(B), A demolition charge placed beside the mine may not be
to be placed against the body of the mine, which results in able to detonate the explosive, wdth the result that the mine
having to clear dirt from the side of the mine, increasing the could be in a more hazardous condition. Placing explosive
risk of activating an anti-disturbance device. The position of blocks on top of a pressure operated mine is wadely regarded an
the detonator will result in a incident shock wave transmitted unacceptable, for obvious reasons. Figure 10(C) show^s how
to the mine, a situation which is not ideal from imparting LEXFOAM can be applied directly above the main charge on
significant pressure through the thick steel wall casing. the top center of the mine. The negligible w^eight of the foam
A more satisfactory, and less dangerous technique would means that only the top surface of the mine be uncovered.
4-103
Initiation of tlie foam destroys the mine by either fiise function Initiation by Gaseous Detonation Waves”, National
or sympathetic detonation. Research Coimcil of Canada, DME Mech. Eng. Report
Figure 11 illustrates the concept of using LEXFOAM to MT-60, Febaiary 1968.
neutralize more than one mine at a time, particularly when, as
is often the case, AP mines may be found close together [4] Tulis, A.J., Austing, J.L., Baker, D.E., “Open Matrix
(clusters). If the mines are very’ close together, then the entire Very' Low Density Explosive Formulations”, J.
area can be blanketed with a 5 cm (2") thick layer of foam. An Hazardous Materials. Vol. 5. 1982. p.387.
alternative would be to connect a number of patches of foam
with either a trail of foam or with detonating cord. |5| Austing. J.L.. Tulis. A.J,. Johnson. C.D.. '’Detonation
It is apparent that the use of LEXFOAM as a mine Characteristics of Very Low Density Explosive
neutralization technique is limited only by the deminer’s Systems”, in Fifth Symposium (International) on
experience and ingenuity. This factor, coupled with the clearly Detonation, ACR-184, ONR, 1970, p.47.
defined advantages of safety, cost and logistics, demonstrates
the viability of LEXTOAM technology for humanitarian [6] Austing, J.L., Tulis, A.J., “Further Studies on the
demining operations. Detonation Characteristics of Very' Low Density
Explosive Systems”, in Sixth Symposium (International)
V. Summary on Detonation, 1976, p.l83
LEXFOAM has been tested by tlie U.S. Army and in field [7] Xueguo, L., “Detonation Characteristics of Polyurethane
tests in Kuwait. These tests have demonstrated that Foamed Explosives”, in Proceedings of the
LEXFOAM is an easily used, versatile, foam explosive International Symposium on Inten.se Dynamic Loading
materia], offering stand-off capability (sprayable), with a >99% and Its Effect.s'\'^Q\]\vig. China 1986. p. 173.
field-proven ability to sympathetically detonate buried, surface
and above ground anti-personnel (AP) mines, anti-tank (AT) [8] Anderson, J.. von Rosen. K., Gibb. A.W., Moen, I.O..
mines, and unexploded ordnance. As a result of LEXFOAM's “Detonation Properties of Explosive Foams”, in Ninth
100% success rate in destroying a variety of mines in the Sympo.sium (International) on Detonation.
No\'ember 1995 A.P. Hill, U.S. Army tests at Fort Belvoir, OCNR113291-7, 1989, p.l364.
Virginia, the U.S. Army's Humanitarian Demining Directorate
concluded (hat the LEXFOAM Snsiciu is ready for immediate [9] Pool. J.E. . U.S. Patent 2.967.099 (1961).
use, and has recommended operational dcplo}'mcnl.
LEXFOAM deliver}' systems arc safe, easy to use, cost flO] Alford, S.C., “E.xplosive Foams”, in Explo.sives
effective and a proficient tool for ordnance demolition. Engineer, 1985, p.76.
Moreover, safety and simplicity make this system particularly
suitable for use by indigenous operators during humanitarian [11] Joachim, C., Welsh, C., “Design and Field Experience
demining overseas. Finally, project managers can be assured with tlie WES 10 kbar Airblast and Soil Stress Gauge”,
that LEXFOAM, unlike conventional high explosives, does not in The Shock and Vibration Bulletin. Part II, Bulletin
lend itself to misuse by terrorists or anti-government groups. 55. Naval Research Laboratory^ Washington. 1985,
p.135.
VI. References
[12] Ingram, J., “Placement Effects on Ground Shock
[1] “Hidden Killers: The Global Landmine Crisis”, U.S. Instrumentation”, Miscellaneous Paper N70-7. U.S.
Dept. OfState 1994 Report to the U.S. Congress on The Army Engineer Watenvays Experiment Station,
Problem with Uncleared Landmines and the U.S. Vicksburg, MS (1970).
Strategyfor Demining and Landmine Control, Office of
Int. Security and Peacekeeping Operations (1994). [13] Charest, J., DYNASEN Inc, Private Communication.
[2] Tulis, A.J., “Anatytical and Experimental [14] Duvall, G.C., “Some Properties and Applications of
Characterization of Explosi^'es’^ in Proceedings of the Shock Waves”, in Response ofMetals to High Velocity
1Symposium on Explosives and Pvrotechnics, March Deformation, Interscience, (1961).
1984*p.l-L
4-104
Device Name Oriein Tvne of Mine or Ordnance
Type 72 China
“
TABLE 1: PARTIAL LIST OF MINES AND ORDNANCE SUCCESSFULLY NEUTRALIZED USING LEXFOAM.
4-105
FIGURE 1: LEXFOAM DETONATION VELOCITY AS A FUNCTION OF FOAM
DENSITY.
DENSITY (g/cc)
100 +
$
0
PRESSURE
(MPa) o A I
10
' I ■ I ■ I ■ I
10 15 20 25 30
GAGE BURIAL DEPTH (cm)
G 2 mm (Thick) DETASHEET
100 -1 A 2.5 cm (Thick) LEXFOAM, p = 0.2-0.25 g/cm3
V 5.0 cm LEXFOAM. p = 0.2-0.25 g/cm^
O 2.5 cm LEXFOAM. p = 0.4-0.5 g/cm^
4* 5.0 cm LEXFOAM, p = 0.4-0.5 g/cm3
+
+
+
O
V
<10
A V
IMPULSE J
(MPa-ms)^® i A O
O A
A
O
O
1 A-1-1-1-1-1-1-1-1-1-1-r
5 10 15 20 25 30
GAGE BURIAL DEPTH (cm)
4-107
FIGURE 8: PHOTOGRAPH OF THE PMR-2A FRAGMENTATION STAKE MINE AND
LEXFOAM CONFIGURATION FOR MINE NEUTRALIZATION.
FIGURE 10: PHOTOGRAPHS OF THE PIVIA-3 PRESSURE OPERATED BLAST MINE
(10A), SHOWING THE PROTECTED EXPLOSIVE CAVITY (10B) AND
DISPERSED LEXFOAM FOR MINE NEUTRALIZATION (IOC).
4-111
FIGURE 11: PHOTOGRAPH OF A TYPICAL LEXFOAM CONFIGURATION UTILIZED FOR
NEUTRALIZATION OF MINE CLUSTERS.
4-112
CHAPTERS: PROGRESS IN
AUTONOMOUS SYSTEMS
FOR MINE WARFARE
AUTONOMOUS VEHICLES
The genesis for this series of Symposia on Technology and the Mine Problem is the vision that
advances in the technologies of autonomous vehicles, sensors, power packs, navigation and control,
and "work packages" will lead to a revolution in how mine countermeasures/countermine operations
are carried out. It is toward the realization of this vision that we stated the objective of the
Symposium was to "change the world". We at the Naval Postgraduate School firmly believe that the
requisite technologies are nearly within grasp - a view shared with us by the late Chief of Naval
Operations, Admiral Jeremy K. "Mike" Boorda, USN.
Accomplishment of this vision will require continuity of effort. The goal is in sight. However,
much remains to be done in proof of concept and operational demonstration. If one imagines a time
line stretching from today's Navy on into the future, these revolutionary approaches will probably
impact squarely on the "Navy after next". Mike Boorda was the kind of naval visionary who was
comfortable with the idea of planning for the Navy after next even as he fought the budget wars to
maintain today's force structure.
In this Chapter we have assembled papers that provide a report on the progress toward
meeting this challenge. However, to use a metaphor from athletics, the address by Major Colin King,
RA (Ret), editor of Jane's new volume on landmines (see Chapter 3), the bar has been raised. Major
King drove home the point that the environments for land countermine operations are difficult. Few
of the vehicular approaches now under development will be able to operate in most of these land
environments.
At sea much the same has happened. ARPA and the Draper Laboratories have successfully
demonstrated the capability to carry out extended mine reconnaissance through water of complex
structure, but development of shallow water and very shallow water autonomous capabilities has been
delayed by budget cuts. (Reference here is to ONR's Autonomous Ocean Network.)
5-1
The quest for truly autonomous mine countermeasures capabilities is entering a crucial stage.
Funds for development and demonstration of components and subsystems that cannot be fielded for
another five to ten years are short. These same resources are claimed by others with more immediate
delivery dates. It is therefore essential that those individuals who serve in various advisory capacities
to both the Army and Navy remain aware of the state of development and potential of autonomous
systems. Unfortunately, these technologies are new and, in many cases, unfamiliar to many
experienced scientific and operational individuals.
It is entirely possible that the more benign operational environments that one encounters in
Humanitarian Demining operations will provide the first opportunities to distance humans from mine
removal activity.
The 1998 Symposium on Technology and the Mine Problem will once again present sessions
that provide status reports on progress toward obtaining autonomous systems for land countermine
and sea mine countermeasures.
5-2
The Basic UXO Gathering System (BUGS) Program for
Unexploded Ordnance Clearance and Minefield
Countermeasures, an Overview and Update
Christopher DeBolt and Christopher O’Donnell
ABSTRACT:
ESTRODUCnON:
The task of removing UneXploded Ordnance (UXO) by Explosive Ordnance Disposal (EOD)
technicians puts these personnel under great risk. The risks are associated with the new technologies us«i in
the submunitions or mines, such as anti-tamper features, and the factors that these objects have been subject
to weather and environmental conditions that could trigger detonation at any time. However, military
operations require that UXOs and mines be cleared from strategic areas. Also, practice ranges and other
lands that are contaminated by ordnance or mines must be cleared so they can be converted back to more
beneficial use. Not only is the cost of training personnel in locating, gathering or disposing the unexploded
ordnance enormous but this activity puts the EOD technician in great physical danger.
The main force behind building most robotics systems is to reduce the human presence in dangerous
task areas such as UXO clearance and de-mining. The difficulties of perftMming complex tasks in the real
world environment present a challenge for engineers in designing a fully autonomous system. Furthermore,
the cost of building a single inteUigent robot fully equipped with complex sensor capabilities is too high for
use in UXO gathering or mine detection because of the risk associated with equipment destruction. The
NAVEODTECHDIV goal is to develop a low cost, easy to use, and simple to maintain system to perform
the Explosive Ordnance Disposal (EOD) mission. The BUGS concept consist of a reconnaissance platform
with suitable sensors to detect and locate the submunition from a safe distance, and then using a low cost,
simple Basic UXO Gatherer (BUG) to perform the Pick Up and Cany Away (PUCA) or Blow In Place (BIP)
function. Once the desirable behavior of a single simple robot is obtained, the same architecture then can be
distributed to all other similar platforms to create a ^oup of robots to accomplish a practical and vital
mission. It is believed the BUGS system of cheap, simple robots, operating collectively to accomplish a
mission, will be faster, cheaper, and easier to build then a single hi^ cost intelligent robot.
5-3
EOD / MCM MISSIONS:
The primary mission considered for the BUGS system is the clearance of scatterable submunitions.
The goal is to have a system which can gather a large number of unexploded submunitions to one location
that can be disposed of at one time. This is the Pick Up Carry Away ^UCA) procedure that is currently
employed in some cases by EOD technicians. The act of disposing (with an explosive charge) a stockpile of
UXO items is much quicker than disposing of numerous individual items that are spread out over a large
area. The EOD technician will be safer since he will not have to traverse an area littered with UXO to place
numerous individual charges, and will not have to handle the UXO’s himself Proximity sensors, tripwires,
and the hke are being used more frequently in submunitions, making the EOD technicians task more
hazardous. Also, submunitions are so cheap to manufacture, large numbers are used at one time, and the low
cost is indicative of low reliability, leading to many duds that must be cleared. The small gatherer robots of
the BUGS system will perform the most hazardous tasks.
Another important mission that is considered for BUGS is mine countermeasures. For breaching a
minefield, the practice of deplo5dng large nets filled with explosives works well, but is not efficient nor does it
take advantage of the knowledge of the location of individual mines. This knowledge of locations is being
developed by several mine detection technology projects currently underway. The individual BUG vehicles
can covertly place neutralization charges over mine targets, and at the appropriate time, a single command
from a remote operator can initiate the charges, rendering safe a required portion of a minefield.
The BUGS system concept consists of a three phased approach. The first phase is to detect and
locate targets to be collected or neutralized. The second phase consists of either reacquiring and gathering
the targets, or placing neutralization charges on the targets. The final phase consists of actually neutralizing
the targets. The approach of using these three phases, using different assets, will accomplish the desired
missions.
The detect and locate target phase is first. A sophisticated sensor platform can be used to perform
this task, such as the USMC’s COBRA or the Army’s ASTAMIDS landmine detection and location systems
that are being developed. For targets that are not buried, such as submunitions in the UXO scenarios, a
human can perform this task. He can visually detect the targets and record locations with a GPS receiver, or
some other local positioning system. NAVEODTECHDIV has recently developed the Remote Controlled
Reconnaissance Monitor (RECORM), which is a teleoperated robot with a camera used by an EOD
technician to remotely survey a hazardous area for ordnance targets. For the UXO scenarios, RECORM is
being automated to autonomously perform an extensive area search, recognize targets optically, and record
the target locations.
The second phase is the reacquiring and collecting targets, or placing neutralization charges on the
targets. For the UXO mission, the approach is to gather the small submunition targets in a central location.
This collection of UXO’s can then be neutralized at one time, providing a savings of time and explosives for
the EOD techmcian as opposed to neutralizing each of numerous targets individually. For the MCM
mission, the approach is to place an anti-mine munition over each mine target for later simultaneous
detonation. The individual BUGs will be loaded with target location information gathered in the first phase.
They will then go to the proximity of the targets, individually, and reacquire the targets using low cost
sensors. These low cost sensors are being developed under a number of other programs, and the technologies
being developed will be inserted into this program. The vehicles to perform the gathering fimction or placing
munition function have to be autonomous and cheap. They must be autonomous to allow the use of
numerous vehicles by one operator, to realize time efficiencies. They must be cheap so that numerous
vehicles can be afforded, and a few can be replaced in the event that one may be inadvertantly be destroyed.
Small vehicles operating in an area littered with UXO’s or mines, and handling explosives, have some
5-4
probability of being destroyed.
The final phase is the neutrahzation of the targets. For the UXO mission, this would typicaUy
consist of an EOD technician placing an explosive charge on the collection of submunitions, and neutrahzing
the collection at one time. For the MCM mission, with the anti-mine munitions in place over the mine
targets, a command from a standard miUtary transmitter, such as the MK186, can simultaneously initiate the
munitions to neutrahze all of the mines. A study has been completed by Naval Surface Warfare Center,
Indian Head Division, that shows the feasibiUty of using a semiconductor bridge initiator on a scaled-up
version of an existing anti-mine munition can meet simultanaety of initiation and neutraliztion requirements.
PROGRAM APPROACH:
The program approach taken for this project is to concentrate on technologies that will enable
different system concepts to be demonstrated, with a strong emphasis on control system technology. For the
small, autonomous gatherers, several contracts were awarded so a variety of concepts could be demonstrated
and evaluated. In addition to these contracts, NAVEODTECHDIV developed its own system concept in-
house. Each of these concepts are being modeled by the Naval Postgraduate School based on inputs from the
developers and observed performance. These system models are being run against various UXO clearing and
minefield neutralization scenarios. Each of the concepts developed is being tested at NAVEODTECHDIV
against a test plan designed to exercise the control subsystems. This testing is further described later.
Information about general technology areas are provided by government agencies to each of the developers to
reduce the burden on them, and allow them to concentrate on the control systems. The Naval Postgraduate
School is providing modehng and simulation assistance, NAVEODTECHDIV is providing information about
sensors and detectors, and NCCOSC is providing information about navigation and control systems. A
technical system concept study, the test results, and the results of the modeling will be used together to
determine which concept will be pursued in the next phase of the project, which is a multiple vehicle system
demonstration.
For the target detection and location platform, we contracted with Lockheed to develop a package to
be integrated into the EOD RECORM vehicle for autonomous detection, location, and indentification of
UXO targets using visual means. The Robotic Work Packages developed for NAVEODTECHDIV for
autonomous survey of underwater ordnanace items is being adapted for terrestial use on RECORM. This
vehicle will work with the gatherers to demonstrate the feasibility of our system concept.
GATHERER CONCEPTS:
For the individual gatherer vehicles, several industry contractors have been developing initial
concept robots. For this initial phase of the project, the emphasis is on the control system that will lead to
expansion into a successful multiple vehicle system. A demonstration of these competing single robot
approaches to UXO clearance and/or minefield neutrahzation was performed in July at NAVEODTECHDIV.
The gatherer vehicle control system concept developers include Foster-Miller, ISX / IS Robotics, Draper
Laboratory, K^T, and NAVEODTECHDIV.
Foster-Miller: Foster-Miller is using a vehicle similar to their Lemmings, which is a battery powered,
tracked vehicle capable of traversing a wide variety of terrain. It is symmetrical, and is designed to flip over
and continue traveling if necessary. It is taller than the Lemmings, since an array of antennae is located on
the top and bottom for receiving homing signals. For the initial concept, beacons vnll be placed on the
targets to guide this vehicle. Beacons may also be used to identify waypoints. Radio frequency transmitted
from the beacon is used for the vehicle’s homing from long distances in to approximately six feet. The
vehicle recalculates the desired angle of travel, and turns to this angle, every six to ten seconds. Closer than
5-5
six feet, an array of light emitting diodes on the beacon is used for homing directly, without any time delay in
changing direction. A magnet is used to pick-up the metallic targets and carry the anti-mine munition.
The Foster-Mller control strategy is quite simple. It is designed to operate in an unstructured
environment, and maximizes the mobihty potential of the vehicle. Sensors are not used to detect obstacles,
and the vehicle is expected to traverse obstacles, or if it gets stuck, it will back up and turn, and continue
travehng. If the vehicle flips over, this is acceptable, since the vehicle concept is to operate even if it is upside
down, like the Lemmings vehicle. The vehicle’s goal is to reach a beacon. A heading towards the beacon is
recalculated every six to ten seconds, and the vehicle makes a turn to that heading and continues forward. In
the final configuration, once the beacon is reached, the vehicle will either perform a pick-up fimction or a
drop/place function, depending on the beacon and the mission (either PUCA or BIP). The vehicle would
then know which beacon should be approached next. The beacon homing navigation scheme could be
replaced in the future with some other type of navigation / location system.
ISX / IS Robotics: IS Robotics is using a variant of their Pebbles HI vehicle, which is a small, battery
powered, tracked vehicle. A DGPS system is used for location and navigation information. An operator
control unit, being developed by ISX, is used to interact with the vehicle. This control unit will serve as a
central data collection for the vehicle, a vehicle activity coordinator, and a control interface between vehicle
and the operator. A mission for the gatherer vehicle is communicated fi-om the control unit to the vehicle. A
joystick at the control unit can be used to tele-operate the vehicle when needed, and a series of “go to”
commands can be entered for the vehicle to communicate the target locations. An electro-magnet mounted to
an arm is used for a manipulator.
The ISX / IS Robotics control strategy is a supervised autonomy, dependent on sensor inputs.
Infi-ared sensors are used to detect obstacles, bumpers detect colhsions, and inclinometers measure the slope
and roughness of terrain. The individual robots are programmed with a behavior control paradigm.
Behavior control facilitates rapid reaction to environmental hazards and robust response to system failures.
Multiple independent behaviors compute commands for the robot’s actuators, based on sensor inputs, and an
arbitration module resolves conflicts. The robots monitor their own progress and alert an operator if there is
an anomaly. As mentioned above, the operator control unit will serve as a central data collection for the
vehicle, a vehicle activity coordinator, and a control interface between vehicle and the operator. Constant
communication is required between the individual robot and the control unit, transmitting location and status
information. A map is maintained of detected obstacles, clear areas, and targets. Constant communication is
also required between the robot and the DGPS base station for precise navigation. Dead reckoning is used as
a secondary navigation system, and can be used exclusively, though with decreased accuracy, if the DGPS
system is unavailable or inoperable.
Ultimately, the operator control unit will plan paths for the robots, based on known obstacles, or
locations of obstacles as they become known by other robots in the field and mapped. If an individual robot
gets itself stuck, and calls for help fi'om the operator, the operator can tele-operate the robot with a joystick,
watching a video display fi-om a camera mounted on the robot.
Draper Laboratory: Draper Laboratory is using an evolution of their MPTy-series of vehicles, which is a
6-wheel drive flexible frame micro-rover driven by battery powered motors. Draper’s local positioning
system is an optical-electrical one, with two tripod-mounted beacons that emit a rotating laser beam. Dead¬
reckoning is used as a backup navigation and positioning system. Draper also employs an operator station
which serves as an automated mission management host. A wide scoop on the front of the vehicle is used for
a manipulator.
The hierarchial control strategy for the Draper Lab concept is dependent on sensor inputs, too.
Sensors are used to detect obstacles and collect information about the world. This requires constant
communication of sensor data and vehicle location with a operator station to create a map. The vehicle has
two speeds. In the slow mode, it is collecting information fi-om the sensors, which are fully active, and the
5-6
operator station is creating a map, melding the sensor information with any other map information that may
be known a priori. In the fast mode, the vehicle is traveling along paths that have been identified by the
operator station as being clear of obstacles. The three critical activities of the operator station are mission
management and mission planning, maintenance of target and path maps, and human operator intervention
as needed. The operator station performs autonomous mission management and mission planning, and sends
four types of commands to the individual vehicles; waypoint-slow, waypoint-fast, collect UXO, and deposit
UXO. The waypoint-slow command is issued to a vehicle to proceed to a commanded location while
detecting and avoiding obstacles and hazards. The waypoint-fast command is issued to proceed to a
commanded location, but a clear corridor is known to exist between the current position of the vehicle and
the waypoint. The collect UXO command is issued when a vehicle locally detects a UXO, and the deposit
UXO command is issued when a vehicle, with a UXO onboard, reports that it has reached the ordnance
disposal area. The collect UXO and deposit UXO functions can be aided, if required, by an operator using a
camera onboard the vehicle. Five kinds of data are transmitted fi-om the individual vehicles to the operator
station. These are the announcement of completion of the current command, announcement of failure of
current command, regular updates of vehicle position, location of detected UXOs, and location of detected
obstacles/hazards.
The autonomous planning performed by the operator station consists of hierarchical planning, route
planning, and road-building. The hierarchical planning decomposes tasks in steps, or levels. At each level, a
set of tasks is decomposed into subtasks. The top level controller deals with mission goals, the middle level
deals with subgoals derived fi-om the mission goals, and the low level deals with commands to the vehicle.
However, the Draper’s plarming does more than simply break down big jobs into httle ones. At each level of
planning hierarchy, the value of accomplishing tasks is traded off against the cost of consuming resources.
An initial plan is repeatedly modified using heuristics in an attempt to generate a plan of maximum expected
utility. This simulation procedure is an iterative improvement scheme wherein, for each iteration cycle, the
heuristic search attempts to improve on the current solution. The steps of the mission planning cycle are
repeated over and over, until the time allotted to the planner for planning has been exhausted, or a point of
diminishing returns has been reached. Each level of the planning monitors the performance of the plan and
compares it to the performance expected when the plan was created to determine the plan s fitness. When
the value of the plan being executed falls below its potential value, the planning process begins anew, and
uses die results to modify the plan during execution. An outcome of this process is the identification of
“roads”, which are used over and over again as the vehicles traverse the field, gathering UXO targets or
placing anti-mine munitions.
I^T; K^T has designing a new legged, walking vehicle that is being investigated for the BUGS
program. This vehicle has eight legs, and is biologically inspired, applying concepts found in the
mechanisms of locomotion, manipulation, and neural control in biological creatures. K^T has teamed with
Case Western Reserve University for the control system. The first vehicle has just been built, and is now
becoming operational. This concept is of particular interest for situations where difficult terrains are
encountered. Walking vehicles are expected to have much better mobiUty than wheeled or tracked vehicles.
NAVEODTECHDIV: NAVEODTECHDIV designed and built a new vehicle for this program. It is a
small, battery powered, wheeled vehicle, and uses a DGPS system for navigation, with dead reckoning as a
backup system. As an autonomous system, it does not employ an operator’s control station.
The control methodology for this concept is a layered subsumption approach. There are three levels
of controllers used. The highest level controller maintains the overall mission goals, and receives and sends
data externally, such as DGPS data and remote operator commands, such as initial mission goals, including
target locations. The lowest level contains the subsumptive control modules in the sensor controller that
generate beha-viors in response to real-time sensor inputs, and a motion controller that controls the motors on
the vehicle. The middle level controller acts primarily as a data handler, channeling data between the other
5-7
controllers. Since the vehicle’s response is based on various asynchronous external stimuli, such as
navigation, communication, or environmental data, specific modular programs can operate on each stimulus
or command independently. In this w^, each specific modular function of the robot is capable of reacting to
the unknown terrain real-time. Likewise, the specific coordination of the overall behavior is still performed
by its central coordinator to achieve this objective. The centralized coordinator is flexible and does not
control the robot’s actuators directly or manage tracking and navigating by itself Instead, the coordinator
provides indirect control by selecting among alternate functional modules. Therefore, a combination of
simple subsumptive modules with some hierarchical central control to prioritize its decisions is used to
achieve the vehicle’s mission objective.
A test plan was written to exercise the various control systems for the different gatherer concepts.
To cany out the testing, a test field was established at NAVEODTECHDIV. The test plan includes
individual tests of local search routines, that are required to reacquire targets, obstacle avoidance and
maneuvering routines, and signal loss tests. Each of these tests were designed to gather information on how
the control systems perform. The obstacle avoidance and maneuvering tests include a long, straight obstacle,
an “L” shaped obstacle, and a bhnd alley. The gatherers must approach these obstacles fi-om different angles
and find a target on the opposite side. For the blind alley, the gatherer must find its way into the alley, find
the target, and find its way back out of the alley. The signal loss tests include the loss of GPS data fi-om the
satellites and the loss of radio fi-equency communications.
In addition to the individual tests, there are two multiple target “field tests”. For the UXO field test,
UXO targets are placed at many locations within a 100 foot by 100 foot test area, behind various obstacles *
and on several different types of terrain, such a sand pile and a pile of rocks. The test is to have the vehicle
autonomously attempt to find as many of the targets as possible, picking up and carrying one at a time, and
depositing them in a central disposal area. For the mine countermeasures field test, the vehicle is to
autonomously deliver a simulated anti-mine munition to each of four landmines that have been buried to be
flush with the surface of the ground. All targets used are metallic for some level of ease of detection.
Three other tests, included in the test plan, were designed to gather information on some particular
feature of the test beds, but not the control systems. These include the ability of the manipulator to pick up
an inert M42 grenade, used as our standard UXO target. Also, the ability of the manipulator to release the
UXO was tested, along with its ability to place the inert anti-mine munition. And the other test was a simple
straight line transit test to see how well the vehicle could travel to a point some specified distance aw^, with
no obstacles.
From the testing performed to date, we have been able to make the following observations. The
Draper Lab system has a very good control station, a very fast vehicle, and can perform waypoint navigation.
The NAVEODTECHDIV vehicle has been able to successfully perform local searches and manipulator
operations. The Foster-Miller vehicle has good terrain mobility, and can home to a beacon very well, even
though the radio fi-equency communication is very close to the ground and the antennae arr^ spacing is
tight. ISX / IS Robotics has a good operator control unit (on a laptop computer), and the vehicle can perform
the different functions autonomously, in an integrated fashion. Much of the value of these tests, though, will
be that they provide real world inputs to the computer modeling of the different proposed concepts.
MODELING / SIMULATION:
Simulation, obviously, is not a substitute for the real world. Only robots operating in the real world
terrain can provide the reliable data and results. There are many different environments that the simulation
cannot be characterized exactly or accurately. Likewise, it is difficult to build a "standard" simulation to compare
the performance of different system approaches to the same application. All BUG concepts are attempting to
solve the same problems; however, their system approach, response, complexity or trade-offs are different. It
5-8
requires extensive work fisr the programmer to integrate 4 or 5 different packages of simulation into one system.
On the other hand, while real world testing is always better, it is time consuming and expensive.
Direct ground testing requires time and experience of the developers to tell where the optimal
tradeoff point is. Depending on its tasks, real robots can damage themselves, their surroundings, other
equipment or possibly people. Real robots often malfunction due to temperature change, break down
mechanically or electrically, or exhaust their battery power. From this perspective, especially for hazardous
missions or big projects, simulation is important both for safety and financial reasons. In the case of BUGS,
simulation is used as a rapid prototype statistical analysis tool and a demonstration tool.
The Naval Postgraduate School is building a simulation that includes a world scale model of the Marine
Corps Air Ground Combat Center, Twenty Nine Palms terrain and a variety of robots that have the “same”
functionality or characteristics as a real robot. Since the study of agent group behaviors is an emerging field, this
simulation is a useful tool for studying the interaction and cooperation between a large number of agents in the
real world. This simulation will solve the difficulties of implementing actual real world testing or analyzing a
large number of agents in a complex world. In addition, the simulation provides the developer with a general
understanding of the total system characteristics through a large number of simulations of diverse situations.
The repeatability of the simulation can provide the designer with insight into the interaction of the ^ent s
behaviors and its environment and provides the operator with historical data records on each simulation for
further analysis. The simulation will be used during the entire life of the project, to support continuous changes
and to validate ideas for future improvement in a safe environment. This will simplify the developers’ tasks in
term of time, space, accessibility and costs over the long run.
Both UXO clearance and mine countermeasures scenarios have been modeled, as well as the various
gatherer vehicle concepts. Variations of sensor capabilities, navigational accuracies, number of vehicles working
in unison, and the hke are being explored throu^ running the model many times. The results of this modeling
will assist in the evaluation of different control strategies.
PLANNED WORK:
The planned work for the BUGS project for FY97 includes completion of the testing of the initial
individual gatherer vehicles, and completion of the computer modeling and simulation by the end of November.
Based on the test results, modeling results, and system concept study reports, a concept will be selected to
advance to the development of a multiple vehicle system. This multiple vehicle system will be demonstrated,
showing the utiUty of using numerous small robots to perform UXO clearance or minefield neutralization
missions. For the sensor platform for the UXO missions, work will continue on the development of the
autonomous RECORM vehicle, to autonomously search an area, detect and classify targets, and provide target
locations to the small gatherers.
SUMMARY:
The BUGS program is proceeding along as planned. Different control system concepts for the individual
BUG vehicles have been developed, demonstrated, and are being evaluated. They are being evaluated as a single
vehicle concept in actual hardware testing at a test site, and as a multiple vehicle concept in computer modeling
and simulation. The real life testing is being used to develq) the computer modeling. Demonstration of a selected
multiple vehicle concept will follow next.
In addition, an initial autonomous survey platform for visible UXOs is being developed, using an EOD
robotic asset, and adapting the Robotic Work Packages that have been developed for autonomous underwater
EOD use. .
The BUGS system vision of cheap, autonomous, simple robots, operating collectively to perform a
hazardous, but important mission, is proving to be a viable idea.
5-9
Small Autonomous Robotic Technician
David Kang
Draper Laboratory
Cambridge, MA 02139
1. Introduction
Clearing unexploded improved conventional land
munitions is an important task that currently requires slow,
expensive processes exposing personnel and equipment to
considerable risk.
Figure 1: Sweep team looking for UXO. Note the
In the current manual UXO clearing approach, areas terrain and natural obstacles that present a
suspected of having UXO are first partitioned into sectors challenge to robotic vehicles.
with comers delimited by flags. In each sector, a four- to
eight-man sweep team (Figure 1) visually scans the area for
UXO. Based on preliminary investigations, the risk during
5-11
Figure 2: UXO gathered by the sweep team In a PUCA
mission. Note the variation in size, shape, Figure 3: MITy-2 Micro-Rover
and appearance of each UXO. These MITy prototypes are the predecessors for the current
SMART system vehicles, which we have labeled the EOD
series. EOD stands for Explosive Ordnance Disposal.
5-13
The grappler is driven by two 24V DC motors'' equipped Table 1: Hardware Components for EOD-1 and EOD-2
with integrated gearheads and optical encoders. These BUGS
encoders provide feedback to the control system necessary
for autonomous PUCA missions. One of the 24 V motors is
used to actuate the scoop linkage, while the other is used to Component/ Capability EOD-l EOD-2
drive the rake. The rake is used to sweep the UXO into the six-wheeled, three- yes yes
scoop during the acquisition process. platform mechanical
architecture
six drive wheel motors on board on board
with integrated encoders
scoop mechanism for not on board
retrieval of UXO available
12MHz Z-World Little on board on board
Giant Microprocessor w/
512K SRAM&PI096
Digital Expansion Board
Systron Dormer micro¬ not on board
mechanical gyroscope available*
video camera & on board on board
transmitter
front bump sensors not on board
available
CONAC™ Local not on board
Positioning System (LPS) available
Polaroid sonar ranging not on board
module array available*
Proxim serial modem on board on board
Figure?: EOD-2 Grappler Mechanism (9600bps)
Radio Shack metal not on board
detector unit available
•Installation Planned
5-14
then locally integrated using a PIC16C84 microcontroller.
3.2.1 Sensors
The integrated signal is then scaled and returned to the
main processor as the relative heading. The signal
3.2.1.1 Sonar integration using a dedicated processor frees the main
Located at the front of the rover are three Polaroid sonar processor from the processor intensive task of numerical
ranging modules. The ranging modules work by emitting a integration.
series of sound pulses and measuring the time elapsed until
the echo returns to the transducer. The time measured can
then be multiplied by the speed of sound at ambient
conditions to calculate the distance to the nearest object.
3.2.1.2 Bumper
Also located at the front of the rover is a bump sensor. The
bump sensor is a small plate, spring mounted in front of
two electrical switches such that depressing the plate causes
the switches to close. The bumper signal is resistor tied
high and the switch is tied to ground such that when the
switch is depressed, the signal is pulled low.
5-15
processor to communicate with the device. With a local 3.2.3 Onboard Microprocessor
microprocessor doing much of the data processing, the bus
can be implemented with a serial communication protocol. The SMART system vehicle microprocessor is a 12 MHz
Little Giant processor made by Z-World Engineering*
based on the Z180 microprocessor chip. The Little Giant
The numerical integration of the rate gyro often leads to a
has 512K of battery backed SRAM, 16 digital I/O lines, 2
drift error if integrated over an extended period of time.
serial ports, an 8 channel A/D converter and a 12-bit D/A
Intermittent software re-calibration is performed by
converter. The Little Giant is programmed with a
tracking position change with the LPS and correlating the
proprietary variant on the C programming language called
movement with the readings from the gyro.
Dynamic C.
3.2.2 Laser Positioning System Programs are downloaded through an RS-232 programming
Using a dead reckoning scheme for vehicle navigation port. Once a program is loaded, the Little Giant can be
eventually leads to the buildup of absolute position error. configured to run the program automatically on power-up.
In order to resolve these navigational errors, an absolute Attached to the Little Giant processor is a PI096 digital I/O
positioning system should be used as a verification of the expansion card. This card adds 96 additional digital I/O
dead reckoning results. The Draper SMART system ports to the Little Giant’s 16. The Little Giant is very
accounts for these inherent errors using a laser positioning simple and quick development platform and is excellent for
system (LPS)^ small embedded projects.
• Easily modified
• Easy to test
• Expandable
• Flexible division of labor between vehicle and
remote control station
5-16
Because testing an embedded system can be difficult, a full
simulation of the vehicle and its control software can be run
on the control station. This simulation is identical to the
vehicle’s control software down to the driver interface
level.
4.2.1 Hardware
The SMART system control station runs on a Pentium
desktop PC under the Linux operating system. This
platform is less expensive than more specialized systems,
but provides ample 32 bit computing power for this
application.
• Obstacles Tasks can be pushed onto or popped from the task stack
• Sonar Hits from the mission control window. When a task is being
• Path History added to the stack, a task creation window opens and
• Planned Tasks displays the task parameters. If the planned tasks overlay is
• Map Grid visible, the mission control window will display the
proposed task as the parameters are changed. The available
tasks include:
5-17
• Waypoint Follow - Execute multiple the most efficient path to that location, given the initial
connected segments in succession. obstacle information.
• Transit - Plan a series of waypoints to a
desired location which avoid known obstacles. The A* search is a method of finding a minimum cost path
• Area Search - Traverses an area using a between any two nodes of a cyclic graph. An A* search
search pattern and attempts to detect UXO. differs from other search algorithms in that the cost
• UXO Pickup “ Pick up a UXO with the associated with a particular node in the solution includes an
grappler mechanism. estimate of the cost (called the heuristic) to complete the
• UXO Dropoff - Drop a UXO held by the search (i.e. to reach the objective node from the current
grappler mechanism. node). By proper selection of the heuristic, trade-offs can
• Simple Control - Allows the user to take be made between the optimality of the solution and the time
direct control of the vehicle. required to generate the solution. In particular, by selecting
a heuristic that is guaranteed to always underbound the
The control station also includes a debug window. This actual cost to complete, it can be shown that the A* search
window displays all state information for the BUG such as: produces an optimal solution.
steering position, wheel velocities, onboard system status,
and battery charge. An example solution of an A* search using an occupancy
map is shown in Figure 13 with obstacles mapped in the
mission area. The environment is broken up into known
empty, known occupied, and unknown (yet to be mapped)
5. Simple Mission Strategy
regions. The planning problem is to provide an obstacle-
free path from the start node, S, to the goal node, X. The
5.1 Basic Assumptions planner returns the shortest obstacle-free path, minimizing
It is assumed that the locations of the UXO within the area exposure to unknown areas when known empty areas are
of operation are knovm a priori, but that the terrain not available. Figure 14 shows the implementation of the
conditions are not. The destruction of the gathered UXO in A* search on the graphical display of the SMART control
station.
the ODA is not addressed directly, although a camera will
be positioned in the ODA for remote examination of the
deposited UXO before a human technician places a
detonation charge.
5-18
5.2.4 Searching
After the vehicle has successfully navigated to the
approximate location of the UXO, a search algorithm is
initiated in order to precisely locate the UXO. During this
phase of the mission, the grappler mechanism is extended
in order to use the metal detector to locate the UXO. Then,
the BUG follows a series of search patterns until the UXO
is discovered. The search routine covers an area over the
approximate location of the UXO that resembles a spider’s
web (Figures 11 & 15).
5-19
6. Advanced Strategies and fi-om the approximate UXO location and the ODA.
This process is expedited using roadways.
While using a control station and a single SMART vehicle
to solve the UXO clearing problem is effective, it is not the
During the mission, the control station is capable of logging
most efficient solution. The lUVC is continuing its
the paths followed by each vehicle. These paths, because
research toward more advanced and efficient strategies for
they have already been traversed by an agent, can then be
UXO retrieval using multiple BUGs, environment mapping,
declared as roadways. As a BUG encounters one of these
“roadway” building, and by determining the optimal
roadways (current coordinates of the BUG coincide with a
distribution of processing power between the robotic agents
and the control station. previous path), it is capable of proceeding at a much faster
pace due to the fact that the path should be free of obstacles
and UXO. This procedure, however, does require that the
6.1 Multiagent Approach
unknown environment is somewhat static.
Using more than one SMART vehicle to perform PUCA
and BIP operations in UXO clearing scenarios is a
challenging task.^ There are, however, advantages to using
more than a single BUG.
5-20
control station is damaged, the entire system is rendered 7.3 f C Serial Bus Architecture
useless. The current architecture has custom parallel interfaces for
each device. This architecture requires a large number of
This would not be the case, however, if each agent were wires making assembly, maintenance and repair difficult
given ample computational resources to carry out its part of and time consuming. The proposed change in architecture
the mission in the absence of the control station. However, will replace the large number of parallel wires with a two
the complexities of inter-agent communication and the wire serial bus. Sensors and actuators will be modularized
additional hardware and software needed on each vehicle and connected in a daisy chain fashion along the two wire
make this a much more complex issue. The optimal bus. The x86 processor will serve as the master and control
distribution of computational resources would be one that the flow of data on the bus. Each device on the bus will
combines the best of each extreme. have a unique 7-bit address.
7. Parallel Development The two wire bus will greatly reduce the amount of manual
labor involved with manufacturing and assembling a
While developing the current UXO clearing system, the
vehicle. A problematic device can be debugged by simply
lUVC has also been looking ahead to create the next
detaching it from the network and testing it in a stand-alone
generation SMART system. This advanced system
environment. In the current architecture, each device has a
includes upgrading to a new vehicle microprocessor (x86),
custom interface requiring a custom test setup. With the
realizing the capabilities of the Global Positioning System,
proposed architecture, each device will have a standard two
and using a new serial bus architecture (PC).
wire interface allowing the same test setup to be used for all
devices.
7.1 Microprocessor Upgrade
Limitations on the processing power and the development To counter the reduced bandwidth of serial versus parallel
language of the Little Giant processor has forced a data buses, each module will process the large amounts of
migration to a more powerful platform in order to perform raw sensor data into high level data. Each module will
more complex behaviors. The proposed platform is the x86 contain an embedded micro-controller dedicated to process
or the Intel PC architecture. The x86 processor will allow the raw data as well as run the low level hardware interface
the choice of many operating system environments such as code. Already in place in the current architecture are
DOS, UNIX, Linux and QNX. The new processor will also dedicated, embedded processors that control the sonar and
increase the processing power from a 12MHz Z180 to a the gyro. With sensor and actuator software being executed
50MHz 486DX as well as allow the use of off-the-shelf locally at the sensor or actuator, the main processor is free
components such as flash drives, Ethernet interfaces, to perform other tasks such as path planning unencumbered
PCMCIA interfaces, etc. by the processing needs of low level hardware drivers.
5-21
Enabling Techniques for Swarm
Coverage Approaches
IS Robotics ISX Corporation
Helen Greiner, Colin Angle Richard Myers
Joseph L. Jones, Art Shectman
Abstract
The use of multiple low cost robotic mine hunters to provide rapid and complete area
coverage represents a promising new approach to the counter mine problem. With this
approach, however, comes a new set of problems for the effective implementation of this
technique. How can a lightly trained technician operate such a complex system? How
much of the terrain of interest in inaccessible due to trees, rocks and bushes? And how can
I be sure that the robots have done their job? By augmenting our Behavior Based local
navigation software with a supervisory control interface and a GPS mapping and directed
search engine, IS Robotics has developed a swarm control system capable of operating
large numbers of hunter vehicles. We have also developed the mine hunting vehicles with
embedded intelligence capable of fully utilizing this control system. The machines integrate
local terrain sensing, high accuracy GPS, robust mobility, support for task specific
sensors, computational assets, and power systems in a cost effective manner.
Fundamental to the Swarm approach is the assumption that large numbers of vehicles can
operate in parallel with little or no operator interaction. Simple intelligence schemes have
been shown in simulation to produce exciting results, but the real world offers far greater
challenges. Unfavorable combinations of obstacles, terrain slope, and poor traction can
introduce systematic effects into "random" search. Such effects eliminate any guarantee of
complete coverage by a robot using random search. Directed search, search based on
global methods, can produce superior results by working from an explicit representation of
areas covered verses areas not covered. However, even directed search coupled with
onboard intelligence does not solve the whole problem.
Despite great advances in navigation and other technologies, the challenge of continuous
duty unsupervised operation in natural environments has not yet been met. Eventually, a
robot that attempts total autonomy will be stymied by an unplanned condition or a
pathological combination of obstacles or other hazards. Our approach is to add a layer of
software that 1) monitors robot progress to detect such conditions and 2) alerts the operator
and allows him or her to intervene. This method produces a robust system that places
infrequent demands on the operator and reduces the robot design challenge to a manageable
level. We show that a production system can be manufactured at an acceptable cost.
5-23
Automated Systems without these
1. Introduction features will lack the functionality and
robustness to effectively assist the MCM
technician. Consequently, the system will
Technology that will largely eliminate
have little chance of being incorporated
human risk in mine clearing is at hand.
into a military mine countermeasures
This high level of risk reduction is
doctrine.
becoming possible because of advances
in microrover, positioning system,
Many MCM scenarios would benefit
communication, and mine detection and
from automation. These scenarios
neutralization technology. In this paper,
include: mapping, marking. Pick Up and
we will describe several components of
Carry Away (PUCA), and Blow in Place
the new technology important to the mine
(BIP) operations. IS Robotics is
remediation problem. Features of a
currently working on two such systems.
control structure that utilize this
The first system is FETCH, a munitions
technology effectively are detailed. In
countermeasures system that is part of the
addition, we review two IS Robotics
EOD BUGS Program. Fetch has
experimental robotic mine counter¬
demonstrated all components of the full
measures systems that bring the required
system in a cluster munition PUCA
technology and control together in one
operation. It has demonstrated placement
package.
of a simulated shape charges for a BIP
operation. The second system is Hum-
This automation assisted mine re¬
De, a joint Tracor/Tracor GDE/IS
mediation is the natural “next step” to
Robotics Internal Research and
current hand-held detection systems and
Development projeet, whieh eoncentrates
remediation practices where humans
on mine detection, mapping and marking.
come in proximity of UXO. In designing
The Hum-De Program combines an IS
an automation assisted mine remediation
Robotics mobile platform and the MCM
system, a control structure for the
control structure with state-of-the-art
complete system must be developed.
Tracer GDE Systems sensor detection
This control structure must make the new
and data processing (recently selected for
remediation system a tool that a technician
further development under the
can use effectively. As such, we need to
HSTAMIDS Program.)
recognize mine countermeasures (MCM)
operational requirements, allow for
human supervision and control when 2. Current Mine Hunting
desired, and incorporate current mine Doctrine
hunting doctrine.
2.1 Buried Mines
IS Robotics has developed a control
stmcture for MCM operations that is In order to understand automated
based on the following important techniques it is important to recognize
capabilities: current mine hunting doctrine. In this
paper, we use the term MCM Engineer
1) Supervised Autonomy (operator may (Mine Counter Measures Engineer) to
take control at any time) mean Army Combat Engineers, ^D
2) Graceful Degradation (tolerates failures technicians, or other personnel involved
or noise on terrain sensors) in mine remediation. MCM Engineers
3) Spatial Coordination of Sensor Data currently sweep minefields by walking
(improves detection and reduces falses the length of the field with hand held
positive rates) pulsed induction mine detectors. The
4) Certifiable Coverage (reports to reliability of this equipment depends on
operator that an area was cleared) environmental conditions. Excessive
5-24
metallic content (bullets, shell casings, operations. Also, unlike buried mines,
frag) in the ground must be filtered out by sub-munitions are found using visual
reducing the sensitivity of the detection inspection and not metal detectors. Just
equipment. At some point sensitivity as buried metallic debris can cause
reduction compromises the effectiveness problems for metal detectors, any debris
of the detector and manual methods must that clutters up a field will make visual
be used. identification of munitions more tedious.
If the engineers locate a potential mine, One method for clearing a field of sub¬
they attempt to identify it using a munitions has MCM Engineers walk
fiberglass stick. These sticks are pushed shoulder-to-shoulder at two arms lengths
into the ground at a 35 degree angle until apart across a swath of the field to be
the stick hits an object. The angle searched. As sub-munitions are
ensures the stick hits the side and not die identified via visual inspection they are
top of the mine. An object discovered in marked with flags. Non-explosive debris
this way must be dug out in order to can also picked up to make a second
ascertain its identity. The mines are inspection easier. Unfused high-
normally buried approximately 4" below explosives may also be collected and
the surface, if they were deeper and they piled near marked sub-munitions.
would not be effective. Because they are Because of the risk of detonation, sub¬
not buried deeply, the probing need not munitions are not handled unless
be very forceful. absolutely necessary. After all munitions
have been marked, shaped charges are
Once identified the mine is either blown placed at each and connected by flash
in place, removed and burned, or simply cord. The munitions are then blown in
removed. Blowing in place is common, place simultaneously from a safe
but in some cases, such as urban distance. Although this procedure was
neighborhoods, the environment is not observed during a range clearing
suitable for a high order detonation. operation at the Twenty-nine Palms
Blowing in place is currendy carried out Marine Base, it may not be relevant for all
using anti-mine shape charges. Burning situations.
of the mine is preferable as it greatly
reduces collateral damage due to 3. Technology
fragmentation. Using this method, the
detonation potential is reduced by
Recent technological advances are, for the
destroying much of the bulk charge
first time, making possible automation
before the detonator fires. Removal of
assistance in the mine clearing problem.
the mine is the most environmentally
In the past, integrating all requisite
friendly method, but it is also the most
capabilities into a man-portable, low cost
time consuming and dangerous. All three
package was not feasible. We feel that
of these methods are being employed in
systems should be designed with
Bosnia. All can benefit from automation.
available technology as field testing in
real conditions is the only way to
determine what the true operational
2.2 Munitions problems will be. As sensory systems
The clearing of unexploded sub¬ and embedded intelligence programs
munitions (such as M41 anti-tank improve, increasingly difficult terrain and
munitions) from a battlefield is more challenging circumstances such as
particularly dangerous due to the large canopy cover will be targetted.
number and small size of these objects. Approaches that argue for limited sensory
The unpredictable state of sub-munition capabilities and no global position
fuses due to ground impact and exposure information need to be seriously
to the elements also complicates clearing questioned as to their operational value.
5-25
1) Positioning Systems increased from virtually no capability
Positioning systems let the rovers cancel prior to 1990 to near 70% probability of
drift in their perceived location and allow detection.^ In the coming year, this
pattern based search strategies, and, more program will focus on multi-sensor
importantly, they allow the creation of a integration, detection algorithms
coverage map. Carrier Phase Differential development, and enhancement in user
GPS systems are giving accuracies of <1 interface. We believe the technology
inch under ideal conditions. These off- developed under this program can be
the-shelf systems are sufficient for a effectively used by automated vehicles.
variety of terrains such as beaches, fields,
and desserts. In terrains with canopy 4) Multichannel Communication Systems
cover, buildings, or other occlusions, Advances in spread spectrum technology
alternate positioning systems may be allow large numbers of robots to
swapped in. For these areas, laser communicate all using the same
position systems' and RF localization frequency band. Commands and data can
methods are available. A more advanced now be effectively shared with a base
ultra wide band radio approach is under station and thus other agents. (However,
development for the military. This compared to purely teleoperated devices,
system, if the production version systems using supervised autonomy need
achieves preliminary specifications, more less communication bandwidth)
than meets the requirements for UXO
cleanup operations. It features good 5) Neutralization
penetration characteristics claiming Blow in Place (BIP) neutralization
range/accuracy of 1 km with 3 inch techniques are proving both performance
accuracy non line-of-sight (LOS). and cost effective. For example. Tracer’s
mine neutralization shaped charge is a
2) Microrovers recently developed and field demonstrated
The size, weight, and footprint of shaped charge explosive device proven
microrovers combine to give the devices a effective against all known mines. They
very low surface pressure thus are capable of penetrating through several
minimizing the risk of accidental inches of soil or water to destroy buried
detonation of pressure mines. IS or submerged mines in-situ without site
Robotics is in the business of designing preparation.^
and building microrovers for applications
such as reconnaissance, surveillance, Continuing trends in cost reduction of
planetary exploration, research, mine advanced technology and production
countermeasures, and hazardous material scale manufacture will reduce the sale
handling. We have proven that price of capable vehicles. In fact, the
sophisticated systems with embedded European Joint Research'' center sees
intelligence can be man portable, and thus vehicle deployment as the only way to
microrovers are now considered feasible meet cost effectiveness goals set by the
for many more applications. Microrover United Nations.
make use of developments in a wide
variety of domains such as increased 4. Enabling Techniques
battery capabilities, electronics
miniaturization, improvements in
connectors, and sensor system
developments. 4.1 Supervised Autonomy
5-26
perform a mission while responding to robotically controlled sensor, such as the
real world conditions (note: behavior Hum-De, can give.
control is described later in this paper).
However, to increase system reliability 4.2.1 Differential Data
and operational control, Supervised Techniques
Autonomy allows a MCM technician to
take direct control of the rover whenever A powerful method for improving the
he needs or desires. We recognize that signal to noise ratio from the sensor in
the current state of artificial intelligence mine detection applications is to
coupled with sensory limitations differentiate the stream of data. This
precludes rovers from autonomously differentiated data strongly shows where
dealing with every contingency. step changes in readings occur, while
Unexpected terrain features or other tending to filter out slow changes and
pathological conditions may cause failure steady state value. Since a mine is a
in the automated routines. The supervised discontinuity in the normal soil,
autonomy approach reduces the rover differentiated data should contain less
control system design challenge to a noise and stronger relevant signal
manageable level. information.
5-27
hunting domain. The use of geometric
methods in dismounted operations is Most parameters have been optimized for
currently left entirely in the hands of the a general case situation, while others can
combat engineer or EOD technician. be manually adjusted to calibrate the
Since it is difficult to predict or assume sensor to a given terrain. This is a
any advanced spatial ability in a given practical solution, but not an optimal one.
soldier (or anyone else), use of any In order to do better, the sensor needs
geometric algorithms is very limited. on-board intelligence which looks at a
Vehicular mounted mine sensing arrays sensor’s readings (and potentially other
have the ability to collect spatially system sensors), and determines a new
correlated sensor data. However, due to sensor parameter set to use.
the arrangement of the sensors, resolution
is a problem (there do not exist arrays The system would also have to have the
capable of collecting spatially correlated ability to go back and rescan an area in
sensor data better than data lines such a way that the results of the first
separated by 6”). This severely limits the scan were spatially calibrated with the
use of geometric techniques against all second. This correlation will allow more
but anti-tank mines, and even there, sophisticated algorithms for data
severe limitations exist. processing to be developed. Specifically,
a high sensitivity setting could be used to
A sensor control system which allowed ensure that a target was not missed, but
the collection of data in fine position once found, this high sensitivity setting
correlated arrays would open up a new would not yield suitable geometric shape
set of powerful tools for the detection and or mine centroid information. By
classification of mines^. Algorithms reducing the sensor’s sensitivity and
designed for image processing could be rescanning the area, a more accurate
brought to bear on such data. The target shape may be found. Additionally,
symmetric or otherwise characteristic multiple sensor runs can be differenced
shapes of land mines could be used to and changes based on parameter
reduce false positive detections, and more adjustments between runs can be found.
accurately determine the position of a Since different materials absorb EM
mine. There currently exist innovative energy differently at different
ways of disabling mines through the use frequencies, this parameter differencing
of shape charges which cannot practically technique may be very powerful in
be used in many instances due to the classifying potential targets.
uncertainty of mine location. Geometric
techniques could solve this problem and
pave the way for complete robotic 4.3 Graceful Degradation
demining solutions.
Graceful degradation means the ability to
function at a reduced level of performance
4.2.3 Parameter Optimization in the face of noisy signals, sensor
failures, or unexpected conditions.
Many mine detection sensors have Graceful degradation is a key feature
operational parameters which affect the (arguably the most important feature) for
way they collect data. In the case of an autonomous system. Most biological
Pulse Eddy induction sensors (PE), the systems exhibit graceful degradation to
magnitude and shape of the current pulse, the extent of functioning with loss of
the gain of the detection amplifier, and limbs or complete loss of sensory
potential analog and digital filtering of the systems. To exhibit graceful
return signal are all variables. GPR degradation, robotic vehicles need to be
sensors have a host of similar parameters designed from the bottom up^. They
including the frequency of the emitter. need to be more than just cars or tanks
5-28
with sensor packages tacked on. information. Compared to traditional
Graceful degradation is facilitated by a sensor fusion approaches, the cost
behavior control approach, a robot sensors and computational requirements
control architecture proposed by Prof of employing sensor data is reduced in a
Rodney Brooks of the MIT Artificial behavioral control system. Bumping an
Intelligence Laboratory^. obstacle while attempting to skirt it, as the
robot does in this example, may not be
Behavior programming is a decom¬ the optimal. However, such a strategy
position of the robot control problem. usually works — hence the term “graceful
Rather than separating the elements of degradation”.
robot control into a sequence of
functional steps, a behavioral control
program decomposes the problem into a 4.4 Coverage Certification
number of task achieving behaviors aU
running in parallel. The control software
in a behavior control robot can be simpler 4.4.1 Coverage Map
to design because the behavior modules
are task specific and, individually, need One important task of a mine
not be made general purpose. countermeasures system is to provide
military personnel with assurance that an
The behavior control approach depends area is cleared. System designs should
on layers of behaviors that perform output coverage maps. A realistic
redundant functions. The most effective example output would show that the
way to explain this is with an example. rover had swept Area A1 with a Ground
Imagine a rover commanded to perform a Penetrating Radar and a Pulsed Eddy
search for munitions in an area. The Induction sensor at a rate of V and height
rover comes to a boulder blocking its H. A map of the area covered would be
path. Under most conditions, proximity provided with potential mines clearly
sensors (IR, sonar, or laser) sense the indicated. Just as importantly the final
blocked path and the rover goes around map will clearly demark areas not
the boulder resuming its course on the covered. Sensor readings and visual
other side. However, now imagine the images from the rover will give an
boulder instead a small bush, the rover indication of why the areas were not
comes to it and the IR fails to detect it. covered. Reasons for non-coverage
Luckily, redundant sensing allows the could include bodies of water, terrain
sonar to sense it and the rover succeeds taxonomy that is too difficult to obtain
negotiating around. To further thwart our accurate readings, excessive metallic
attempt, now the sonar fails to sense the clutter, or dense brush. This information
bush because it is positioned too high. is vital in evaluating how sweeping
The rover hits the obstacles; tactile should progress.
sensors detect it. Tactile sensing is an
integral part of any autonomous system.
Using these readings the rover can “bump 4.4.2 Data Logging
and turn” around the obstacle. This list In our approach, a coverage map is
of possible failure modes and backup created at an Operator Control Unit
systems can be extended to include stall (OCU). At the OCU, data from multiple
and velocity sensing on the drive wheels, assets is integrated into a coherent
inclination sensors, and impact sensors. picture. The system architecture allows
for local storage and processing of data
Behavioral programming doesn’t depend and off-board mass storage and
on a particular sensor to be absolutely archiving.
reliable under all circumstance, rather we
depend on many types of sensors to Local storage of sensory data is important
provide redundant sources of so that the mine countermeasures system
5-29
is able to run automatic target recognition operations. Automatic detection
routines. Data that would require too algorithms seek not to replace humans
much bandwidth to transmit may be highly evolved, acute sensory systems
manipulated on board. such as sight, but to augment them with
non-intuitive sensory information such as
Off-board storage is essential to provide pulsed induction sensors and ground
the coverage map and integrate penetrating radar. Performance will be
information from multiple assets. Off- more consistent as reliance of the training
board storage also provides the following and perception skills of individual
benefits: technicians is avoided for the non-
intuitive sensors.
1) Provides backup in case of accidental
detonation
2) Logs “keep away” zones around 5.3 Force Muitiplier
detected mines and makes this
information accessible to all assets. A swarm of rovers are performing the
3) Allows predictive search- Many mine clearance task in parallel. Each rover may
fields are laid out according to known be slower than a technician performing
mine doctrine. While this is not the case the same task'. However, since one
in Bosnia or most of the Third World, in technician is able to control many rovers,
situations where this practice was his overall effectiveness is increased.
followed, knowledge of the location of Under supervised autonomy, tasks which
some mines can be used to predict the vehicles are able to perform well, such as
location of others. A mine hunting driving a sensor at a constant speed and
system can keep track of this position^ local obstacle avoidance are left to the
information in such a way that pattern robot. Tasks requiring higher level
matching algorithms could be easily cognitive capabilities, recognition, or
applied to the data. This approach makes reasoning are left to the operator. Under
use of the information to improve the this paradigm, the cognitive load on the
speed and ultimate accuracy of a clearing operator from each robot is reduced, and
operation. he can effectively supervise many rovers.
4) Data on mass storage devices can be In addition, the rovers can each signal the
saved for later analysis and improvement operator for aid if, for example, the
in algorithms. onboard diagnostics signal a problem or
if a particular command fails to be
completed in the expected time.
5. Advantages of Carefully
Designed Automated 5.4 Sensor Data Quaiity
Approaches
Human factors, the proficiency of a
5.1 Risk Reduction particular operator to sweep the sensor
consistently, currently play a large role in
Humans need never enter the mine field. the detection ratios. More consistent
But, their ability to supervise increases application of the sensor yields more
the ability of the system to adapt to totally consistent results.
unforeseen circumstances.
5.2 Training
The level of training - or the operator’s ‘ Coverage speed will probably be limited by the
ability to pick up audio and visual cues maximum speed ratings of the GPR and PI
that can help indicate the presence of a sensors. Thus, we expect the operating speeds of
mine - plays a large role in the detection the vehicles to be equivalent to that of handheld
probabilities in current demining systems.
5-30
Robots excel at scanning sensors at a (OCU) and a number of small robotic
constant speed on flat terrain. Ability to agents. These robots use a strategy of
both drive slowly at a consistent speed supervised autonomy to locate, pickup,
and the ability to speed up during terrain and carry away unexploded munitions.
traversal is desirable. Position feedback (A blow-in-place capability has also been
and a well-designed velocity controller shown.)
are necessary for application of a variety
of sensory elements. The robotic
approach is expected to surpass current
capabiUties where sweeping speed is
dependent on the proficiency of the
operator.
5-31
and the ability of one subsystem to
backup another. Fetch includes
systems that detect and respond to
hazards, a software architecture
(Behavior programming) in which
system backup is inherent, and a
facility that allows an operator to take
either supervisory or direct control.
Further, Fetch monitors its own
progress and can alert the operator if
an unexpected condition arises.
5-32
progress and alert the operator if an Munition pickup
anomaly is detected.
Fetch includes a rudimentary pickup
Navigation system. This system is composed of a
one degree-of-freedom arm with attached
To direct the robots to known locations of electromagnet. A break beam sensor
munitions an accurate navigation system monitors the area along the bottom
is used. This system also allows the surface of the magnet. In this way the
operator to verify coverage of any areas robot can determine when it is holding a
that require search. Currently, Fetch munition. An automatic munition pickup
employs a carrier phase GPS system that procedure was developed. To pickup a
provides two centimeter accuracy under munition the robot; lowers the arm, turns
fevorable circumstances. Heading on the electromagnet, raises the arm, and
information is obtained from a magnetic checks the break beam sensor. If a
compass with nominal 2 degree accuracy. munition is present, the pickup sequence
Dead reckoning complements the GPS successfully terminates. If not, the robot
system by providing navigational data to backs up and repeats the sequence.
the robot between GPS updates. Dead
reckoning can be used exclusively (but Operator Control Unit
with reduced performance) if GPS fails.
The Operator Control Unit (OCU)
Obstacle avoidance/escape implements three major control functions:
situation awareness, supervisory control
Onboard sensors allow the robot to sense and simple task management. Situation
and respond to hazards such as obstacles awareness gives the operator a rapid
and rough terrain. Currently, Fetch uses understanding of the status of the Hunter
infrared reflective sensors to detect local vehicles and of the area currently being
obstacles. A front mounted bumper searched. We accomplish this by
monitors collisions and inclinometers displaying a two-dimensional scrollable
measure the slope and roughness of the map of the search area marked with icons
terrain. to indicate the relative positions of the
robots, obstacles, target munitions and
Search terrain features (see figure 2). A Hunter
icon changes color to indicate the current
Because the robot cannot depend on status of the robot; red indicates a fault or
precise advance knowledge of munition lack of progress and green indicates
locations, a search capability is included. operation within normal limits.
This search component consists of a
systematic procedure to hunt for These icons also indicate the heading of
munitions and a metal detector to alert the the robot and its position relative to
robot that a munition has been found. In domain objects and other robots. The tab
operation, the robot spirals outward from key rapidly cycles the focus of attention
the expected position of a munition. If from one robot to the next. Obstacles
the robot encounters an obstacle in its detected by the Hunter or manually
path, it reflects, i.e. the robot reverses entered by the operator are displayed as
and continues the spiral in the opposite gray boxes with two black crossed lines.
direction. This strategy insures Steep parts of the terrain that are detected
maximum coverage even in the presence by the Hunter will be marked as either red
of obstacles. If no munition is detected, or yellow areas, depending on the
the search routine stops when the robot inclination detected by the robot and areas
exceeds a predefined radius from the that have been searched will be marked
start location. with open gray squares. When the robot
detects a munition, the approximate
5-33
location of the detected munition is find and pickup a munition, transport the
indicated with a red circle. munition to the disposal area, and place
the munition on the ground. In addition
By selecting a Hunter icon and clicking to autonomous operation, high level
the mouse, the operator can issue simple supervision and direct operator control
supervisory control commands to the were also demonstrated.
robot. These include a "goto" command
which can be issued by placing a flag on 6.2 Hum-De
the screen which the robot will then
autonomously travel to. A Hunter can
also be told to "search" an area. This will
6.2.1 Hum-De System
cause the robot to begin a spiral search
pattern, stopping when a munition has
been detected. Once a munition has been The Highly Mobile Mine Mapping,
detected the operator can command the Marking & Detection System, HMMMM-
robot to perform an automatic "pickup" D or Hum-De, is a Joint Tracor/IS
maneuver to grasp a munition. The Robotics IR&D program to demonstrate
operator can also use the OCU to remote detection, mapping, and marking
teleoperate a robot by using either a of landmines and UXO.
joystick or keyboard commands and live
video transmitted by the robot. This Hum-De is a high mobility platform
allows the operator to maneuver the robot designed to deploy an advanced
out of terrain traps, to identify detected GPI^metal detection sensor. The Hum-
munitions and to manually pickup De platform combines its autonomous
munitions that are difficult to navigation system with remote
automatically grasp. supervision by a human operator through
a portable base station. The small
System integration mobility platform is only 30” L x 26” W
X 14” H and weighs under 60 lbs. The
All the above components must be fully vehicle was designed and fabricated in
integrated in a final system. With such less than 4 months and is currently being
integration in place the operator can bring prepared for field demonstration tests.
about a complete clearance operation by
issuing a single command of the form:
Dispose {robot-designator pickup-
location drop-off-location). Upon
receiving such a command, the robot will
autonomously navigate to the pickup-
location, search locally for the munition,
find and pickup the munition, navigate to
the drop point, and deposit the munition.
All this occurs without further operator
attention.
Fetch demonstrated all the above aspects The Hum-De combines state-of-the-art
of the munition clearance task at its final Tracer GDE Systems sensor detection
evaluation at the EOD site in Indian Head, technologies and data processing onto an
Md, November 1996. Operating IS Robotic mobile platform with unique
autonomously. Fetch was able to navigate navigation capabilities and embedded
to a given point, perform a local search. intelligence. The Hum-De system
5-34
consists of the the mine detector built by minefield until it has been cleared of
GDE, the Mobility Platform, Sensor mines. In addition, automated systems
Sweep Arm, and Operator Control Unit have the potential to out perform hand¬
(OCU). held systems through precise sensor
application and spatial tagging of data.
Thus, more effort should be spent on
applying the sensing technology to
automated techniques.
7. Conclusion
Applying automation to the world-wide
mine remediation problem makes sense.
These systems look toward a future
where no human need be sent onto a
5-35
5-36
Control of Small Robotic Vehicles In
Unexploded Ordnance Clearance
A. J. Healey\ Y. Kim
^ Point of Contact
(408)-656-3462
(408)-656-2238
healey@me.nps.navy.mil
5-37
note that the time is inversely proportional to the number of
The analytical consideration is useful in that it shows robots, N, and p, the conditional probability of detection
the importance of the traverse speed, the detection radius and given that the target is within range of the sensor. While
the proportional influence of the number of robots in the this performance indicates that the faster vehicle clears in
field as well as the importance of a high probability of shorter time, and that increasing the number of working
detect, p. vehicles and the detection radius has a proportional benefit,
increasing N also reduces y so that a limit exists to the
Random search using cheap robots has been proposed in benefits of increasing to number of vehicles.
[3]. In [4, 5], we show that the random search methodology
together with a bounding signal (electronic fence) would be TARGETED SEARCH
possibly preferred for low cost vehicles (without precise
navigation) It was also shown that depending on the With the benefit of high precision navigation, it is now
placement point used, the coverage by multiple robots may possible that not only could an exhaustive search be
be skewed towards the placement point so that multiple undertaken by a fleet of robots, but also, if an external
placements are desirable. Homing to a pile point can be means of providing targeting data (expected location of
accomplished with a placed radio beacon targets to be found and recovered), then, advantage may be
taken of the knowledge of the terrain freeways to increase
The requirement of having to perform obstacle travel speed in certain paths, while slow speed search with
avoidance maneuvering while in transit adds time to the obstacle avoidance in unknown sections will produce the
search. Results have shown that there is an added time knowledge necessary to map building.
consumed by obstacle avoidance (including avoidance of
other vehicles) that reduces the effective speed, so that At this point, not all segments of area need to be
searched, and only those local areas where targets are located
need to be searched. In this case, the expected clearance time
is
where y, is a reduction factor based on the density of To =d('LWr*
obstacles, time lost to obstacle avoidance and the number of
vehicles in the search. in which, i^Jo) is the average time spent in obstacle
EXHAUSTIVE SEARCH avoidance for obstacles, i( p) is the average time spent
in locally searching targets with sensor of detection
Studies of the threat to robotic clearance systems probability, p and d is the average distance traveled in
indicate tliat the majority of items will be ferrous in nature pickup and return of all targets.
so that magnetic detection coils could be used to advantage.
On a limited size / cost platform, detection radii not more SIMULATION AND MODELING
than approximately 20 cm. are possible. It follows that
directed searching is not likely to be better than random Operation of the robot vehicles is complicated by the
searching unless navigational accuracy within centimeters is fact that navigation over rough terrain is required at the same
available. With the recent developments in differential time, obstacle avoidance behaviors have to be running.
GPS positioning, accuracy to within standard deviations of Behavior based control [7] is used with the exception that
less than 2cm. are now claimed [6], which opens the arbitration between concurrently running behaviors is
possibility of directed searching to be accomplished with the simplified to that of switching between discrete modes while
detection sensors available. algorithmic control laws are used to control the behaviors.
An overall canonical automaton for the discrete event control
In directed search, the area is swept a constant rate - of each vehicle is given in Figure 1.
either in spiral directions, or in a lawnmower pattern. The
mean clearance rate is constant at Robot Navigation
until the field is cleared. The expected time for 100% ^com Wcom y^(0)
clearance is then,
subject to rate limits from the actuators while the
To^A/r(n,,N)U(2r)pN commanded heading is randomized as appropriate and given
an additive bias depending on its position relative to the field
5-38
BUGS Canonical State Diagram
T0=Suit
SI =Read_NMt_Tirgel_Loc*tion T1 sReodve^SdraitH
SiOo.Wiypoint _Trinwt_to _Next_Tiige{ T2=0bttjckJ><ecled
T3=Obftacfc_ClearftAln_LociI_A«a
S3=Do OtwUcle^Avoidanoe
S4=Do U)CtI_Search T4=OiUck QMrAAto TVtnritto.Pile
T5=Oi>ftick_QMr AAln_Wiypoinl_Trai»it
S5=PerfonTJ PickJUp
S6=i n_TranBLtoIRifc
Figure 2 Obstacle Avoidance Scheme Using Forward
T7si*ict_ipJ)one
TfeTuiie_0«L Seirching_E»oee<fe<l Looking Sectors
TS>=Dropped_OQ _Pile
T10=At_T«i«LAiei
Tl lsTin»e_ew»e<Jed_Pick^Up
Target Detection and Pickup
System effectiveness results using a "C" coded program Figure 3 Uniform Target Detection Probability Density
have supplemented a concurrent graphics based simulator Distribution
development, and now allow for large numbers of Monte
Carlo simulations to be conducted in short times. A successful pick up is assumed if the vehicle can
position itself such that the target (xj,y2 ) actually lies
within R, the detection circle of radius r.
Obstacle Avoidance Behavior
Since no sensor can be guaranteed to always give a
Obstacle avoidance has been simulated with different correct signal, the conditional probability, p (r) <1, is
algorithms and the simplest has been to stop upon detecion, applied to determine if, given that a target lies inside the
backup turn right,go forward and check again. THis tends to region R with the nominal detection radius r from any
get trapped in complex obstacles but the forward sector vehicle, a detect signal is given.
avoidance shown in Figure 2 appears to execute quickly and
is robust to trapping.
5-39
p is an appropriate function of radial displacement,
although the "cookie cutter” model has been used in this
work to date. More representative distributions are easily
implemented. A detection signal is declared positive if a
uniformly distributed random number, : [0,1] is such that
rn <p.
RESULTS
Sensor Imperfection
Probability of Casualties
CONCLUSIONS
ACKNOWLEDGMENTS
REFERENCES
5-42
Unexploded Ordnance Clearance
and Minefield Countermeasures
by Multi-Agent, Small Robotics
Because a full paper was not received by publication date, the above Abstract appears in this
Proceedings. The authors can be reached at NAVEODTECHDIV, 2008 Stump Neck Road,
Indian Head, MD 20640; telephone 301-743-6850, X 281.
5-43
20 November 1996
BUGS
:3 ■c:
0
Technologies
c s CO p
.0
.9> _
CO 8
o ■J2
CO 0
§;
so CD
• •
CO
CQ <c CO
• • •
CD
CD i?
Terrain/Conditions
CQ
O
c:
}2
■SJ "o ^ ■o
-c CD "(B c
o .o CD
,CD 5 § t3
O § CD
.bJ
Operations
-2 CD
9- S -2
2 2a §
0) 'J;;;
S iS c: 5> o ^
:s> 2 <D :3
CD CD
cc 2 =5 2 CD
o Q OC Q ^
CD
•
o
CO
o
o
o
cc
CO
o c CO
CO
""I
O CD 03
CO
■O ■O ::3 CO O) 5 §
c: -2 I- ^ o .c: < CO
:3 iS -S CD c: D)
8
::3 Q) .0) •c
o
?0
c ^ §
m
t I I I CD I
i? -J
CD
•
QQ
'•’tiliikir'
W..r
91
.O o p)
O) CO is
o
CO O
c: S
o .P
-2
•9 N
-5
=3 ^
:3 O' 5
o >b
BUGS
■9-
c CD O
CD <1> CD
q: ^
■? O' a -2
O p o 2 2
CD qj D)
P > CD >2 -2
IS
Q
• • •
5
Model Based Simulation
a
CD
O
c:
o
Clearance Time
BUGS
CD O
CD CD
CD CD CD
CD
CO CO CO .O
-c
c: CD "^3
.CD
^ c: O
::3 .CD
CD o
Q: QQ Q K.
CO
CD
2 S
(1) CO
g CD CO
CD is
S
Q. 2^ O
^0 ^
•£ >
CD O 'cD
S -2
^CD
BUGS
O CO c]L
CD
CD ^ :3
o
CO ^ o c: o
52
p -J
S
CD
i5
o CD
c: c: I
.2^
c:
S o
0 3;^ §
a> , p
Ct ' '
I I
%
CO
CD O
::d CO
c
PQ CD
CO
■ System
.o
CO
CD
CQ
5=54
BUGS
5-56
Weight: 19 lb
c
.O
CO
.2>
CD
c
.O
•J2
^ S
^ S
(D CO .o
CO
CD ^
O) I
I £2
BUGS
•S
o CD
9 .o
o CJ
O O £ c
c
CM CD O LU
o • ft CL
CO o CO
CD CD
CL
CD
Ci
If'?
Q ' '
CD
BUGS
5-58
CD c:
CD
■t!
S
S
C o
o O
O Q)
o
s
Q. CD
S q:
CO ^
■Q O
3 CD
CO :5
Q '
Q. y CO
c:
Io
5
:d I
~C
P -2
c
a. CD
P-
Q
5
(D
s CO
O
CD
9
9- i ■S
CD
c:
CO
CD
oc
C
o i fi
O
5-60
- Start, Stop
- Pick-up, Drop
Mobility Motor Control
05
■S c
CO CD
CO n-
•2 P
Co
P -2
B t I
o .9^
QQ .2 ^CD
i CO
O •
Abstract - The C. S. Draper Laboratory, Inc. prototyping of UUV systems and subsequent at-sea test
(Draper) recently completed the at-sea test phase demonstrations of Navy mission concepts would allow the
of the Autonomous Minehunting and Mapping Navy to determine whether the concept should be taken to
Technologies (AMMT) Program for the Defense full scale development.
Advanced Research Projects Agency (DARPA). The first UUV was delivered for at-sea testing 19 months
The primary objective of this program is to after the start of the contract, and underwent preliminary
develop and demonstrate advanced minehunting
performance testing and evaluation. Since these initial sea
technologies that will enable Unmanned Undersea
trials, the vehicles have been modified and configured to
Vehicles (UUVs) to clandestinely survey an
undersea area for mines and collect data for post validate several Navy operational missions. The first
mission mapping of the surveyed area. The mission to be validated was the Tactical Acoustic System
survey data must be of sufficient quality t o (TAS), which was a classified mission and will not be
support selection of an amphibious operating discussed. The second mission demonstrated high data rate
area and subsequent neutralization of mine or underwater laser communications between an AUV and a
obstacle threats. manned submarine. The third mission was the Mine
As integration contractor for the AMMT Search System (MSS), which used a fiberoptic or acoustic
Program, Draper modified one of DARPA’s data link and demonstrated that a UUV could guide a surface
existing UUVs; which was previously designed ship or submarine through a minefield in a semi-
and built by Draper, and used for DARPA’s Mine
autonomous mode. In a fully autonomous mode, the MSS
Search System Program. State-of-the-art techno¬
vehicle performed a survey of an area and subsequently
logies in the areas of Sonar Mapping,
Navigation, Acoustic Communications, Imaging, transferred mine target data from the surveyed area to a host
and Mission Planning were incorporated into the via radio from a rendezvous point.
AMMT vehicle, resulting in a system having the In recent years, DARPA has responded to the priority
capability to perform an autonomous survey and need for mine countermeasures clandestine reconnaissance
meet program objectives. The vehicle was with the Autonomous Minehunting and Mapping
subsequently tested at-sea to demonstrate the Technologies (AMMT) Program. The AMMT Program is
advanced minehunting technologies and concepts. a follow-on effort to DARPA’s MSS Program and builds
This paper describes the develop-ment and significantly upon MSS achievements in five technology
integration of technologies required to perform
areas: Sonar Mapping, Inertial Navigation, Acoustic
the clandestine AMMT mission. Details of the
Communications, Undersea Imaging, and Mission
autonomous adaptive mission planner and the
Planning. These minehunting technologies were integrated
execution of the at-sea tests are presented.
into a modified MSS vehicle and underwent a five month
at-sea test program, which concluded in May of this year.
I. Background
The intent of the AMMT Program is to demonstrate the
successful integration of these minehunting technologies
C.S. Draper Laboratory, Inc. (Draper) has been
into an autonomous vehicle, and in so doing gain insight
developing Autonomous Undersea Vehicles (AUVs) and
and information which will prove beneficial to the Navy’s
associated vehicle subsystem technologies for the Defense
present off-board sensor programs; the Near Term Mine
Advanced Research Projects Agency (DARPA) for several
Reconnaissance System (NMRS) and Long Term Mine
years. The DARPA Unmanned Undersea Vehicle (UUV)
Reconnaissance System (LMRS), and also support other
Program began in May, 1988 when Draper was contracted
Navy UUV Program priorities.
to design, fabricate, assemble, test and deliver two UUVs.
The goal of the joint DARPA/Navy UUV Program was to
demonstrate that UUVs could meet specific Navy mission
requirements with emphasis on the use of state-of-the-art
technology and “rapid prototyping” of hardware. Rapid
5-63
11. Goals AND Objectives ni. System Development
5-64
acoustic modem’s two projectors are installed in the
vehicle’s aft free-flood section. The modem’s eight
receivers are mounted in the vehicle’s forward ffee-flood
section and the modem’s computer is installed in the
vehicle’s electronic section. The Doppler aided-inertial
navigation system components include a broadband
Doppler Sonar, located in the forward free flood area, and an
inertial reference unit, navigation computer and recorder
located in the vehicle’s payload section. Power to the
various subsystems is distributed from the vehicle’s 300
Kw-Hr silver-zinc battery. In the AMMT configuration the
vehicle is capable of diving to 460 m, and achieving speeds
of 2-7 knots.
The mapping sonar utilizes the MSS Ahead Looking
Sonar and performs additional functions of bathymetric
mapping and precise navigation. Computer aided detection
enhancements were added to the sonar data processing,
which also generates bottom relief profile maps. A
byproduct of the mapping process is the ability to improve
vehicle navigation using sonar data. The Acoustic
B. Vehicle Systems Tracking and Navigation (ATN) function of the mapping
sonar, which estimates vehicle position based on ping-to-
The AMMT vehicle, as shown in Fig. 2, is a modified ping correlation and tracking of bottom features, provides
version of the MSS vehicle. The 1.1 m diameter hull was input to the vehicle’s navigation filter for integration with
extended to an overall length of 12.5 m, by the addition of the navigation estimates from the inertial navigation
a new 0.7m titanium hull insert. The additional insert system.
volume houses two mapping processors for the sonar Acoustic communications at rates to 10 kbps over
subsystem. The basic MSS vehicle and control system ranges up to 10 km were goals for the program. The
remain intact along with the Ahead Looking Sonar acoustic communications system also had the additional
System. Mechanical mounts were added to the bow of the function of image management. Prior to transmission
vehicle to allow a bridle to be used for vehicle towing, and of the laser line scanner images and sonar maps topside,
to provide for attachment of a tow cable for submerged images and map data were compressed to reduce
sonar diagnostic towing. A GPS system is integrated into transmission time. Subsequent decompression and
the vehicle with the antenna added to the vehicle’s 1.1 m enhancement of the transmissions was provided in the host
erectable mast, also used for the radio antenna. The communications support equipment.
5-65
The primary imaging system on the AMMT vehicle is a
Laser Line Scanner System (LLSS) provided by Raytheon.
The system uses a 400 mw laser and has the ability to
provide sampling resolutions of 512, 1024, 2048, and 4096
pixels over a 70“ field of view with 14 bit amplitude
resolution. The laser sensor is mounted to the underside of
the vehicle hull in a faired, fiberglass pod. For the AMMT
Program application, the LLSS was designed to provide high
resolution images at altitudes of 11.3 - 14.0 m over the target
at speeds up to 6 knots. Because the LLSS had not been
previously used in autonomous operations and to mitigate the
potential risk that the system might not operate satisfactorily
in the vehicle, a second imaging system was integrated into
the vehicle. The backup system utilizes an electronic still
camera and strobe lights, which were mounted in the aft and
forward freeflood sections of the vehicle respectively.
Early DARPA LTUVs guidance and control system required Fig. 3. Actual Path (dark) vs. Desired Path (light).
operators to specify a pre-planned sequence of guidance
commands to control vehicle trajectories. This capability has
proven to be very reliable and robust and yet limited relative -modating ocean current. Note the effects of varying ocean
to contigent maneuvers. To be a truly autonomous system, current on the shape of the northerly and southerly turns.
the vehicle must be able to execute obstacle avoidance, terrain Fig. 4 shows the actual groundtrack for a survey, imaging
following, conditional target imaging, and optional path and depth excursion mission overlaid on the pre-mission
planning in real-time as well as simple high level input display. After launch, the vehicle went to the start
specification of mission objectives and constraints. The on point, conducted a survey over the desired region while
line mission planner achieves this goal via real-time avoiding known obstacles, perfonned imaging maneuvers,
assessment and trajectory planning and interfaces with the sampled the water column with a depth excursion and finally
UUV guidance and control system. ended the mission at the desired waypoint. The survey region
Planning is a search through the infinite space of possible
decisions to find that sequence of decisions, or plan, that best
achieves the given objective. A search algorithm generates
possible vehicle trajectories which are then scored using a
cost function. The cost function is a weighted combination of
the estimated resources (time and fuel) needed to perform each
activity in the mission and the estimated value of each
mission objective. The value of an objective is specified by
the user and is relative to other activities in the list. If an
activity is twice as important to complete as another activity,
then its value would be twice as much as the other. The
mission planner scales this value by the probability of
completing the objective in the mission. In this manner, a
near optimal vehicle trajectory is selected given the specified
goals and constraints.
Examples of mission planning capabilities are presented in
Figures 3 and 4. Both include surveys where the lane growth
was North - South (North on tope of page). Fig. 3 shows the
planned and actual path for a 150 foot lane spaced survey.
For planning purposes, the vehicle’s turn diameter was
limited to 500 feet. The planner generated detailed
maneuvers based on constant diameter turns while accom-
5-66
was well covered and all obstacles were avoided. The topside (host) acoustic modem’s array, amplifier, and
additional maneuvers or loops in the groundtrack occurred projector.
because the planner was trying to attain the proper vehicle In addition to the equipment mounted on the MA^ SeaCon,
heading at the desired location. These maneuvers are an a temporary shelter was constructed dockside to house the
artifact of how the planning problem was decomposed. For vehicle during assembly, disassembly, maintenance, repair,
instance, if the planning time horizon were longer, there and battery charging. The vehicle was rolled out of the
would be no need for additional maneuvers. shelter on its handling carts and then shore-craned onto a set
of support cradles dockside where the LESS pod was mounted
A Schedule and Milestones to the vehicle. The vehicle was then either crane launched or
transferred to the pedestals on the SeaCon for travel to the
The AMMT Program originally began in April 1993, as a test site and launch at-sea. Fig. 5 shows the vehicle in the
follow-on effort to DARPA’s MSS Program. The program maintenance shelter on it’s handling carts.
was stopped in October 1993, due to a congressional delay in In the water, the vehicle was handled with the use of
GFY 1994 funding, and restarted in July 1994. Following inflatable support craft, including a pontoon type inflatable
in-laboratory systems integration, mobilization for the boat developed by Draper, to control the vehicle during
program’s at-sea test phase at NSWC’s South Florida Testing surface tow operations. The vehicle could be surface towed
Facility began in December 1995. The at-sea test phase of from the dock to the test site using the pontoon boat; or
the program occurred over a five month period ending with carried to the test site on the pedestals onboard the SeaCon,
demobilization of the test site in mid-May 1996. During the and craned into the water, sea-state permitting. The vehicle
test program, many of the subsystem test demonstration was able to be handled using the pontoon boat in seas to sea-
milestones were successfully met. However, full mission state 3, and launched/recovered using the support ship crane in
demonstrations of the integrated AMMT system were not sea-states of 1 or less. Generally, the vehicle was towed
completed due to program schedule and cost considerations. either to or from dockside using the pontoon boat. After
being towed dockside, the vehicle was lifted to the pedestals
IV. System Test Preparations on the support ship using the support ship crane for
subsequent transfer to the shore cradles. Fig. 6 shows the
A. Test and Support Facilities vehicle in the water with the pontoon boat.
An exercise minefield was installed in the off-shore test
Potential test sites were identified from the Pacific area, to provide targets for the Ahead Looking Sonar.
Northwest (Nanoose and Dabob) to San Diego to New Approximately 40 bottom and moored test mines were
England waters to Florida. deployed in water depths ranging from 46-180 m. A pattern
The Naval Surface Warfare Center’s South Florida Testing of mine lines parallel to the shore was used to represent a
Facility was selected as the site for the AMMT Program’s at- mine barrier which might be used to deter an amphibious
sea tests and mission demonstrations. Lxx:ated in the Ft. assault from the sea. Special optical targets were also placed
Lauderdale area of Florida, the facility was selected for its on the ocean to evaluate the dimensional and constrast
proximity to the open ocean, and the suitable bottom features resolution of the vehicle’s imaging systems.
and water depths available in it’s test area. The size,
availability, and launch/recovery capability of NSWC’s
support ship, MA^ SeaCon, was also a major reason for
selection of the test site. The site is located on the Port
Everglades channel and permits rapid access to the open ocean
without the penalty of a long surface transit. The M/V
SeaCon is equipped with a 22 ton crane which is capable of
lifting the AMMT vehicle for launch and recovery operations
and a winch capable of supporting vehicle diagnostic tow
operations. The MA^ SeaCon is also large enough to permit
the AMMT Control Van, a standard 40 foot ISO container, to
be mounted on its deck along with pedestals for vehicle
support during transits to and from the test area. The Control
Van housed the control and tracking equipment to support the
AMMT vehicle and was the central monitoring location for
vehicle performance and data collection. The SeaCon also Fig. 5. UUV in Maintenance Shelter.
housed and deployed the V-Fin assembly which carried the
5-67
Fig. 6. UUV and Pontoon Boat.
Draper’s simulation facility provided support to the The Test Director is responsible for the conduct of the at-
program during laboratory vehicle integration and at-sea sea tests. He is supported by:
testing. All vehicle software, including the mission planner,
was run in the high fidelity hybrid simulation for verification 1) Ship’s Captain: Responsible for all support ship
and validation. Mission software for each individual test day activities, including safety and emergency response. Also,
was verified in the simulation facility before its use at-sea. In ship’s crew will be responsible for UUV handling and
addition, the facility was used to troubleshoot and resolve the associated ship maneuvering activities.
problems and anomalies, encountered during at-sea testing, by
recreating the actual test conditions in simulation and running 2) Test Coordinator: Every at-sea event has a test
the tactical software. coordinator who is responsible for the conduct and execution
of the specific event.
B. Test Plan Formulation
3) UUV Tracking Coordinator: Maintains range and
A very large effort was expended to formulate the AMMT bearing of UUV relative to the support platform.
At-Sea Master Test Plan. Once the objectives were finalized, Recommend maneuver for optimal system monitoring.
the proper sequence of events could be established, test
organizations created, program requirements identified and 4) UUV Support Computer Operator: Conduct pre-launch
executed, and Master Test Plan designed. and post-recovery checkout of UUV and payload. During test,
vehicle will be queried for status and subsequent interrogation
Support Groups of vehicle fault status. During operations, operator can
download mission activity changes as directed by the Test
The Test Working Group (TWG) consisted of the major Director.
system suppliers identified in the Test Organization. This
group had the responsibility to provide coordination in the 5) Sonar Coordinator: Responsible for sonar mapping
planning of the test program. The Chairman of the TWG system operations, monitoring of displays, and data collection
was the AMMT Test Director. Meeting’s were scheduled as during Diagnostic Tow Testing.
necessary to ensure that all test requirements were evaluated
and/or incorporated during engineering design review phases C Preliminary Requirements
and test plan development. The TWG was replaced by the
Joint Test Group (JTG) at the start of the at-sea testing. The AMMT Test Program included preliminary sequential
The JTG had the responsibility for overseeing the conduct phases followed by the at-sea test and demonstration. The
of the at-sea test program. They operated as the day-to-day following tasks had to be successfully undertaken prior to the
guidance, review, approval, and on-scene authority for at-sea at-sea testing phase:
operation. The head of the JTG was the Test Director.
1) Test and Evaluation and Factory Acceptance Tests
Command and Control Organizations (FATs) of GFE UUV prior to delivery to test site.
5-68
2) FATs conducted on the various subsystems to be conditions where there was eminent risk of vehicle loss. The
incorporated in the UUV, including the GFE mapping sonar plan was a stand-alone document that permitted the Test
system. Director to react based on specific, well defined, prevailing
conditions.
3) Testing and integration of AMMT sub-systems and The Master Test Plan identified the complete AMMT at-sea
stand-alone items. test when operating in a perfect, no-problem, no-failures
world. Of course, that did not take place. The JTG reviewed
4) Testing and integration of AMMT sub-systems in the each days activity based on the past and prevailing conditions
UUV. and modified the next event to take place as required. Test
Execution will identify the actual sequence of events.
Prior to the start of the at-sea tests, a number of program
requirements had to be generated to meet program goals and V. Test Execution
perform technology demonstrations.
The System Integrator had to provide a reliable fault After a Test Readiness Review (TRR) held in early
tolerant, integrated testbed vehicle. This required the January, 1996, AMMT was given the go ahead to proceed
engineering of Interface Control Documents (ICDs) for each with at-sea testing. The test program consisted of a sequence
subsystem incorporated into the vehicle, and execution of the of events, each built upon the other, culminating in
preliminary pre at-sea test tasks identified above. All autonomous missions of increasingly greater complexity and
program requirements were reliably met for timely challenge.
commencement of the AMMT at-sea test. The objective of the at-sea test program was to demonstrate
Numerous support equipment suites had to be designed, the utility of the AMMT technologies in conducting a
built, and incorporated into the vehicle system for proper clandestine reconnaissance full mission profile. Specifically,
execution of the AMMT tests. The vehicles Emergence the vehicle would transit some 25 nautical miles over-the-
Recovery System (ERS) was augmented with an Argos horizon, perform a minehunting and mapping survey over a
beacon and an inflatable enhanced radar target buoy. During broad area, review Computer Aided Detection results, then
autonomous operations, the vehicle’s activities and position revisit - for imaging purposes - the more mine-like of the
were shown on a real-time tactical display. During pre and detected objects. After imaging with the laser, a full
post test activities, methods to monitor and execute data resolution image would be stored for post-mission retrieval
transfer were formulated along with an external cooling while a second copy was compressed for immediate transfer
system. by the acoustic modem to the support ship along with
mapping products. A key assumption of this operational
D. Master Test Plan scenario and throughout the test program was that the vehicle
had no prior knowledge of its environment; it would have to
Based on objectives, requirements, system and budgetary rely upon the ahead looking sonar to provide real-time input
constraints, a Master Test Plan was formatted. Each days to the mission planner for such critical functions as obstacle
activity was planned from start to finish and presented in a avoidance and terrain following at low altitudes for imaging.
Gantt chart format. Consideration was given to maximum
use of UUV battery capacity between recharging and A. Preliminary Tests
minimizing night operation for safety reasons and weekend
operation because of the very high volume of large pleasure Vehicle handling equipment and procedures were first
craft traffic in the operating area. Each daily operation’s verified with a mock-up or dummy vehicle. Then came
Gantt chart identified the activity name, duration, start time, handling of the actual vehicle. A specially designed inflatable
support ship start and end time, and daylight hours. Each test catamaran could safely fasten itself to the bow of the vehicle
also identified: objectives, test method, test coordinator, test and still employ an outboard motor for local maneuvering and
documentation, data products and format, data distribution, control. The crews of the highly versatile support craft
data analysis, and other test support. This information and handled a number of chores such as attaching and removing
the Gantt chart provided sufficient information for the JTG to liftings slings from the crane and removing the surface tow
plan and re-plan activities on-site as conditions dictated. The bridle upon arrival at the dive point.
charts also presented the inter-relationship of activities. A deep berth on the Intracoastal Waterway was utilized for
The Master Test Plan also included a Safety, Search and calibration of doppler sonar bias parameters and for acoustic
Recovery (SSAR) Plan intended to assure a high state of interference tests. Draper Lab, as system integrator,
safety and recovery readiness during AMMT UUV sea trials. developed a Ping Management technique to ensure that the
The safety plan was intended to account for all situations acoustic modem and ahead looking sonar did not interfere with
ranging from normal vehicle testing conditions to these each other.
5-69
The first autonomous AMMT operation was a Preliminary meaningful mine identification test for the imaging
Performance Test with calibration of waterspeed versus equipment and image compression process. Special optical
propeller RPM. Hydrodynamic performance in the horizontal targets were also placed on the ocean bottom to evaluate the
and vertical plans was validated with the underhull imaging dimensional and contrast resolution of the imaging systems.
pod and the added length of the mapping processor hull insert. A revised sonar software release was down-loaded to the
Theoretical endurance with silver-zinc batteries was estimated vehicle’s mapping processors and a limited amount of
36 to 48 hours depending on speed. Diagnostic Tow was conducted which verified fully
During the first autonomous dive, initial insights into autonomous Computer Aided Detection of the mines. As the
acoustic modem performance were obtained. The modem was vehicle was towed over the exercise minefield, information
relied upon heavily, not only a development technology, but passed over the umbilical allowed one to observe the system
as an essential tracking and status tool for the untethered making target calls without operator intervention and noting
vehicle. Messages were transmitted from the vehicle at target locations near known positions of exercise mines.
operator-selected time intervals such as once per minute.
Detailed status content consisted of vehicle position, velocity, C. Autonomous Tests
attitude, keel depth, altitude above the bottom, estimates of
ocean current, and subsystem status including battery and Autonomous operation and testing of the mission planner
variable ballast. Data rates of 5 kilo-bits per second were came next. In a dense survey over a small area, the Mission
achieved with the modem and image processing software later Planner demonstrated the ability to solve and execute
demonstrated the ability to transmit compressed images using trajectories for lane spacing closer than the vehicle minimum
two methods; JPEG as a baseline and a wavelets-based turn diameter. Avoidance trajectories were demonstrated
method. Time to transmit the compressed images was about around obstacles manually entered into the pre-mission
4 and 2 minutes respectively. planner. At-sea performance showed close correlation with
The next item in the test sequence was calibration of the predictions based on Draper’s Hybrid Simulation. Surveys
inertial navigation subsystem. The vehicle was towed on the over larger areas permitted fully autonomous operation of the
surface 45 miles north along the coast from Fort Lauderdale. mapping sonar. Terrain following using sonar inputs was
Differential GPS fixes were periodically used to reset the briefly demonstrated. A “comb” survey, consisting of
inertial navigation position and to calibrate doppler scale shoreward probes along the coast at 1 nautical mile intervals,
factor and boresight misalignment - a one-time procedure for yielded rapid reconnaissance over a large area. The UUV
each vehicle installation. Upon completion of the demonstrated its inherent stealth for clandestine
calibration, navigation accuracy was evaluated during the reconnaissance when it surfaced for a GPS fix and le-
return tow south. Portions of the two were conducted on the submerged within 10 minutes. Successful test days resulted
surface and later submerged for concurrent inertial navigation in completion of back-to-back missions of 4 hours duration.
subsystem after 15 nautical miles was less than 10 yards and Each the single longest AMMT autonomous mission lasted 6
after 35 nautical miles was less than 50 yards. Refinements hours, covered 20 linear miles of survey, and mapped a 2 by
to the calibration process may provide further reduction in 2 nautical mile area with 100% overlap.
errors.
VI. Accomplishments
B. Diagnostic Tests
Significant accomplishments were achieved during the test
The final item remaining to be completed before further program. In addition to meeting the four major program
autonomous operation was Sonar Diagnostic Tow. requirements; provide a reliable, fault tolerant integrated
Investment was made in the submerged tow capability in testbed vehicle; develop support equipment suite; provide
order to connect Ethernet channels from the vehicle to the capability to launch, recover, perform diagnostic tow and
Control Van to permit sonar data recording at higher rates and service vehicle; and select test site and mobilize for conduct of
for longer periods than would be possible with vehicle tests; many program objectives were met and test
recorders. The Diagnostic Tow umbilical also provided real¬ demonstrations conducted. At-sea tests included subsystem
time insight into the behavior of the mapping sonar. demonstrations of the various advanced technologies.
Recording and real-time insight were invaluable for diagnostic However, full mission demonstrations of the program’s
testing in spite of some acoustic interference from the tow integrated minehunting and mapping capabilities were not
vessel. accomplished due to time and cost considerations.
A threat-representative exercise minefield was deployed to On-line adaptive mission planning was implemented, and
provide targets for the ahead looking sonar. About 40 bottom successful generation and execution of mission plans in
and moored mines were laid in water depths from 150 to 600 varying ocean environments was demonstrated. The vehicle
feet. Some mines were painted olive drab to provide a achieved trajectories and maintained safe operating conditions.
5-70
All preplanned mission planner activities were demonstrated for the water conditions. EESS images were obtained at non-
at-sea, except for data upload; with the planner’s contingent optimal altitudes during surface tow operations but they lack
activities demonstrated in simulation, as the program ended detail and recognizable objects.
before these activities could be executed at-sea. Originally, the at-sea test program was designed to include
The sonar system demonstrated the ability to perform real¬ an overall demonstration of the fully integrated vehicle
time mine detection and classification by detecting all conducting a clandestine reconnaissance mission profile,
deployed targets. The ability to autonomously map bottom utilizing all of the advanced technologies. Specifically, the
topography was demonstrated by generation of maps which vehicle would autonomously transit 25 nautical miles, over-
agreed with surveyed data. The ability to track and navigate, the-horizon, perform a minehunting and mapping survey of
using bottom features in the acoustic tracking navigation an area, revisit one or more of the previously detected
mode, was demonstrated over short intervals. However, the minelike objects, image the objects with the EESS, and
acoustic tracking navigation data was not integrated with the compress and acoustically transmit the image and mapping
Inertial navigation system data before the program ended. products to a host. The detailed maps and images would be
The navigation system demonstrated it can be operated in generated post-mission to support “quick-look” data products.
an autonomous mode and can perform erect and align Although pieces of this mission profile were demonstrated
sequencing. Test data indicated performance accuracies which individually, they were not demonstrated collectively in one
approached the desired goal of 0.02% of distance traveled, autonomous mission due to test schedule limitations.
exceeding state-of-the-art systems in use. The results are
based upon limited test data and require additional at-sea VIE Conclusions
testing to establish firm quantitative values.
The acoustic communications system demonstrated reliable The AMMT Program was conceived and executed in the
uplink communications from the vehicle to the host at data DARPA tradition of having aggressive goals with an
rates of 5 kbps up to ranges of 2 km. Uplink rates of 10 aggressive schedule. At-sea testing was terminated in May
kbps were demonstrated up to ranges of 700m. Downlink 1996 before all goals were completely met, because of
communications from the host to the vehicle, typically at 2.5 funding limitations and other commitments for the NSWC
kbps, were not demonstrated due to the thermal layer in the facility and support ship. At termination, the Program was
operating area and the geometries of the V-Fin positioning on proceeding toward completion of all goals. At the time, there
the host and the directional receiver patterns of the vehicle. were no known technical challenges which were considered
Image processing demonstrations showed the ability to insurmountable.
compress optical images by either of two methods, JPEG or The AMMT Program went a long way towards proving the
EPIC, and transmit the images acoustically to the host. The concepts that a complex autonomous vehicle, integrating
system was used to compress and transmit a sonar map post several state-of-the-art technologies, could perform a
mission but not in real-time, before the program ended. clandestine survey of an undersea area and transmit
Optical imaging was not successfully demonstrated during information in near real-time such that it could be used as an
AMMT Program testing. Although both the LESS and effective mine countermeasure asset. Autonomous real-time
camera were integrated into the vehicle and operated correctly mission planning was demonstrated, a first for UUVs.
producing images dockside in air, the LESS did not produce a Navigation accuracy improvement of a factor of 5 over known
suitable image at desired altitudes at-sea due to environmental system was validated. Additional testing would have received
conditions. Autonomous imaging runs were made at the list of demonstrated technologied achievements.
conservative altitudes of 15 m, which proved to be too great
5-71
5-72
GPS and Mine Warfare
James R. Clynch
Department of Oceanography, Naval Postgraduate School
Monterey, CA 94943
clynch@nps. navy .mil
the time of transmission ( received time minus the transmit time) by the
Abstract - The Global Positioning System (GPS) is now the standard
for navigation and precise positioning in the military and dvili^ speed of light. These are called pseudoranges. They are used to
worlds. There are many techniques for utilizing this system, all with generate a position. The phases are used in a like manor to generate a
velocity. In an advanced differential techniques call kinematics, these
different levels of accuracies and limitations. Techniques will be
reviewed that yield accuracies between 100 m and 2 cm for dynamic phases are used for cm level positioning.
platforms. Many of the latest techniques and systems are now
coming out of the civilian sector. The applications of these te<^ques It is the message data that generates the limitations on the GPS system
to mine warfare will be discussed, including subtle limitations. A for the military user. In order to use the ranges to generate a location,
new system under development to achieve 2 m absolute positions at the position of the satellites is needed. This is obtained from a simple
sea and serve as a mobile differential GPS base station will be orbit model and coefficients contained in the message. These
coefficients produce predictions of the satellite positions based on data
presented.
between 2 hrs and 26 hrs old. These coefficients are called the
Broadcast (BC) ephemeris. The inaccuracy in the BC ephemeris, and
I. Basic GPS Information Flow how the error grows with age, is the major error source for military
users.
In order to understand some of the capabilities and limitations of GPS
in any application, the flow of the information and the contribution of
various error sources to the position error needs to be understood. [1-3] It takes a minimum of data from 4 satellites to generate a GPS three
In Fig. 1 a schematic of the information flowing from the satellite to a dimensional solution and the receiver time bias. The users clock error
mobile user is shown. The phrase "effective transmissions" is used is determined at each solution time line. Modem GPS receivers utilize
because pulses and carriers are not actually present in the signal. They 6 to 12 satellites in a solution to reduce the error. Most receivers
produce solutions at a 1 Hz rate. Receivers that produce 10 Hz
are regenerated in the receiver.
solutions are available.
In Fig. 2 the latitude errors from two receivers sharing an antenna are
shown. The curve with the wide oscillations is taking data in the
civilian Standard Positioning Service (SPS) mode. The curve with much
lower errors is from a military receiver using the Precise Positioning
Service (PPS). Four hours of data are plotted here on a vertical scale
that spans 100 m. Both these curves may be important to the military
user in precision navigation applications.
Notice that the errors are really slightly slopping lines indicating a very
slow rate of change of these errors. However, if the error were
determined, they could be considered constant over an hour or so. This
GPS Satellites means that if a differential system used the PPS system, it would have
Effective Transmissions to transmit a very small amount of information at a rate measured in
units of bits per hour. The resulting error, due to the fuzz on the PPS
Figure 1 error lines is at the 30 cm per axis level. It should also be possible to
preload corrections into a system if the mission duration is only an hour
or so and still maintain this 30 cm error level.
5-73
10.0
UT Hours
Figure 2
Distance (km)
B. Differential GPS Accuracy GPS signals do not penetrate seawater. Only a few mm will effectively
block them. [13] Accordingly, for underwater vehicles an antenna must
The other lines on Fig. 3 are for differential systems. [2,3] In this case be placed above the surface in order to get a position. An antenna docs
a reference receiver is set over a known point and measures the errors not need to permanently be on the surface, a pop up of the antenna will
in each satellites signal on a real time basis. These errors, or do.
corrections, are transmitted by some means to the remote, often mobile,
user. The remote user corrects his measurements before he computes his The time the antenna needs to be up varies greatly with several factors.
location and velocity. There have been several generations of receivers First different vendors vary a great deal in the time to fix (solution). In
used in DGPS systems and are there are different processing techniques addition the past tracking history can be important. If the receiver
all leading to the variety of lines. The length of the lines on this graph thinks it knows the approximate range of the satellite and the Doppler
reflects specifications and may often be greater. offset frequency, the acquisition can be very rapid. This can be 2 to 5
seconds in receivers optimized for reacquisition. Receivers are available
The first generation of DGPS has and error of about 4 m. This that can get ranges from 4 or more satellites in about 10 sec if they have
generation is exemplified by the USCG system [6] for maritime use. It been tracking these satellites in the past hour of so.
5-75
Selective Availability
(Civilian System Only} Orbit
Ionosphere Atmosphere [
{Single Frequency) j / Sat. Clock
Components of GPS
Range IVIeasurement
Figure 4
In order to do a real time navigation solution, the receiver must have not
are not canceled. Multipath is the error generated when the signal enters
only the range measurements, but also the coefficients in the message
the antenna after bouncing off some object. In general the receiver
that determine the location of the satellites. It takes 30 seconds for this
cannot distinguish this signal from the direct path and a composite of the
data to be received. Therefore it often takes 40 - 60 s to get a first fix
two is measured.
on the best receivers if it has no ephemeris. However these ephemeris
can be preloaded or a periodic extended tracking period can be used to
The Selective Availability error ( SA ) is intentionally introduced to
acquire them. They arc nominally good for 6 hr.s.
degrade the position accuracy of SPS users. It is about 30 in and causes
the large oscillations in Fig. 1. It is not periodic however, following a
If real time navigation is not required, then only range data is needed.
complex path designed to be hard to filter out. The ionospheric error
In this case the solutions can be generated post mission using ephemeris
collected by another receiver. is normally not present in military receivers because they often use two
frequencies. The second frequency is there Just to remove this error.
However the Precision Lightweight GPS Receiver (PLGR) is a single
frequency receiver. The second frequency is unavailable to inexpensive
V. GPS Error Sources
civilian receivers. This has lead to most DGPS systems being signal
frequency systems. 3’he atmospheric error is about 2 m vertically and
The errors in position solutions using GPS are rooted in errors in the
more for paths at lower elevation angles.
measurements of ranges and phases of the satellite signals, in the
broadcast ephemeris, and in any differential corrections. The errors
ITie orbit and satellite clock error is the most important one for military
enter into the solutions with a multiplicative factor. [1-3] Assuming that
users. It is caused by the ageing of the information in the broadcast
the range measurement errors are about the same magnitude from all
satellites: ephemeris. These model positions of the satellite position are
predictions of where the satellite will be in the future. Data used for
these prediction can be from 2 to 26 hours old. Tlie current positions
Solution Error = DOP x Range Error
are always known at the control center in Colorado Springs, but the
once/day upload cycle only delivers aged data to the user. The error
Here the DOP, or Dilution of Precision, is the multiplicative factor. It
grows slowly however, and only a very low update rate would be
is generally between 1.5 and 12 with values less than 6 being
required to provide the military user with current information. Of
considered acceptable. DOP is basically a measure of how spread out
course this is effectively what is done in all DGPS systems.
the satellites are in the sky. If all the satellites are in one segment, the
DOP will be large, (There are other bad geometries. For example if
three satellites are along a great circle, only two are useful.)
VI. Shipbome Reference System
The errors in a range measurement are shown in Fig 4. The true range
The Naval Postgraduate School is currently developing a system to serve
and the receiver clock error are used in the solution. The receiver clock
as a differential base station on a ship. There are many problems that
error is a part of the solution at each time line allowing very inexpensive
a shipborne reference system (SRS) will encounter that are absent or can
oscillators to be used. All the range errors that contribute to solution
be avoided with land reference sites. The most obvious complication is
error are contained in the small sliver that has been expanded. Most of
that the antenna is moving. However there are other issues, such as the
these errors originate in the satellite or in the atmosphere. They are
inevitable multipath on a ship. A multisensor system has been proposed
essentially identical for close receivers and are removed in DGPS.
to solve all these problems. A block diagram of the SRS is shown in
Only the site specific errors, the receiver thermal noise and the multipath
Fig. 5.
j-76
The orientation might be obtainable at the required accuracy from the
A three antennas themselves. However multipath errors might alias into
Corrections this estimation. Therefore an inertial sensor has been added to the
to Users system. This sensor will also allow the system to coast over outages
such as passage under a bridge. The inertial sensor will also add
considerably to the tracking of the antenna motions over ship roll
periods. This inertial can be of modest accuracy, with error
characteristics on the order of 1 deg/hr.
The motion of the ship will be tracked using the phase data, which is
about 1000 times as accurate as the ranging data. The inertial will be
usefiil here in detecting cycle slips and in averaging over residual phase
multipath. Once a solution is initialized, the system should be able to
keep track of the motion of the antennas at a level below a wavelength,
that is in the 10 to 20 cm range.
The key item being averaged out is the orbit and clock error in the
broadcast ephemeris. Having to estimate individual orbit clock errors
means that the system cannot depend entirely on the GPS satellite clocks
to estimate user clock errors. To solve this problem an atomic oscillator
will drive all the receivers. It is not clear that the ship clock bias can be
obtained at the nsec level, but drifts should be well controlled and only
biases should remain.
The key to the system is a data processing unit that will process a very
large quantity of data in a batch mode to initialize the system. With
GPS Ship Reference System modern computers and large disks it will be possible to keep and analyze
one or two days of 1 sec data from multiple receivers. The use of
Figure 5 batch mode will made the detection of cycle slips and bridging data gaps
easier by a factor of 4 than a Kalman filter. Once initialized the system
should maintain the solution via kinematics (phase based solutions) at a
high precision, but with some much larger bias. The initialization
The system will consist of one to three GPS receivers. The multiple process does not need to stop with the availability of computer power
receivers will be used to average down the multipath. Therefore the today. It can be run every few hours and the new initialization will
antennas will have to be fairly widely separated, at least by 3 to 6 m ( serve as an integrity check on the ongoing solution.
one to two P-code chips ) to give independent multipath measurements.
Of course the measurements from these antennas will have to be brought A simple error model has been developed for this system. There are
together, and orientation information is needed for that. inputs for the amount of multipath, the GPS receiver noise level, and the
14 ,-
This system is currently under development at the Naval 8. Kelly,D.A. and et al., "Navigation Performance Analysis for the
Postgraduate School. It is planned to demonstrate a system at sea in FY EDGE Program", Proc. lON-GPS-95, Sept. 12-15, 1995, Palms
98. The work on the Shipborne Reference System is sponsored by the Springs, CA, pg 413.
Office of Naval Research.
9. Van Dierendonck,A.J., and P. Fenton, "Theory and Perfonnance of
Narrow Correlator Spacing in a GPS Receiver", Proc. National
VII. Conclusions Technical Meeting ION 92, Washington DC, 1992, p 115.
The accuracy levels available to mine warfare operators covers a large 10. Hatch,R. "The Synergism of GPS Code and Carrier
range from 10 m to 30 cm utilizing various techniques. Very good Measurements", Proc. of 3rd International Geodetic Symposium on
results are obtained with a PPS differential system which need only Satellite Doppler Positioning, Las Cruces NM, Feb. 1982.
transfer a few bytes per hour from the reference station to the users.
The sources of the errors and techniques to deal with them have been 11. Remondi, B.W. Using the Global Positioning System (GPS) Phase
discussed. Finally a system to obtain 1 to 2 m absolute positions on a Observable for Relative Geodesy: Modeling, Processing, and Results.
ship has been discussed. PhD Thesis, Center for Space Research, Univ, Texas at Austin, May
1984.
VIII. References
12. Neumann, J.B. et al. "Test Results from a New 2 cm Real Time
1. Milliken, R.J., and C. J. ZoIIer, ’’Principles of Operations of Kinematic GPS Positioning System", Proc. ION-GPS-96. Kansas City
NAVSTAR and System Characteristics," Global Positioning System, MO, Sept. 17-20, 1996, pg 883.
Papers in Navigation. Vol. /, Institute of Navigation, Washington DC.
1980. pg 3. 13. Kwak,S.H. et al. "An Experimental Investigation of GPS/INS
Integration for Small AUV Navigation", Proc. of 8th International
2. Kaplan, E. (Ed), Understanding GPS, Principles and Applications. Symposium on Unmanned, Untethred Submersible Technology. Durham
Artcch House, Boston, 1996 NH, Sept. 1995.
5-78
The Phoenix Autonomous Underwater Vehicle
Abstract. The Phoenix autonomous underwater collaboration with other scientists interested in either
vehicle (AUV) is a robot for student research in robot or virtual world. Repeated validation of
shallow-water sensing and control (Figure 1). Phoenix simulation extensions through real-world testing
is neutrally buoyant at 387 pounds (176 kg) with a hull remains essential. Details are provided on process
length of 12 feet (2.2 m). Multiple propellers, coordination, reactive behaviors, navigation, real-time
thrusters, plane surfaces and sonars make this robot sonar classification, path replanning around detected
highly controllable. The underwater environment obstacles, networking, sonar and hydrodynamics
provides numerous difficulties for robot builders: modeling, and distributable computer graphics
submerged hydrodynamics characteristics are complex rendering. Finally in-water experimental results are
and coupled in six spatial degrees of freedom, sonar is presented and evaluated.
problematic, visual ranges are short and power
endurance is limited. Numerous Phoenix contributions
include artificial intelligence (AI) implementations for
multisensor underwater navigation and a working
three-layer software architecture for control.
Specifically we have implemented the execution,
tactical and strategic levels of the Rational Behavior
Model (RBM) robot architecture. These three layers
correspond to hard-real-time reactive control,
soft-real-time sensor-based interaction, and long-term
planning respectively. Operational software
functionality is patterned after jobs performed by crew
members on naval ships. Results from simple missions
are now available.
In general, a critical bottleneck exists in AUV
design and development. It is tremendously difficult
to observe, communicate with and test underwater
robots because they operate in a remote and hazardous
environment where physical dynamics and sensing
modalities are counterintuitive. Simulation-based
design using an underwater virtual world has been a
crucial advantage permitting rapid development of
disparate software and hardware modules. A second
architecture for an underwater virtual world is also
presented which can comprehensively model all
necessary functional characteristics of the real world in Figure 1. Phoenix AUV testing in Moss Landing
real time. This virtual world is designed from the Harbor, California.
perspective of the robot, enabling realistic AUV
evaluation and testing in the laboratory. 3D real-time
graphics are our window into the virtual world, 1 INTRODUCTION
enabling multiple observers to visualize complex This work describes software architectures for an
interactions. autonomous underwater robot and for a corresponding
Networking considerations are crucial within and underwater virtual world, emphasizing the importance
outside the robot. A networked architecture enables of 3D real-time visualization in all aspects of the
multiple robot processes and multiple world design process. Recent work using the Phoenix AUV
components to operate collectively in real time. is notable for the successful implementation and
Networking also permits world-wide observation and integration of numerous software modules within
5-79
multiple software layers. The three-layer software Chapter Organization. Section 2 presents
architecture used is the Rational Behavior Model motivations for artificial intelligence (AI) approaches
(RBM), consisting of reactive real-time control in underwater robotics. Section 3 describes robot
(execution level), near-real-time sensor analysis and hardware for Phoenix. Sections 4 through 7 examine
operation (tactical level), and long-term mission the Rational Behavior Model (RBM) software
planning and mission control (strategic level) architecture, detailing the execution, tactical and
(Byrnes 96) (Marco 96b). In effect a higher robot strategic levels. Section 8 describes robot networking.
software layer also exists: an off-line mission assistant Sections 9 and 10 discuss virtual world design criteria
that uses rule-based constraints and means-ends and visualizing control algorithms. Section 11
analysis to help human supervisors specify mission presents AUV-virtual world communications which
details, followed by automatic generation of strategic permit real-time physics-based response in the
level source code. Results for simultaneous operation laboratory. Sections 12 and 13 discuss interactive 3D
of the three onboard robot software layers (running an computer graphics and sonar visualization. Section 14
autogenerated mission) have been verified by virtual evaluates experimental results. Section 15 points out
world rehearsal and in-water testing (Davis 96a, 96b). areas for future work. The chapter closes with
Theoretical development stresses a scalable conclusions, references, and pointers to a repository for
distributed network approach, interoperability between software and documentation.
models, physics-based reproduction of real-world
response, and compatibility with open systems 2 MOTIVATION
standards. Multiple component models are networked Untethered underwater robots are normally called
to provide interactive real-time response for robot and Autonomous Underwater Vehicles (AUVs), not
human users. Logical network connectivity of physical because they are intended to carry people but rather
interactions is provided using standard sockets and the because they are designed to intelligently and
IEEE standard Distributed Interactive Simulation independently convey sensors and payloads. AUVs
(DIS) protocol (IEEE 95). Implementation of the must accomplish complex tasks and diverse missions,
underwater virtual world and autonomous robot are all while maintaining stable physical control with six
tested using the actual Phoenix AUV (Figure 2). spatial degrees of freedom (i.e. posture, meaning
3D position plus 3D orientation).
The underwater environment is highly challenging.
Hydrodynamics forces are surprisingly cross-coupled
between various axes because of asymmetric vehicle
geometry and the nonlinear drag "added mass" of
water fluid carried along with moving vehicles. Active
sonar returns provide precise range but poor bearing
accuracy, and can be subject to frequent dropouts.
Sonar range maxima are highly frequency-dependent.
At moderate ranges (beyond several hundred meters)
sonar paths can bend significantly due to continuous
refraction from sound speed variation, which is caused
by changes in water temperature, salinity and pressure
(i.e. depth). Vision is only possible for short ranges
Figure 2. Phoenix AUV shown in test tank (tens of meters at best) and is often obscured if water
(Torsiello 94). is turbid. Underwater vision also requires powerful
lighting, which is an unacceptable power drain due to
In order support repeatability of our results, already-severe power and propulsion endurance
documentation and source code are available constraints. Laser sensors are usable to approximately
electronically (Brutzman 96b, 96c). Current work 100 m range and provide good range and bearing data,
includes model validation as well as adapting but remain expensive, hard to tune and subject to
hydrodynamics and controls coefficients for other turbidity interference. Typically little or no
submersibles. Ongoing work also includes making 3D communication with distant human supervisors is
graphics and networking compatible with the Virtual possible. When compared to indoor, ground, airborne
Reality Modeling Language (VRML 2.0), to permit or space environments, the underwater domain
Internet-portable rendering and interaction via any typically imposes the most restrictive physical control
computer connected to the World Wide Web. and sensor limitations upon a robot. Underwater robot
5-80
considerations remain pertinent as worst-case is performed piecemeal and incrementally. For
examples relative to other environments (Figure 3). example, a narrow problem might be identified as
suitable for solution by a particular AI paradigm and
• Complex Hydrodynamics then examined in great detail. Conjectures and
o coupled in six spatial degrees of freedom theories are used to create an implementation which is
o accompanying "added mass" of water tested by building a model or simulation specifically
o instability can be severe or fatal suited to the problem in question. Test success or
failure is used to interpret validity of conclusions.
• Sonar
o accurate ranges but bearings poor Unfortunately, integration of the design process or
o numerous nonlinear factors affect even final results into a working robot is often difficult
reverberation and attenuation or impossible. Lack of integrated testing prevents
o sonar path bending at long ranges due to complete verification of conclusions.
sound speed profile (SSP) effects AUV design must provide autonomy, stability and
reliability with little tolerance for error. Control
• Vision and Laser
o range limited by turbidity systems require particular attention since closed-form
o lighting requires excessive power solutions for many hydrodynamics control problems
are unknown. AI methodologies are thus essential for
• Endurance typically a few hours
numerous critical robot software components.
o limited power available
o constrains all other equipment Historically, the interaction complexity and emergent
behavior of multiple interacting AI processes has been
• Navigation
poorly understood, incompletely tested and difficult to
o ocean currents vary with time, location
o acoustic navigation requires calibrated formally specify (Shank 91). We are happy to report
prepositioned transponder field that these problems can be overcome. Our three-layer
robot software architecture, in combination with a
o GPS and inertial methods possible
physically and temporally realistic virtual world, has
• Conununications
enabled effective research, design and implementation
o tether is an unacceptable encumbrance
o acoustic limited in bandwidth, range of an autonomous underwater robot.
The charter of the Naval Postgraduate School
o optical extremely limited range_
(NPS) Center for AUV Research group is to support
graduate student thesis research. Certainly there is no
Figure 3. Environmental constraints for underwater robots
are severe.
shortage of problems that underwater robotics
researchers might work on. We believe that having a
clear and compelling objective is fundamentally
A large gap exists between the projections of theory important. Mission drives design. A well-defined goal
and the actual practice of underwater robot design. provides priorities that can be understood by a large
Despite numerous remotely operated vehicles (ROVs) research group, clear criteria for making difficult
and a rich field of autonomous robot research results, design tradeoffs, and a finish line: success metrics are
few complete AUVs exist and their capabilities are defined. We have chosen shallow-water minefield
limited. Cost, inaccessibility and scope of AUV design mapping as our driving application. At the 1995
restrict the number and reach of players involved. Symposium on Autonomous Vehicles for Mine
Interactions and interdependencies between hardware Countermeasures (MCM) (Bottoms 95), consensus was
and software problems are poorly understood. reached that all technical components exist which are
Equipment reliability and underwater electrical needed to build effective MCM AUVs. Our motivating
connections are constantly challenging. Testing is goal is to demonstrate such a vehicle. We intend to
difficult, tedious, infrequent and potentially hazardous. demonstrate that there are no fundamental technical
Meaningful evaluation of results is hampered by impediments to mapping shallow-water minefields
overall problem complexity, sensor inadequacies and using affordable underwater robots. We are
human inability to directly observe the robot in situ. integrating component technologies necessary for
Potential loss of an autonomous underwater robot is underwater autonomy in a working system, and are
considered intolerable due to tremendous investments making good progress toward reaching that goal.
in time and resources, the likelihood that any failure Related efforts. Over a dozen other research
will become catastrophic, and difficulty of underwater groups are active in underwater robotics. The
recovery. Massachusetts Institute of Technology (MIT) Sea
Underwater robot progress is slow and painstaking Grant has deployed several Odyssey-dsiss AUVs
for other reasons as well. By necessity most research notable for open-ocean and under-ice oceanographic
5-81
exploration leading to the possibility of autonomous meter and a depth pressure cell. Five rotational gyros
oceanographic sampling networks (AOSNs) mounted internally are used to measure angles and
(Curtin 93). The Florida Atlantic University (FAU) rates for roll, pitch and yaw respectively. Small
ocean engineering department has built a series of cross-body thruster tunnels were locally designed and
vehicles which include fuzzy logic controllers and built for the Phoenix AUV. An in-line bidirectional
special sensing techniques (Smith 94). The Woods propeller inside each thruster can provide up to 2 Ibf
Hole Oceanographic Institute (WHOI) Deep (8.9 N). Detailed schematics and specifications of all
Submergence Lab (DSL) has specialized in long-term Phoenix AUV hardware components are presented in
bottom monitoring, acoustic communications and (Torsiello 94).
remotely teleoperated task-level supervision of
manipulators (Sayers 96). An excellent introductory
text on underwater robot design and control is
(Yuh 95). Annual AUV technical symposia
are sponsored in alternate years by the
IEEE Oceanic Engineering Society (OES)
{http://auvibmLtamu.edu/oes) and the Autonomous
Undersea Systems Institute (AUSI)
{http://WWW. cdps. maine. edu/A US I).
Important problem domain for AL Despite many
handicaps, the numerous challenges of operating in the Figure 4. Exterior view of NFS Phoenix AUV.
underwater environment force designers to build robots
that are truly robust, autonomous, mobile and stable.
This fits well with a motivating philosophy of Hans
Moravec:
5-82
corresponding to strategic planning and tactical strategic level. This architectural relationship is
coordination. Sensors are also controlled via execution illustrated in Figure 6 (Holden 95).
level microprocessor-hardware interfaces, although
some sensor functions may be optionally commanded RBM Manned
by the intermediate tactical level, such as steering level Emphasis Submarine
individual sonar transducer heading motors during
classification. A
Strategic
Mission
Plan Logic
Commanding
Officer
Due to the large variety of critical tasks an
autonomous underwater robot must perform, a robust Sequencing Officer of Deck,
multilevel software architecture is essential. Tactical Behaviors Watch Officers
Underwater robot software architectures are a Hardware Watchstanders
Execution Control
particular challenge because they include a many of
the hardest problems in robotics, control and AI over
short, medium and long time scales. Figure 6. Rational Behavior Model (RBM)
Rational Behavior Model (RBM). The software software architecture (Holden 95).
architecture used by the Phoenix AUV is the Rational
Behavior Model (RBM) (Byrnes 93, 96). The Rational
Behavior Model (RBM) is a trilevel multiparadigm Human analogies are particularly useful for naval
software architecture for the control of autonomous officers working on this project who already know how
vehicles. Execution, tactical and strategic levels to drive ships, submarines and aircraft, since they
correspond roughly to direct interaction with vehicle provide a well-understood partitioning of duties and a
hardware and environment, intermediate clearly defined task lexicon. The naval analogies used
computational processing of symbolic goals, and here merely express common and essential robotics
high-level planning, respectively. The three levels of requirements using terminology familiar to the many
RBM correspond to levels of software abstraction officer students who have worked on Phoenix. This
which best match the functionality of associated tasks. approach permits them to intuitively apply at-sea
Temporal requirements range from hard-real-time experience and domain knowledge. The RBM
requirements at the execution level, where precise paradigm continues to serve well as a formal robot
control of vehicle sensors and propulsion is necessary architecture which scalably composes numerous
to prevent mission failure or vehicle damage, to critical processes having dissimilar temporal and
soft-real-time long-term planning at the strategic level. functional specifications.
RBM provides an overall structure for the large RBM three levels summarized. Execution level
variety of Phoenix AUV software components. A software integration includes physical device control,
particular advantage of RBM is that the three levels of sense-decide-act, reactive behaviors, connectivity, a
RBM can be informally compared to the mission script language, and stand-alone robustness in
watchstanding organization of a submarine crew (i.e. case of loss of higher levels. Tactical level software
a manned AUV). Watchstanders operating vehicle includes Officer of the Deck (OOD) coordination of
sensors, the propulsion plant and diving station parallel tactical processes, telemetry vector state
controls correspond to the execution level. Precise variable updates as a form of shared memory, sonar
real-time control is needed at this level. The Officer control, sonar analysis and classification, path
Of the Deck (OOD) is represented in the tactical level, planning, DiveTracker acoustic navigation,
carrying out Commanding Officer (CO) orders by DiveTracker acoustic communications,
sending individual commands capable of being carried DGPS/GPS/INS navigation, and fail-safe mission abort
out by watchstanders at the execution level. Due to the if strategic level commands are lost. Strategic level
diversity of tactical tasks and the complexity of some software integration includes cross-language message
orders from the CO, the OOD has assistants at the passing, linking dissimilar binary executables, and
tactical level to assist in their decomposition. These several functionally equivalent strategic level
departments (navigation, sonar, path replanner etc.) variations: missions prescribed by Prolog rules, static
permit the OOD to concentrate on sequencing and mission scripts or an off-line mission generation expert
coordinating overall vehicle operation rather than system. There are numerous three-level robot
exhaustively directing every detail. Finally the CO is architectures and many are similar to RBM.
responsible for mission generation and successful Operating Systems and Compilers. Interestingly
completion. CO tasks include mission-related enough, operating system and compiler considerations
planning and decision making, all performed at the have been most notable for their incompatibilities
rather than their power. Aside from multitasking and
5-83
interprcx^ess communications, we have not yet found it development of software in all three software levels,
necessary (or desirable) to take advantage of real-time independently and in concert, first in the virtual world
operating system constructs. The execution level and then in the real world.
resides on a GesPac 68030 under OS-9 written in Declaring that combined models create a virtual
Kernighan and Richie (K&R) C, a precursor to world rather than a simulation is not an overstatement.
ANSI C. The tactical and strategic levels currently From the robots perspective, the virtual world can
reside on the Voyager Sparc 5 laptop under Solaris effectively duplicate the real world if robot
Unix, written in ANSI C and Prolog respectively. hardware/software response is identical in each
Additionally, tactical and execution software can domain. In effect, this is a type of Turing test from the
identically compile under SGI Irix 5.3 Unix in robot’s perspective. Such a concept is controversial,
ANSI C. Compilation of single version source files perhaps especially among reactive behavior-based
across a variety operating system architectures and approaches which assume world models are
language variants is achieved through use of #ifdef unavoidably overcomplicated and use "the world is its
and Makefile constructs (Brutzman 96c). This own best model" (Brooks 86). In our case the
prevents "versionitis” or multiple file versions which challenges of the underwater environment eliminate
inevitably lead to programmer confusion, incompatible relying on world availability throughout robot
source code interoperability and wasted effort. We are development. Development of a virtual world
continuing this interoperability trend by porting to the architecture that can realistically support the robot
well-supported public domain compiler g++ architecture has produced a new paradigm for robot
(GNU ANSI C/C++). software development (Brutzman 92a, 93, 94).
Hierarchical versus reactive. Only a few years
ago, robot architecture designers seemed preoccupied 5 EXECUTION LEVEL
with bipolar arguments between hierarchical and Disaster and divergence. In 1994 the execution
reactive approaches. Hierarchical stereotypes included level was the only software which effectively existed
phrases like deliberative, symbolic, structured, inside the Phoenix AUV. A second networked version
"top down," goal-driven, explicit focus of attention, of execution level was adapted to run in conjunction
backward inferencing, world models, planning, search with developmental tactical routines and the
techniques, strictly defined goals, rigid, unresponsive underwater virtual world. A disastrous hydrogen
in unpredicted situations, computation-intensive, and explosion occurred in 1994 which required over a year
highly sophisticated performance. Reactive stereotypes to repair. During this reconstruction period many
included phrases like subsumptive, "bottom up," changes and enhancements were made to the AUV
sensor-driven, layered, forward inferencing, robust software. Unfortunately the two versions of execution
subsuming behaviors, avoid both dynamic planning level software grew far apart as they progressed, with
and world models, behave somewhat randomly, the in-water version emphasizing new hardware
succeed without massive computations using interfaces (Healey, Marco 95) and the virtual world
well-considered behaviors, difficulty scaling up, version emphasizing increased functionality
elusive stability and nondeterministic performance. (Brutzman 94).
RBM is a hybrid architecture that is hierarchical at the Two versions into one. The top priority for 1995
top layer, reactive at the bottom layer and a mixture in efforts was to merge the two different versions of the
between. Real-time responsiveness varies execution level. The in-water code was painstakingly
correspondingly at each level. From our experience reintegrated with the virtual world version, one
with Phoenix it appears clear that a three-layer hybrid function at a time. This approach permitted frequent
architecture is essential for a robot that must meet a testing in the virtual world as well as continuous
broad range of timing requirements. Similar execution level accessibility to other tactical level work
three-layer hybrid architectures now appear to be the which proceeded in parallel. Laboratory bench tests
norm for many mobile robots. were also conducted to ensure that software functions
World models. Numerous Phoenix AUV theses and controlled the proper hardware and direction of
source code implementations have been handicapped rotation of moving components was correct. A single
by inadequate end-to-end hardware and software version of the combined execution level source code
functionality within the vehicle. Such constraints are had to run on different computer architectures, using
common for AUVs. Availability of networked different compilers, and with different physical and
hydrodynamics and sonar models for integrated logical interfaces. The new source code also had to
simulation during robot development have been run identically in the real world and the virtual world,
invaluable for development of robot control all without error. This effort was successful (Burns 96)
algorithms. This approach has permitted realistic (Brutzman 96a).
5-84
Telemetry state vector The execution level runs in interface to sensor and hydrodynamics models when
a tight sense-decide-act loop and provides real-time operating in the virtual world.
control of vehicle sensors and effectors. Sensor data
HELP Provide keywords list
and effector orders are recorded in a telemetry state
vector. This state vector is updated at the closed loop WAIT # Wait/run for # seconds
repetition rate, typically 6-10 Hz. The state vector is WAITUNTIL # Wait/run until clock time
used for mission data recording, sharing critical
QUIT do not execute any more
parameters among all tactical processes, and providing
a data-passing communications mechanism which RPM # [##] Prop ordered rpm values
permits identical operation in the real world and the COURSE # Set new ordered course
virtual world (described later). State vector
TURN # Change ordered course #
parameters, message-passing semantics and relation to
flow of control are described in detail in RUDDER # Force rudder to # degrees
(Brutzman 94). DEPTH # Set new ordered depth
Vehicle control. As current AUV research
PLANES # Force planes to #
indicates, a great variety of control modes are possible
when controlling vehicle posture and movement. A THRUSTERS-ON Enable vertical and
lateral thruster control
primary goal for the execution level is to provide
robust open-loop and closed-loop control using NOTHRUSTER Disable thruster control
propellers, cross-body thrusters and fin surfaces. ROTATE # open loop rotation
Direct open-loop control of all these effectors is control
available, singly or in combination. Closed-loop NOROTATE disable open loop rotate
control is available for course, depth and position,
LATERAL # open loop lateral control
either in waypoint-follow mode or hover mode.
Waypoint-follow mode relies on propellers and plane
GPS-FIX Proceed to shallow depth,
surfaces, which works well while.transiting but poorly take GPS fix
when stationary. Hover mode relies on propellers for
GPS-FIX-COMPLETE Surface GPS fix complete
short-range longitudinal motion, and thrusters for
lateral/vertical/rotational motion. Hover mode allows GYRO-ERROR # Degrees of gyro error
[GYRO + ERROR = TRUE]
precise station keeping in position, heading and depth,
at least while dead-reckon position and ocean current LOCATION-LAB Vehicle is operating in
lab using virtual world.
set/drift estimates are accurate.
Mission script language. In keeping with our goal LOCATION-WATER Vehicle is operating in
water w/o virtual world.
to make vehicle control understandable, we have
implemented execution level functionality using a POSITION # ## [###] reset dead reckon
i.e. navigation fix.
series of script commands. Each command consists of
ORIENTATION # ## ### (phi, theta, psi)
a keyword followed by a variable number of
parameters. The mission script language controls POSTURE #a ttb #c #d #e #f
(X, y, z, phi, theta, psi)
operating modes and state flags in the execution level.
A subset of the mission script language appears in OCEANCURRENT #x #y [#z]
Figure 7. TRACE verbose print statements
Commands can originate from tactical level
STANDOFF # Change standoff distance
processes, a prepared mission script file or a human for WAYPOINT-FOLLOW,
operator. Each command is designed to be HOVER
unambiguous and readable either by the robot or by WAYPOINT #X #Y [#Z]
people. Prescripted missions and tactical
HOVER [#X #Y] [#Z] [#orientation]
communications are intelligible because they sound [#standoff-distance]
similar to OOD orders and ship control party
communications aboard ship. We believe this
Figure 7. Mission script language (from file
approach has general applicability for most AUVs.
mission, script.HELP) (Brutzman 94).
Another feature is text-to-speech conversion in the
virtual world, simplifying human monitoring of
mission progress. Overall execution level functionality 6 TACTICAL LEVEL
also includes plotting telemetry results, replaying Officer of the Deck (OOD) Coordination. Of the
recorded mission telemetry data, and acting as network three levels of the RBM architecture, the tactical level
5-85
was the last developed onboard Phoenix. Creation of The Phoenix is designed for precision navigation
an OOD module is crucial. The OOD controls the requiring position accuracy of 1 m. The standard
flow of information between other levels and within deviation of the position available from GPS is
the tactical level, yet cannot become overburdened by approximately 60 m, with DGPS being accurate within
unnecessary details. By forking parallel processes, the 2 m. The DiveTracker short baseline acoustic ranges
OOD creates several departments which are available have a geometry-dependent standard deviation within
to assist in processing commands and sensor data. 20 cm (with an occasional range out to 33 cm) which
Reuse of execution level functions and data structures can cause a transiting position uncertainty of 1-3 m.
reduces the amount of unique code needed by the Using raw positions results in fix-to-fix position
tactical level. A modular interface design permitted uncertainty, control chattering and hydrodynamic
the departments and OOD to be developed stability problems for Phoenix. Kalman filtering
simultaneously. Figure 8 shows interprocess corrects these difficulties.
communications (IPC) from OOD to strategic level, Kalman filtering is a method for recursively
execution level and other tactical level processes updating an estimate of the state of a system by
(Leonhardt 96). processing a succession of measurements. The
Phoenix implementation uses a model-based
movement estimator for state, combined with
measurements, to produce the most probable estimate
of the vehicle’s position. A discrete Kalman filter is
used to process measurements, and the use of acoustic
range data requires an Extended Kalman Filter mode
of operation due to the nonlinearity of range
measurements (Bachmann 96).
Accurate and efficient navigation from point to
point also requires the knowledge of the local ocean
currents to prevent undershooting the intercept course
towards the desired location. If a vehicle fix
determines that the vehicle is not where the motion
model predicts, then the likely causes are ocean current
or AUV speed/heading errors. Using a non-zero mean
movement model (where input vehicle speed is
Figure 8. Interprocess communications (IPC)
assumed truth) results in the filter solving for both an
(Campbell 96).
updated position data and estimates of ocean current.
Estimated ocean currents are actually the combined
Properly implementing IPC is crucial. Forked Unix sum of actual ocean current, errors in reported speed
processes have duplicate variable stores but do not and heading errors. The ocean current values
share memory. Thus state variable changes in the produced can thus change with the vehicle heading,
parent (OOD) and children processes (navigation, but the root mean squared value of the currents will
sonar, replanner) must be performed individually for converge to a steady state number. This number can
each process. We use standard Unix pipes for this be resolved to X/Y or set/drift (polar) components for
communication since the tactical level is always within dead reckoning use. As with most processes at the
a single processor (Stevens 95). BSD-compliant tactical level, the algorithmic basis for this approach is
sockets are used for communications to the execution similar to techniques used by human navigators.
level since that operates on a different processor (or By monitoring the difference between a motion
even on a different network). Separate communication model and measurements, the Kalman filter can
channels are used for updating state vectors and determine if it has possibly lost track or received a bad
exchanging orders/ acknowledgements. measurement. If the difference is briefly too high, then
Navigation. The navigation module is a parallel the measurement is ignored. If the difference is too
forked process of the tactical level. It uses an high for longer than 15 seconds, then it is assumed
Asynchronous Discrete Kalman Filter to filter GPS that the filter has lost track. Upon loss of track, the
satellite navigation data received from a Motorola tactical level is informed and the OOD surfaces to gain
8-channel GPS/DGPS unit and ranges received from a a GPS fix and reset the filter state and parameters.
commercial short baseline sonar range system This GPS-FIX procedure is designed to work equally
(DiveTracker). well in hover and waypoint control. Full navigator
details are in (McClarin 96) (Bachmann 96).
5-86
Real-time Sonar Classification. Real-time sonar and is treated as a spurious return. We have developed
classification and run-time collision avoidance are more robust imminent collision avoidance algorithms
essential for AUV autonomy and survivability. An off¬ independent of near-real-time sonar classification
line sonar classification expert system was originally using the second steerable sonar. Using multiple
written using the CLIPS expert system shell noninterfering sonars permits employing search
(Brutzman 92b, 92c). Successful development of rules techniques that are otherwise mutually exclusive when
was originally dependent on the support of the expert sharing a single sonar transducer head. The collision
system rule-matching engine. Once the expert system avoidance sonar (usually the ST725) is directly
was developed, translation to C was practical and the controlled by the execution level for reliability and
optimized sonar classifier is now capable of running in rapid response.
real time to meet robot sensing requirements Path Planning and Replanning. Path planning is
(Campbell 96). a tactical function. The strategic level contains the
The sonar module initializes sonar transducer commanding officer (CO) and controls the overall
parameters for maximum range scale, orientation mission plan. The CO decides (in general terms)
change step size and transmitter power settings. Three where the ship will operate. Meanwhile, achieving the
modes are available: "transit search," "sonar search," ordered track is the responsibility of the tactical level
and "rotate search." The transit search consists of a Officer of the Deck (OOD). To determine a safe route
60"" sonar scan in front of the AUV. This search is to the location the CO has requested, the OOD tells the
primarily conducted for collision avoidance. The other tactical-level replanning department the desired
two modes are conducted in a search area to detect, location and the ship's present position. The sonar
localize and classify any unknown objects. Sonar department (via the OOD) provides the replanning
search and rotate search are 360® searches. Sonar department with the current physical environment, i.e.
search is performed by mechanically rotating the sonar where all the "circled" obstacles are. The replanning
head, whereas rotate search is accomplished with the department takes this data and provides the OOD with
sonar head fixed while the full Phoenix body performs the best path to the CO's ordered location after adding
a 360® rotation. a safety distance around any obstacles. If a new
Sonar processing begins with filtering, obstacle is found by sonar while the ship is transiting,
thresholding and smoothing of the raw sonar data to the OOD will call upon the replanning department to
produce a return bearing and range. The returns are check the path. Replanning does not constantly
then fitted to line segments using parametric process data but rather is called when the OOD needs
regression. Line segments are started when a sliding it.
window locates four returns that form an acceptable As a final step, smooth motion planning algorithms
line. Points are subsequently added based on distance are applied to the output of the circle world path
from the line segment and whether the new resultant replanner in order to provide precise control of
line segment is acceptable. Completed line segments Phoenix and allow for rapid travel around obstacles
are then combined based on proximity and without slowing into hover mode (Brutzman 92c)
orientation. (Kanayama 95) (Leonhardt 96) (Davis 96a). Hover
To remove the directionality effects of sonar scan mode is inefficient when transiting waypoints, since it
rotation, comparison of line segments is performed by requires Phoenix to stop and maintain posture at a
first using the segment that is more clockwise relative given location. Given the turning radius of a vehicle,
to the AUV. Once objects and line segments are smooth motion planning allows the vehicle to go from
formed, heuristic rules are applied to classify the one point to another along a path that does not require
objects. The last part of the classification process is to the vehicle to perform instantaneous changes in
relay object information in a manner suitable for path direction. Thus the vehicle does not need to rotate in
planning purposes. A circle representation is used place when negotiating around obstacles. Replanner
with the center at the centroid of the object. details are in (Leonhardt 96). Figure 9 illustrates the
Particularly long line segments (i.e. walls) are end-to-end process of detecting, classifying, localizing
converted to a set of small adjacent circles. This and avoiding a sonar obstacle.
methodology works. Additional experimental results
are needed to ensure that system coefficients are
properly tuned for current Phoenix sonars.
Imminent collision avoidance is achieved with a
simple relative bearing and range check for all valid
returns that contribute to any line segment. If a return
does not contribute to a line segment it is not evaluated
5-87
• Symbolic computation only, contains mission-
independent doctrine predicates and current
mission guidance predicates
• No storage of internal vehicle or external world
state variables
• Rule-based implementation, incorporating rule
set, inference engine and working memory
(if required)
• Non-interruptible, not event driven
• Directs tactical level via asynchronous message
passing
• Messages may be either commands or queries
requiring Boolean responses
• Operates in discrete (Boolean) domain
independently of clock time
• Building blocks: goals
• Abstraction mechanism: goal decomposition
(backwards chaining) and rule partitioning
(forward chaining); both are based on
goal-driven reasoning_
INITIALIZE VEHICLE
Figure 9. Obstacle detection, classification, localization abort initialize: abort entire mission
and avoidance.
TRANSIT
waypoint process abort: abort entire mission
7 STRATEGIC LEVEL setpoints system failure: abort entire mission
I GPS failure: abort entire mission/ignore
Prolog. The RBM strategic level is typically \ Obstacle log failure: abort entire mission/ignore
5-88
Mission Generation Expert System. The strategic recovery phase (Byrnes 96), Since there may be
level can also take the form of a deterministic finite multiple phase sequence solutions for a mission, each
automata (DFA). A mission controller initiates the solution generated by the system is the next solution
phase associated with the current DFA node upon found as opposed to an optimal solution. In addition,
arrival, transitioning to a new node when the current missions generated through means-ends analysis are
node’s phase completes successfully (or aborts because linear and proceed phase-by-phase to the end. In any
of a time out). A representative mission phase case, users are allowed to choose among the candidate
template appears in Figure 12. Individual tactic solutions generated (Davis 96a, 96b).
predecessors and successors can be composed using More complicated missions can take full advantage
this template to create missions of arbitrary complexity of this strategic level DFA structure. They are
(Davis 96a, 96b). specified phase-by-phase using the second piece of the
Mission Generation Expert System, the mission
specification tool. This tool allows an experienced
ii user who understands the DFA structure of the
e„pppcefni double boxes are
oredecessor composite templates strategic level to define missions one phase at a time.
label □_ n _
Regardless of whether the mission planning tool or the
mission specification tool is used, the system
automatically checks input for correctness and logic
tactic/strategy name and will not allow specification of an invalid mission
_K , .
,, =)
1
predecessor
parameter values (if any)
\ failure
(Leonhardt 96) (Holden 95) (Davis 96a, 96b).
The final aspect of this system is the code
generation facility. By using specified phases, either
the mission planning or mission specification tool, and
exception templates for valid phase types (e.g. hover, search etc.)
success (not used)
the system can generate executable code in either
Prolog or C++. Earlier theses demonstrated that the
strategic level can be equivalently instantiated using
Figure 12. Template for tactic and strategy composition. either the Prolog backwards chaining engine or the
CLIPS forward chaining engine. Alternate languages
Advantages of the strategic level DFA structure are are possible because there are multiple ways to plan.
twofold. First, an arbitrary mission can be modeled Backwards chaining can be unambiguously
simply as a set of phases that are executed in an order implemented using forward chaining, forward
defined by the transitions of the DFA. Second, chaining can be unambiguously implemented using
mission control using the Prolog search engine is backwards chaining, and both can be implemented
powerful enough that complex behavior can be using fully enumerated decision graphs. Use of C++
implemented without needing computationally has become possible because improved understanding
intensive mathematical calculations. Arithmetic is and tighter constraints on mission primitives has
confined to the tactical level, conceptual mission eliminated the need for the full functionality of the
planning is confined to the strategic level. Prolog search engine. Nevertheless such
Since a prime motivation for Phoenix is shallow simplifications were only possible following extended
water counter-mine operations, the mission generation experimentation using Prolog code.
process must be substantially simpler than writing Extensive testing of autogenerated Prolog and C++
Prolog programs if typical human operators are to code has been conducted in the virtual world, and
deploy the AUV. One solution to this problem successful in-water testing has been conducted at the
combines a graphical user interface for mission Phoenix AUV test tank. Moss Landing Harbor and the
planning and specification together with a goal-driven NPS swimming pool. Further in-water tests are
expert system for strategic level code generation. planned. Accomplishing our goal of simplifying
There are three aspects to the AUV Mission mission generation is indicated by a significant
Generation Expert System. The first is a mission reduction in the time required for mission coding
planning tool, which specifies vehicle launch and (minutes when using the expert system as opposed to
recovery positions and what the mission is supposed to hours without it). Finally, syntactic programming
accomplish. Means-ends analysis then computes a errors have been completely eliminated by the source
sequence of phases which can accomplish the desired code autogeneration system and logical programming
mission. Failure of any single phase will cause a errors have been substantially reduced.
mission to either abort or follow an alternate failure-
5-89
8 ROBOT NETWORKING
Perhaps surprisingly for a small robot, networking
is a major consideration. Within the Phoenix AUV is
an Internet-connectable local-area network (LAN).
This enables network communications between and
within the three software levels, external connectivity
in laboratory via tether cable, and (optionally) external
connectivity during harbor testing. Remote connection
of the LAN to the campus Internet backbone is
achieved using multiple wireless bridge boxes.
Multicast Backbone (MBone) connectivity permits
local or world-wide transmission of audio, video and
DIS streams (Macedonia 94). World Wide Web links
to online software documentation, multiple research
group accounts and properly networked LANs with
group access around campus further strengthened this
software development collaboration. Ease of use and
remote access translate into significant productivity
gains and regular discovery of new capabilities. We
expect to someday extend this approach underwater by
developing Internet Protocol over Sea Water (IP/SW)
connectivity (Brutzman 95a). Other network
considerations are elaborated in Section 11 as part of
virtual world connectivity.
9 VIRTUAL WORLD
The harsh environment in which an AUV must
operate calls for extra precautions in its design to
prevent damage to or loss of the vehicle. We have
developed a medium-scale virtual environment which Figure 13. Underwater virtual world for an AUV
(Brutzman 94).
enables meaningful end-to-end testing of robot
software and hardware in the laboratory (Figure 13),
As noted in earlier work on the virtual world: The objective of the underwater virtual world is to
reproduce real-world robot behavior with complete
It is tremendously difficult to observe, fidelity in the laboratory. Many questions pertain.
communicate with and test underwater robots, What is the software architecture required to build an
because they operate in a remote and hazardous underwater virtual world for an autonomous
environment where physical dynamics and underwater vehicle? How can an underwater robot be
sensing modalities are counterintuitive. An connected to a virtual world so seamlessly that
underwater virtual world can comprehensively operation in the real world or a virtual world is
model all necessary functional characteristics of transparent to the robot? How can 3D real-time
the real world in real time. This virtual world interactive computer graphics support wide-scale
is designed from the perspective of the robot, general access to virtual worlds? Specifically, how can
enabling realistic AUV evaluation and testing computer graphics be used to build windows into an
in the laboratory. 3D real-time graphics are underwater virtual world that are responsive, accurate,
our window into the virtual world, enabling distributable, represent objects in openly standardized
multiple observers to visualize complex formats, and provide portability to multiple computer
interactions. A networked architecture enables architectures? Overview answers to these questions
multiple world components to operate are provided here. Detailed analyses and example
collectively in real time, and also permits solutions are presented in (Brutzman 94). In effect,
world-wide observation and collaboration with the virtual world requires a separate software
other scientists interested in the robot and architecture for networked world models that
virtual world. (Brutzman 94) complements the robot software architecture.
The real world is a big place. Virtual worlds must
similarly be comprehensive and diverse if they are to
5-90
permit credible reproductions of real-world behavior. planes/rudders/propellers simultaneously when such
A variety of software components have been shown operation does not provoke mutual interference. Most
necessary. In every case, 3D real-time visualization Phoenix control code has been developed and tested in
has been a crucial tool in developing AUV software. conjunction with the construction of a real-time six
Ways to scale up and arbitrarily extend the underwater degree-of-freedom hydrodynamics model. Design,
virtual world to include very large numbers of users, tuning and optimization of control algorithms in
models and information resources are also isolation and in concert is the subject of active research
incorporated in this work. (Healey 93, 96) (Fossen 94) (Marco 96a) and remains
Virtual world capabilities were utilized for testing an important area for future work. Control algorithm
and verification throughout the software development robustness is a particularly important topic since
process. Use of this tool allows a number of potentially fatal nonlinear instabilities are possible and
programmers to work independently and in concert. vehicle reliability is paramount.
Virtual world capabilities have been incrementally Typical efforts at hydrodynamic development are
improved to match increased vehicle software based on mental interpretation of multiple time-series
capabilities, such as hydrodynamics and controller such as Figure 15. Dozens of two-dimensional
response rendering (Figure 14). Scientific time-series plots are necessary for quantitative
visualization techniques have provided further performance analysis, but this approach remains
significant benefits (Brutzman 95b). notoriously difficult to use when attempting to
mentally integrate and visualize all aspects of vehicle
behavior. The successes of individual control
algorithms created as part of this effort were highly
dependent on 2D and 3D visualization techniques.
Complete derivations of the full hydrodynamics model
and corresponding control equations are in
(Brutzman 94, 96c).
5-91
string types, starting with a predefined keyword and
followed by entries which may optionally have
significance depending on the initial keyword.
Messages with unrecognized keywords are treated as
comments. These two kinds of messages (telemetry
and commands) can be used for any communication
necessary among robot-related entities. Employment
of string types facilitates data transfer between
different architectures, data transfer via network
sockets, and file storage. String types also ensure that
all communications are readable by both robot and
human, a trait that is particularly useful during
debugging. An open format for command messages
permits any user or new application to communicate
Figure 16. Evaluating control response while broaching.
with little difficulty.
Within the AUV, the basic communications flow
between execution level and tactical level is
straightforward. All telemetry vectors are sent from
the execution level to the tactical level, providing a
steady stream of time-sensitive, rapidly updated
information. The tactical level may send commands to
the execution level as desired, and the execution level
may return informational messages between telemetry
vectors as appropriate. Nonadaptive tactical level
functionality can also be provided by carrying out
prescripted mission command files. Telemetry vector
records and command messages are logged in separate
mission output files for post-mission analysis and
replay.
The telemetry vector serves several essential
purposes. In addition to providing a steady stream of
Figure 17. Evaluating control response after broaching.
information from the execution level to the tactical
level, the telemetry vector also serves as the data
11 AUV-VIRTUAL WORLD transfer mechanism between execution level and
COMMUNICATIONS virtual world. Efficient communications between robot
Since RBM is a multilevel architecture, and virtual world are essential if rapid real-time 10 Hz
communications between levels must be formally robot response is to be maintained. The telemetry
defined. Communications between robot and virtual record is a concise and complete way to support all of
world must also be clearly specified. Defining these data communications requirements. Figure 18
communications includes establishing a physical path shows in detail how the flow of control proceeds and
for data transfer as well as defining the syntax and the telemetry vector is modified during each
protocol of exchanged messages. Our design sense-decide-act cycle.
objectives include reliability and clarity so that Robot execution software is designed to operate
messages are easily created and easily understood, both in the virtual world and in the real world. While
either by software processes or by people. Details sensing in the virtual world, distributed hydrodynamics
follow in order to illustrate the precise relationships and sonar models fill in pertinent telemetry vector
between robot, virtual world and graphics-based user slots. While sensing in the real world, actual sensors
viewing windows. and their corresponding interfaces fill in pertinent
Two kinds of messages are defined for use between telemetry vector slots. In either case, the remainder of
robot and virtual world. The first is the telemetry the robot execution program which deals with tactical
vector, which is a list of all vehicle state variables communications, command parsing, dynamic control,
pertinent to hydrodynamic and sensor control. interpretation etc. is unaffected. While operating in
Telemetry vectors are passed as a string type. The the virtual world, robot propulsion and sensor
second kind of messages allowed are free-format commands are communicated via the same telemetry
commands. Free-format command messages are also vector. While operating in the real world, robot
5=92
propulsion and sensor commands are sent directly to
auv posture: 1 ody-frame 'orld-frame ordered ordered sensed
hardware interfaces for propellers, thrusters, planes, telemetry time position, orientation velocities velocities udders, planes, sonar sonar
vector propellere. thrusters bearings values
rudders, sonar steering motors, etc. Again almost all Jj x.y.z.4>.9.4'. u v.w.pfl.r, X
_1
□
ordered values are changed by tactical level orders, mission script & execution level control
Act 7
execution level updates clock each timeslep and sends orders to virtual world or hardware
12 INTERACTIVE 3D GRAPHICS
Several important requirements are needed for the
creation of object-oriented graphics viewers for
visualizing a large-scale virtual world. Open
standards, portability and versatility are emphasized
over platform-specific performance considerations in
order to support scaling up to very large numbers of
users, platform types and information sources. The
Openinventor graphics toolkit and scene description
language has all of the functionality needed. The
potential integration of network connections to
logically extend graphics programs is also examined.
Open standards, portability and versatility are
emphasized over platform-specific performance
considerations in order to support scaling up to very
Figure 18. Data flow via the telemetry vector during each large numbers of users, platform types and information
sense-decide-act cycle. sources.
A good graphics toolkit for building a virtual world
viewer has many requirements to fill
The telemetry vector is therefore a key data transfer (Foley, van Dam 90). Rendered scenes need to be
mechanism. Telemetry vector updates also define the realistic, rapidly rendered, permit user interaction, and
communication protocol between execution level and capable of running on both low end and high end
virtual world. As might be expected, this works well workstations. Graphics programmers must have a
because the execution level program follows the wide range of tools to permit interactive
common robotics cyclic paradigm of sense-decide-act. experimentation and scientific visualization of
Figure 19 provides an overview of the telemetry vector real-world datasets (Thalmann 90). The ability to read
update sequence as an alternate means of portraying multiple data formats is also important when using
the validity of this approach. Given the perhaps-worst- scientific and oceanographic datasets. Scientific data
case computational complexity of underwater world format compatibility can be provided by a number of
models, this networked virtual world software data function libraries which are open, portable,
architecture for real-time performance in the reasonably well standardized and usually independent
laboratory also appears applicable to other robot of graphics tools (Fortner 92). Viewer programs need
domains. to be capable of examining high-bandwidth
information streams and large archived scientific
databases. Thus the ability to preprocess massive
datasets into useful, storable, retrievable graphics
objects will be particularly important as we attempt to
5-93
scale up to meet the sophistication and detail of the However, sound waves can be bent by variations in
real world. Adequate standardization of computer depth, temperature and salinity, A variety of problems
graphics and portability across other platforms is also including ambient noise, multipath arrival, fading,
desirable but has been historically elusive. shadow layers, masking and other effects can make
Openinventor is an object-oriented 3D graphics sonar use difficult. Since active sonar typically
toolkit for graphics applications design (Strauss 92). provides good range values with approximate bearing
Based on the Open GL graphics library, Openinventor values, algorithms for sonar recognition are much
provides high-level extensions to the C+H- (or C) different than vision algorithms. In the short sonar
programming language and a scene description ranges used by Phoenix, simple error probabilities and
language. It is designed to permit graphics linear geometric sonar relationships are adequate.
programmers to focus on what to draw rather that how Figure 20 shows the perspective gained by observing
to draw it, creating scene objects that are collected in AUV sonar from an "over the shoulder" perspective,
a scene database for viewpoint-independent rendering. one of several vantage points needed when developing
The ability to store graphics objects as readable, sonar classification algorithms.
editable files is especially appealing for the creation of
large-scale virtual worlds. Since the performance of
computer graphics is highly dependent on the
computational complexity of scenes to be rendered, it
is inevitable that truly large-scale world scene
databases will eventually overload viewing graphics
workstations. Such overload will occur regardless of
the efficiency of viewpoint culling algorithms and
graphics pipeline optimizations, unless partitionable
and networked scene databases are used. Furthermore,
since populating a virtual world is a task that needs to
be open and accessible to large numbers of people, an
open graphics data standard is needed for virtual world
construction. The ability to selectively load graphics
objects and scenes from files is an important
distribution mechanism which can take advantage of Figure 20. Local viewpoint of active sonar in test tank.
Web connectivity.
Ubiquitous portability for analytic, hypermedia, Since sonar is the most effective detection sensor
network, multicast and graphics tools is therefore an used by underwater vehicles, sonar visualization is
essential feature for virtual world model builders. A particularly important when designing and evaluating
superior alternative is now available using the Virtual robot software. Sonar parameters pertinent to
Reality Modeling Language (VRML) specification visualization and rendering include sound speed
(Carey 96). VRML is the Web standard for interactive profile (SSP), highly-variable sound wave path
3D representation. VRML scene description files are propagation, and sound pressure level (SPL)
the best approach for object definitions in a large-scale attenuation. Several questions are prominent.
virtual world (Brutzman 96d). How can a general sonar model be networked to
provide real-time response despite high computational
13 SONAR VISUALIZATION complexity? How can scientific visualization
Sensor differences distinguish underwater robots techniques be applied to outputs of the sonar model to
from ground, air and space-based robots. Since the render numerous interacting physical effects varying in
oceans are generally opaque to visible light at three spatial dimensions and time? Initial
moderate-to-long ranges, vision-based video systems investigations indicate that this area may yield
are ordinarily of use only at short distances and are significant results. The high dimensionality of sonar
unreliable in turbid water. Vision systems also usually data is best served by scientific visualization
require intense light sources which deplete precious techniques.
energy reserves. In comparison to underwater Sonar sensing is crucially important (Stewart 92).
computer vision, active and passive sonar (acoustic Previously only a single geometric sonar model was
detection) has long been a preferred sensing method available for Phoenix, derived by hand to model the
due to the long propagation ranges of sound waves AUV test tank (Figure 21). Although effective in a
underwater. small regular volume, this approach was too limited
and did not permit easy addition of artificial targets or
5-94
obstacles. We adapted the computational geometry of both hydrodynamics and sonar models in the virtual
routines included in the Openinventor interactive 3D world. Recent results include precise vehicle
graphics library to shoot rays into the scene database to maneuvering and rendezvous with a docking tube
produce a general geometric sonar model. Now the (Davis 96a, 96b) (Figure 23). Much more
same scene database (made up of Openinventor and experimental testing awaits.
VRML files) can be used for both virtual world
visualization and real-time 3D sonar ray intersection
calculations (Figure 22) (Davis 96a) (Brutzman 96b).
Figure 21. Manually derived geometric sonar model for 15 FUTURE WORK
AUV test tank (Brutzman 94). An underwater vehicle which can transit through
waypoints and hover in the presence of currents
enables a variety of capabilities which are not possible
14 EXPERIMENTAL TEST RESULTS for vehicles that must retain forward way to remain
Once Phoenix functionality was correct in the hydrodynamically stable. We intend to examine
virtual world, test tank experiments were conducted to whether the Phoenix hull form can stably approach
fine tune hardware and properly move the AUV and neutralize a moored mine-like object. Figure 24 is
through the water. Diving, forward, backward, lateral a notional diagram that shows how sonar can be used
and rotational movement checks were all performed to carefully approach a target broadside, keep station
during these test tank experiments. However, the against the ocean current, take confirming video, and
calibration of speeds during these movements could attach a beacon or neutralizing device using a simple
not be tested due to the relatively small size of the test one- or two-degree-of-freedom effector. For low sea
tank (6m x 6m x 2m deep). states, we see few limiting factors in this approach.
The next vehicle tests were performed in the
relatively calm sea water harbor in Moss Landing
California. A variety of logistical problems were
overcome but a seemingly endless series of minor
hardware failures then thwarted each attempt to run a
complete minefield search. Although a complete
mission was never accomplished beginning to end, all
components of the mission were individually exercised.
We now believe that the functionality and logic of the
AUV software is correct (Brutzman 96b). Remaining
tests include repeated mission testing, verification of
aggregate software behavior under a variety of
scenarios, tuning of control constants, and validation
5-95
16 CONCLUSIONS
The underwater environment is extremely
challenging for robots. Counterintuitive
hydrodynamics response, poor visual capabilities,
complex sonar interactions, communications
inaccessibility and power endurance are significant
design constraints. Robot builders must provide stable
control and reliable operation at all times due to the
unacceptably high cost of failure. A variety of AI
processes must be used few planning, sensing and other
complex tasks.
Systems integration is significant due to the many
sensOTs and effectors required for nontrivial operation.
The Phoenix AUV demonstrates that a three-layer
rdx)t architecture can be effective at combined system
Figure 23, A mobile stable AUV might precisely place an
control over time scales ranging from hard-real-time
explosive charge on an underwater mine.
sense-decide-act response to temporally unconstrained
mission planning.
Using an underwater virtual world for interactive
Phoenix is only directly controllable in five degrees 3D graphics rendering is an essential capability for
of freedom since roll is unconstrained. Pitch effective AUV development. The networked software
stabilization is straightforward using vertical thrusters. architecture and various results described here
Testing will determine whether roll stabilization is also demonstrate that a real-time physically based
necessary, p^haps by using an additional thruster. We underwater virtual world is feasible. It enables
are further interested in development of automatic repeated testing of all aspects of underwater vehicle
diagnostics that reconfigure control algorithms to control, stability, sensing, autonomy and reliability.
handle equipment faults. We also intend to explore Graphics viewer requirements include scientific
local measurement of cross-body ocean current flow visualization and portability across multiple platforms.
using acoustic doppler current profilers (ADCPs), in The use of multicast DIS messages, Web access and
order to permit precise maneuvering in the midst of VRML scene descriptions that include dynamic
highly varying flow fields and high sea states. Finally, behaviors promise the possibility of scaling to very
future work on underwater virtual world network^ large numbers of participants. Network connectivity
graphics includes compatibility with common Web allows us to use the global Internet as a direct
browsers using the Virtual Reality Modeling Language extension of our desktop computers, permitting global
(VRML) (Brutzman 96d). collaboration on a routine basis.
After years of effort, the RBM architecture is fully
instantiated onboard the Phoenix AUV and is being
successfully tested and refined by in-water testing. A
networked underwater virtual world has been crucial
to this development project. Experimental results
indicate we are close to demonstrating that affordable
underwater robots can operate autonomously in
challenging environments.
5-96
17 REFERENCES
Bachmann, E.R., McGhee, R.B., Whalen, R.H,, Steven, Brutzman, Don and Reimers, Stephen, "Internet Protocol
R., Walker, R.G., Clynch, J.R., Healey, A.J. Yun, over Seawater (IP/SW): Towards Interoperable
X.P., "Evaluation of an Integrated GPS/BSfS System for Underwater Networks," Ninth International Symposium on
Shallow-Water AUV Navigation (SANS)," Proceedings of Unmanned Untethered Submersible Technology (UUST)
the IEEE Oceanic Engineering Society Conference 95, University of New Hampshire, Durham
AUV96, Monterey California, June 3-6 1996, New Hampshire, September 25-27 1995. Available at
ftp://taurus. cs. nps. navy.mil/pub/auv/ipoversw.ps
pp. 268-275.
Bacon, Daniel Keith Jr., Integration of a Submarine into Brutzman, Don, "Virtual World Visualization for an
NPSNET, Master's Thesis, Naval Postgraduate School, Autonomous Underwater Vehicle," Proceedings of the
Monterey California, September 1995. Available via IEEE Oceanic Engineering Society Conference
http://www-npsnet. cs. nps, navy, mil/npsnet OCEANS 95, San Diego California, October 12-15 1995,
pp. 1592-1600. Available at
Bottoms, Al, chair and editor. Symposium on Autonomous ftp://taurus. cs. nps. navy. mil/pub/auv/oceans95.ps. Z
Vehicles for Mine Countermeasures, Naval Postgraduate
School, Monterey California, April 1995. Brutzman, Don, "Tutorial: Virtual World for an
Autonomous Underwater Vehicle (AUV)," IEEE Oceanic
Brooks, Rodney A., "A Robust Layered Control System for Engineering Society Conference OCEANS 96,
a Mobile Robot," IEEE Journal of Robotics and Fort Lauderdale Florida, September 23-26 1996.
Automation, vol. RA-2 no. 1, March 1986, pp. 14-23. Available at
http:/wv^. stl. nps. navy. mil/~auv/uvw_tutorial. html
Brutzman, Donald P. and Compton, Mark A., "AUV
Research at the Naval Postgraduate School," Sea Brutzman, Don, Burns, Mike, Campbell, Mike, Davis,
Technology, vol. 32 no. 12, December 1991, pp. 35-40. Duane, Healey, Tony, Holden, Mike, Leonhardt, Brad,
Marco, Dave, McClarin, Dave, McGhee, Bob and Whalen,
Brutzman, Donald P., "From virtual world to reality: Russ, "NPS Phoenix AUV Softw^e Integration and
designing an autonomous underwater robot," American In-Water Testing," Proceedings of the IEEE Oceanic
Association for Artificial Intelligence (AAAI) Fall Engineering Society Conference AUV 96, Monterey
Symposium on Applications of Artificial Intelligence to California, June 3-6 1996, pp. 99-108. Available at
Real-World Autonomous Mobile Robots, Cambridge ftp://taurus. cs. nps. navy. mil/pub/auv/auv96.ps
Massachusetts, October 23-25 1992, pp. 18-22. Available
at ftp://taurus. cs. nps. navy. mil/pub/auv/aaai92ws.ps. Z Brutzman, Don, "Graphics Internetworking: Bottlenecks
and Breakthroughs," chapter. Digital Illusion, Clark
Brutzman, Donald P., Compton, Mark A. and Kanayama, Dodsworth editor, Addison-Wesley, Reading
Yutaka, "Autonomous Sonar Classification using Expert Massachusetts, to appear 1996. Available at http://
WWW. stl. nps. navy, mil/'-brutzman/breakthroughs. html
Systems," Proceedings of the IEEE Oceanic Engineering
Society Conference OCEANS 92, Newport Rhode Island,
October 26-29 1992, pp. 554-559. Available at Brutzman, Don, NPS Phoenix AUV Software Reference,
ftp://taurus. cs. nps. navy. mil/pub/auv/oceans92.ps.Z November 1996. Available at http://
WWW. stl nps. navy. mil/~auv/software_reference. html
5^97
Campbell, Michael Scott, Real-Time Sonar Classification Healey, A. J., Marco, R.B. and McGhee, R.B.,
for Autonomous Underwater Vehicles, Master’s Thesis, "Autonomous Underwater Vehicle Control Coordination
Naval Postgraduate School, Monterey California, Using a Tri-Level Hybrid Software Architecture,"
March 1996. Proceedings of the IEEE Robotics and Automation
Conference, Minneapolis Minnesota, April 1996.
Carey, Rikk, Marrin, Chris and Bell, Gavin, ’’The Virtual
Reality Modeling Language (VRML) Version 2.0 Holden, Michael J., Ada Implementation of Concurrent
Specification," International Standards Organization/ Execution for Multiple Tasks in the Strategic and Tactical
International Electrotechnical Commission (ISO/BEC) draft Levels of the Rational Behavior Model for the NPS AUV,
standard 14772, August 4 1996. Available via the VRML Master's Thesis, Naval Postgraduate School, Monterey
Repository at http://www.sdsc.edu/vrml California, September 1995.
Curtin, Thomas B., Bellingham, James G., Catipovic, IEEE Standard for Distributed Interactive Simulation
Josko and Webb, Doug, "Autonomous oceanographic (DIS) - Communication Service and Profiles, IEEE
sampling networks," Oceanography, vol. 6, 1993, Standard PI 278.1, Institute of Electrical and Electronic
pp. 86-94. Additional information at http:// Engineers, New York, 1995. Information available at
web. mit. edu/afs/athena/org/s/seagrant/www/auv. htm http://www. sc. ist. ucf edu/~STDS
Davis, Duane, Precision Maneuvering and Control of the Kanayama, Yutaka, "Introduction to Motion Planning,"
Phoenix Autonomous Underwater Vehicle for Entering a CS4313 Lecture Notes, Naval Postgraduate School,
Recovery Tube, Master's Thesis, Naval Postgraduate Monterey California, March 1995.
School, Monterey California, September 1996. Includes
video appendix. Available via Leonhardt, Bradley J., Mission Planning and Mission
http://www. cs. nps. navy, mil/resear ch/auv Control Software for the Phoenix Autonomous Underwater
Vehicle (AUV): Implementation and Experimental Study,
Davis, D,, Brutzman, D., Leonhardt, B., McGhee, R., Master's Thesis, Naval Postgraduate School, Monterey
"Operational Mission Planning and Mission Control for California, March 1996. Available at
the Phoenix Autonomous Underwater Vehicle," IEEE http://WWW. cs. nps. navy, mil/research/auv
Journal of Oceanic Engineering, in review, 1996.
Macedonia, Michael R. and Brutzman, Donald P.,
Flagg, Marco, "Submersible Computer for Divers, "MBone Provides Audio and Video Across the Internet,"
Autonomous Applications," Sea Technology, vol. 35 no. 2, IEEE COMPUTER, vol. 27 no. 4, April 1994, pp. 30-36.
February 1994, pp. 33-37. Available at
ftp://taurus. cs. nps. navy. mil/pub/i3lo/mbone. html
Foley, James D, van Dam, Andries, Feiner, Steven K. and
Hughes, John F., Computer Graphics: Principles and Marco, D. B. and Healey, A. J., "Local-Area Navigation
Practice, second edition, Addison-Wesley, Reading Using Sonar Feature Extraction and Model-Based
Massachusetts, 1990. Predictive Control," IEEE Symposium on Autonomous
Underwater Vehicle Technology, Monterey California,
Fortner, Brand, The Data Handbook: A Guide to June 3-6 1996, pp. 67-77.
Understanding the Organization and Visualization of
Technical Data, Spyglass Inc., Champaign Illinois, 1992. Marco, D. B., Healey, A. J. and McGhee, R.B.,
"Autonomous Underwater Vehicles: Hybrid Control of
Fossen, Thor I., Guidance and Control of Ocean Vehicles, Mission and Motion," Autonomous Robots, vol. 3, 1996,
John Wiley & Sons, Chichester England, 1994. pp. 169-186.
Healey, A.J. and Lienard, D., "Multivariable Sliding Mode McClarin, David W., Discrete Multi-Mode Kalman
Control for Autonomous Diving and Steering of Filtering of Navigation Data for the Phoenix Autonomous
Unmanned Underwater Vehicles," IEEE Journal of Underwater Vehicle, Master's Thesis, Naval Postgraduate
Oceanic Engineering, vol. 18 no. 3, July 1993, School, Monterey California, March 1996.
pp. 327-339.
Moravec, Hans, "The Stanford Cart and the CMU Rover,"
Healey, A.J., Marco, D.B., McGhee, R.B., Brutzman, D.P. Proceedings of the IEEE, vol. 71 no. 7, July 1983,
and Cristi, R., "Evaluation of the NPS Phoenix pp. 872-884.
Autonomous Underwater Vehicle Hybrid Control System,"
Proceedings of the American Controls Conference (ACC)
95, San Francisco California, June 1995.
5-98
Sayers, Craig P., Yoerger, Dana R., Paul, Richard P. and 18 SOFTWARE AND DOCUMENTATION
Lisiewicz, John S., "A Manipulator Work Package for All source code, support files and compiled
Teleoperation from Unmanned Untethered Vehicles - executable programs are available via the Internet
Current Feasibility and Future Applications," (Brutzman 96a). This software reference includes
International Advanced Robotics Programme (IARP) on help files, Phoenix software, 3D graphics viewer,
Subsea Robotics, Toulon France, March 27-29 1996.
hydrodynamics, sonar modeling, networking and
Additional information at http://www.dsLwhoi.edu
Multicast Backbone (MBone) resources. AUV
Shank, Roger C., "Where's the AI?," AI Magazine,
dynamics software is parameterizable for other
voL 12 no. 4, Winter 1991, pp. 38-49. vehicles and all work is in the public domain.
Available at http://www.stLnps.navy.mil/-auv
Smith, Samuel M. and Dunn, Stanley E., "The Ocean
Voyager II: An AUV Designed for Coastal Acknowledgements. The authors thank Mike Zyda
Oceanography," Proceedings of the IEEE Oceanic and Yutaka Kanayama for help and advice during
Engineering Society Conference Autonomous Underwater the conduct of this research. We are also grateful to
Vehicles (AUV) 94, Cambridge Massachusetts, approximately eight dozen colleagues and students
July 19-20 1994, pp, 139-147. Additional information
of the NPS Center for AUV Research who have
available at http://www.oe.fau.edu/AMS
made valuable contributions to Phoenix. Financial
Stevens, Richard W., Advanced Programming in the Unix support for this ongoing work has been provided by
Environment, Addison-Wesley, Reading Massachusetts, the National Science Foundation under Grant
1992. BCS-9306252 and the Naval Postgraduate School
Research Initiation Program.
Stewart, W. Kenneth, "Visualization resources and
strategies for remote subsea exploration," The Visual To appear: AI-Based Mobile Robots, Kortenkamp,
Computer, Springer-Verlag, vol. 8 no. 5-6, June 1992, David, Bonasso, Peter and Murphy, Robin, editors,
pp. 361-379. MIT/AAAI Press, Cambridge Massachusetts, 1997.
Strauss, Paul S. and Carey, Rikk, "An Object-Oriented 3D
This chapter is available online at
Graphics Toolkit," COMPUTER GRAPHICS,
vol. 26 no. 2, July 1992, pp. 341-349. http://www. stl nps. navy. mil/-auv/aimr. html and
http://www, stl. nps. navy. mil/-auv/aimr.ps
Thalmann, Daniel, editor, Scientific Visualization and
Graphics Simulation, John Wiley & Sons, Chichester
Great Britain, 1990.
5-99
5-100
A Small Co-Axial Robotic
Helicopter for Autonomous Mine-Field
Search and Destroy Missions
Charles Colby
Aero-Nautical Robotics Corporation
2991 Alexis Drive
Palo Alto CA. 94304
phone 415.941.9090 fax 415.949.1019
e-mail: COLBY@SURF.COM
I have worked in Silicon Valley for 25 years and I knew Then I had a friend design a very small, lightweight, low
where to find most of the high-tech components of a mine power, state-of-the-art inertial stabilization system using gyros
identification and locating system...and the people who could and a solid state accelerometers that make the vehicle
put it together. What was missing were 4 key elements: inherently stable.
1. The Vehicle to carry the equipment...a vehicle that can We then located a company called Geometries that makes
fly 5 to 40 feet high at 5-10 miles per hour. very sensitive, and very lightweight cesium Magnetometers.
This Geometries unit is the most sensitive type
2. A very sensitive Magnetometer to detect the mines...one Magnetometer available, and can detect most buried land
that was available in a small, lightweight package. mines.
3. High precision GPS to guide the vehicle. We are designing an on-board computer system to
manage the vehicle’s autonomous operation.
4. The Funds to pay for the development of the system. It Uses 20 mips embedded computer that is the size of a
A closer look at each of the 4 missing element reveals the credit card.
following; • It uses fuzzy logic to make decisions on its own.
1. The Vehicle:
What is needed is an anti-gravity machine that will carry
about 25 pounds of payload.
5-101
•The airborne system is autonomous which means Look at the progress I have made on this project with only
you tell it what to do and it goes and does it.., with no further $100,000 invested so far...
contact needed from the ground. ( Giant companies like Lockheed and Boeing would have
•You can uplink a script file from the ground spent 10 million dollars by now to get this far on a project
control laptop to the airborne vxhicle and it goes and does it’s like this)
prescribed maneuvers unless you override it with a command
given from the ground. I feel it is time to give something back to the world...
With 100 million indiscriminate -killing-machines
That brings us to number 3 on the list of missing (buried mines) in place in the world and 1500
elements.
people..mainly children., being maimed and killed each
month., something needs to be done and the US government
3. GPS system: (or any government) is not doing their part to help this
The Global positioning system uses 24 satellites. In situation.
1994, the best accuracy a GPS receiver could give you was The only US government funded demining research that I
about 6 meters.
know of is the project at Fort Belvoir called the “Humanitarian
This was not good enough. We need accuracy of less Demining Program”, It is being run by Harry N. (Hap)
than 10cm to be able to pinpoint the position of a mine and Hambric. Hap and his crew are doing a fantastic job with the
then come back and dig it up , destroy it in place , or carry it limited funding they have available. What is needed is
away.
additional funding for Hap’s organization. Then possibly there
would be funding available to continue my co-axial robotic
Most people are under the false impression that GPS is helicopter project. Hap Hambric can be reached at
only accurate to 30 meters. Now, with the latest differential 703.704.1086. Here is what you can do to help make a
GPS systems available, the accuracy is down to 2 cm. difference: Call your Congressperson and explain that more
funding is needed to combat this terrible buried mine problem.
•The price is now a great deal lower now.
•The size and weight are much less now.
2. Conclusion
So now an off-the-shelf GPS system is available that
meets the requirements of this project.
The good news is we have an answer to the problem with
our Robotic Helicopter... the bad news is that it is only 75%
The last missing element was the money to pay for the
completed and I have maxed out my Visa cards so
R&D for this project..
development has stopped.
This is THE BIG PROBLEM. We have tried numerous
money raising efforts:
So I come here today to ask for help in finishing this
• We have written a business plan for the project.
project.
•We have distributed about 50 of the business plans.
• Nobody, no institutions, no government agency,
We need: Strategic Relationship Paitners and Funding to
no venture capitalist...nobody was willing to even consider
finish the R&D on the Project.
this project. “Land mines are not a very sexy subject” and
additionally, unless your business plan has the word “Internet”
If you have access to a budget that can support this
in it every two paragraphs, forget it with any venture
project or know of one, please contact me at
capitalists.
415.941.9090
So, I bit the bullet and financed this whole R&D project
Thanks for your help.
my self myself with my VISA gold cards. My own personal
VISA cards are now maxed out at $100,000.
5-102
5-103
COLBY CO-AXIAL
HELICOPTER
PRELIMINART SPECIPICATION HIGHLIGHTS
5-105
FIGURE 5
^• LOW COST
AUTONOMOUS
MINE-FIELD
SEARCH AND
DESTROY VEHICLE
• COMPUTER AND GPS CONTROLLED
• PATENT-PENDING IN-PLACE
MINE DETONATION METHOD
• 30 LB PAYLOAD
WE ARE SEEKING:
‘ STRATEGIC RELATIONSHIP PARTNERS arc
• FUNDING TO FINISH R U D Aero fl Nautical Robotics Corp.
• CUSTOMERS AND END USERS
5-107
2991 ALEXIS DRIVE, PALO ALTO, CA. 94304 PH 415-941-9090 FIGURE 7
FAX 415-949-1019 E-MAIL: COLBY@SURF.COM
5-108
Fully Autonomous
Land Vehicle For
Mine Countermeasures
Raymond C. Daigh, RAHCO International
Phil Rice, Lockheed Idaho Technologies Company
5-109
transport modules, (waste containers), were
redesigned and implemented in disposable
materials. The waste transport container and bed
plate were re-configured for end loading. A
systems health monitor was also installed.
5-110
Differential GPS
RF Modem k
Base Receiver
GPS Rover 1
Compass Navigation
Lid Open/Close
OAS Solenoid Out
-H Latch Lock/Unlock
Engine Temp Controller
Engine Oil Press >_
Battery Voltage Cradle Open/Close
Enclosure Temp RS-485
Hydraulic Oil Temp
Camera Select
Camera Digital Out
Pan/Tilt
Controller
Latch Unlocked Zoom/Focus
Left/Right
Front/Rear PWM Out Track Pumps
Ultrasonic Sensors
Servo
Controller Power Out Travel Siren
Left/Right
Track Encoders
I Video Transmitter
Emergency
Stop
RS-232
▼
Video Receiver r Genie Micro
*1 Supervisor's
Travel Joystick —, 1 RS-485 Emergency
Stop
Camera Controls
L
Genie Micro
Container Switches
5-111
Compass heading signals were fused, each In addition to a primary navigation system, it was
computational cycle (20 Hz), with derived necessary to develop a collision avoidance
heading information from the track encoders and system that complemented the basic navigation
the rate gyro which produced a combined mea¬ functions. The collision avoidance system
surement (drHdg) better than its parts. The final consisted of an array of broad beam ultrasonic
sensor fusion was averaging the two track sensors mounted on each end of the vehicle.
encoder's linear velocity (vAng). This system of Each array of coordinated pulse sensors consisted
redundant signals provided a high level of of nine transmitter receivers that operated with
flexibility in sensor usage and greater overall overlapping convergence zones in a ring configu¬
dead reckoning system reliability. Pitch and roll ration. This resulted in 3 feet of side coverage at
data derived from the compass was used to a range of 17 feet, with the only discontinuance
correct DGPS position (adjN, adjE) for operation in beam coverage occurring between the vehicle
on uneven terrain. Reported position and heading and 5 feet of range. These spaces uncovered by
(drN, drE, drHdg) was the result of correcting the the beam were very narrow triangular areas
dead reckoned position divergence with the located inside the emergency stop range. For the
adjusted DGPS position in a supervisory control purpose of velocity control, there were two sonic
loop. This correction took place at 2 Hz while zones.
the dead reckoning system calculated positions at
20 Hz. Important safeguards incorporated into
the overall design included the compass' ability Detection Zone -
5-112
Obstacle Avoidance Flow
3
between the first echo and confirmation. In through use of the RS485 busses between the
order to provide controlled vehicle deceleration, a microcontrollers on board the vehicle and the
collision to detection zone ratio of 20:15 was remote control station. The real time operating
chosen for proper algorithm balance. Incorpora¬ system allowed a high degree of customizing and
tion of hydraulic latency and braking time streamlining communications within the data
resulted a minimum stopping distance of 6 feet transfer between modules. Telemetry data
at 5 feet per second velocity. Therefore, an transfer between the vehicle and the remote
active emergency stop zone was included for any operator's station was a tightly defined optimized
single echo inside of six feet. An allowance was data packet utilizing Cyclic Redundancy Check
made to identify and register nine separate as the main control communications. Additional
obstacles and calculate avoidance velocity telemetry systems included video communica¬
reductions for the closest one, as shown in the tions for the vehicle vision system, differential
central loop of the Obstacle Avoidance Flow GPS corrections, supervisors dead man switch,
Diagram.. When an obstacle disappeared from and emergency stop. Independent communica¬
the vehicle’s sensing zone, based on a lack of tions' channels allowed isolation of critical
echoes within the tolerance band of a registered information from nonessential data and improved
obstacle, the system would reset and then restore vehicle reliability. The remote station's video
normal velocity control. For close approach and vision system camera controls, telerobotic remote
vehicle docking, the emergency stop zone control, and primary operator interface communi¬
stopping distance was coordinated with the cated by close coupled serial links. The control
vehicle’s actual speed to gradually approach but center was configured as a three part system
never reach, a zero distance. This permitted a consisting of the modified TTV remote control¬
very slow speed docking without eliminating the ler, mission planning computer, and vehicle
emergency stop zone. The bottom loop of the control command generator (VCCG). The
Obstaele Avoidanee Flow diagram shows func¬ mission planning and VCCG algorithms were
tional implementation of the supervisory velocity written in visual basic and tested in interpreted
controls. Another important system feature was code for monitoring and modification ease. This
coordinating the sensor bank selection with travel approach was satisfactory for the relatively low
direction and implementing bank switching on speed demonstration requirements of real time
the basis of net linear velocity as derived from kinematic control.
the track encoders.
Mission Planning GUI
AUXILIARY SENSOR SYSTEMS
The mission planning interface was designed to
Vehicle auxiliary sensor systems included all provide ease of use and point and click operabil¬
normal engine and hydraulic alarms and waste ity. Fundamental setup was based on a scaled
transport container lid and latch system robotics operating area map with known obstacles-
control. These systems were integrated into the buildings, trees, and other fixed equipment-
overall on-board control architecture and distrib¬ overlaid by no go zones with adjustable exclusion
uted among the on board micro controllers. This borders. The pre-planned vehicle route was
resulted in auxiliary subsystem reliability and drawn on this scaled map. A predefined
equal processor loading between micro control¬ waypoint was established at each route location
lers. where the vehicle heading changed. Each
waypoint had its own set of parameters within the
SYSTEM INTEGRATION transition area.
5-114
Waypoint Navigation will proceed to this point in a direct line from its
current location.
The waypoint method of vehicle control was
chosen to provide segmented mission control and Fly by Point -
definitive means by which navigation precision A point on the course where a pivot turn is
was measured and logged in real time. The performed.
VCCG logged vehicle data during operation,
which was used as the basis for statistical system Smooth Flyby Point -
performance analysis. There were five different A point on the course where a radius turn is
types of waypoints, each with similar characteris¬ performed.
tics. Through the linear segments. Global
operational parameters were in effect for linear Pause Point -
and angular velocity. At each waypoint the A stopping point along the path where the path
Washin and Washout Circles permitted customiz¬ tracking may be resumed.
ing the global control parameters for each
waypoint type. End Point -
A docking location at the end the path.
Start Point -
The starting point for the mission. The vehicle
Path Definition
5-115
During normal operation the vehicle operated The control algorithms running under the VCCG
continuously on the basis of off track error performed all travel control and systems safety
correction to the next waypoint. The vehicle interlocking functions. The autonomous control
heading was therefore corrected continuously to system and TTV joystick controller outputs were
ensure arrival at the next waypoint. Minor identical in design. In fact, the on board systems
deviations in track and heading were expected at did not differentiate between the two control
the Washin Circle. methods. The control station’s telemetry router
determined which control signal source was
Vehicle Control Command Generator GUI transmitted to the vehicle, either TTV joystick or
VCCG. From this MMI it was possible to define
This interface was the primary monitoring all mission tuning parameters. These included
interface for vehicle operation. It included all of velocity, acceleration, and deceleration param¬
the mapping and path planning generated in the eters for both linear and angular motion, mini¬
mission planning MMI and all of the vehicle mum and maximum radius’ for flyby waypoints,
operating system monitors. The display also and tuning parameters for control loops.
traced the vehicle path against the planned path
in real time and allowed operator intervention
when necessary due to receiving any one of
several alarms or abnormal condition reports.
^max = 10 dcg/s
Pause Radius = 2 ft
Position Radius = 1 ft
Flyby Radius = 2 ft
Path 1
5-116
DEMONSTRATION RESULTS of the three month checkout or testing periods.
5-118
equipped with robotic manipulators to retrieve REFERENCES
and transport plutonium or other radioactive
materials is currently possible. Additionally, the 1. P.M. Rice, et. al, “ Evaluation of a Self Guided
basic technology exists today to provide a Transport Vehicle For Remote Transportation of
completely autonomous vehicle capable of Transuranic and Other Hazardous Waste”, INEL,
performing mine countermeasure activities on November 1995.
land, in shallow water, and in surfzones without
risking the lives of U.S. military personnel. 2. H.J. Tucker, et al,” Design, Development,
Integration and testing of a Self Guided Transfer
SUMMARY Vehicle and Remote Excavator for Cooperative
Buried Waste Retrieval Integrated Demonstra¬
Autonomous vehicle technology advancements tions”, December 1995.
are expected to be a continued priority well into
the 21st century. This is primarily due to the
increased need and interest in physically safe
methods of remediating environmentally hazard¬
ous areas. Current vehicle technology allows
accurate, short range, mission deployments. The
SGTV provides a perfect platform for a multitude
of application modifications yet maintains
flexibility for future enhancements. During the
technology demonstration the SGTV proved it
could travel pre-planned courses within 9 inches
at speeds up to 5 miles per hour. Obstacle
avoidance and supervisor safety systems com¬
plete the GPS strapdown dead reckoning naviga¬
tion package ensuring safe, reliable operation.
The technology demonstration was deemed a
success considering the low research and devel¬
opment costs and short nine month schedule from
conception to completion.
ACKNOWLEDGMENTS
5-119
ADVANCED TECHNOLOGY: IT’S AVAILABLE AT JPL
James R. Edberg
Jet Propulsion Laboratory
California Institute of Technology
james.r.edberg@jpl.nasa.gov
5-121
orbits for a design period of four years, exploring the planet and its sateiiite system and
discharging a probe into the atmosphere of Titan. The spacecraft is about three stories
high, weighs about 5 tons, and carries the most extensive and sophisticated set of
instruments ever fiown. It is also a billion dollar project. Missions on that scale, and
perhaps launched only once in a decade, are no longer affordable or desirable - nor are
they necessary. The emphasis is on smaller, lighter, iower powered, and cheaper
vehicles.
JPL’s Space Microelectronics Program is developing the technology to satisfy all of
those criteria. Fig. 1 shows a comparison or progression (downsizing) from Cassini to a
possible micro-spacecraft of the future, utilizing a variety of the JPL developed micro¬
sensors and associated technology. This effort has produced a number of miniature
sensor packages, including ,for example: accelerometers, seismometers, radiometers
.hygrometers. A hydrophone developed for the Navy is encapsulated in a 1 inch sphere.
(Fig 2). In addition to small size, these devices exhibit extreme sensitivity, ruggedness,
and very iow power consumption.
Of particular interest may be a Reversed Electron Attachment Detector (READ). It is a
man-portable device capable of unambiguous detection of unique chemicai signatures
associated with mines. READ has demonstrated the ability to detect 2,4-DNT; 2,4,6-TNT:
PETN; and RDX in parts per trillion concentrations, as weil as nerve and blister agents
and non-conventional explosives such as perchloro and peroxy compunds.
Utilizing complementary metal-oxide semi-conductor (CMOS) technology, JPL has
developed a new imaging sensor - virtually a camera on a chip - promising smaller and
cheaper imaging systems but comparabie in performance to the current state of the art
(Fig. 3). This active pixei sensor technique represents a considerable leap beyond the
widely used charged coupled device (CCD) technology. Use of the CMOS sensors
presents the opportunity for reducing imaging costs, power and size, and improving
reliability.
The thrust toward miniaturization piaces a concomitant need for improved power
sources. The Laboratory has an extensive effort devoted to advanced power sources
featuring small size, long life and increased specific energy. In the range up to lOOkW
and specific energies to over 200 watt-hours/kg. Fig 4 shows a number of ceil types
under consideration, development and test. An “AA” LiTiS^cell has been successfully
cycled 1,000 times to 50% discharge at ambient temperature. Also under development
(Fig 5) is a direct methanol, liquid feed fuel cell where a 3% methanol/water mixture is the
fuel and air is the oxidant. Advantages include simplicity, start up at room temperature,
operation at 70®to 90®C, and no resulting pollutants, the only outputs being potable water
and COg. The system is modular, with a 4 x 6 inch cell providing 50 amperes
continuously at 0.4 volt.at 90®C with air. Plan is to demonstrate a 1kW fuel cell stack next
year.
In addition to the above, other JPL technologies which merit investigation for marine
applications, include - but are not necessarily limited to: teleoperators/robotics; roving
vehicles: communications: data collection.processing compression: digital imaging and
visualization: guidance.control, navigation: and certainiy quality control and systems
integration.
Teleoperator/robotic activities range in size and function from a large, seven-degree
of freedom arm being deveioped for NASA use as an autonomous surface inspection
device for the Space Station, to a micro-surgery device for medical applications, e.g.,
inside the eyeball surgery. The NASA arm incorporates an eddy current sensor for
detection of minute pits or cracks, as well as a proximity sensor to avoid actual contact
5-122
with the surface. The micro-surgery device is being developed in conjunction with an eye
surgeon, and has an accuracy/repeatability of ten angstroms. The device will also
eliminate any tremors resident in the surgeons hands, even his pulse beat. Also in the
laboratory is a modular, eleven degree of freedom arm, about 2 inches in diameter, which
permits access into intricate, complex passages.
An autonomous roving vehicle, “Sojourner” (Fig 6), will be mounted inside the
“Pathfinder” spacecraft, which will be launched this coming December and will land on
the surface of Mars on July 4,1997. After landing, “Pathfinder” will unfold and deploy the
rover onto the surface of Mars. It will be directed to explore certain targets or areas,
navigating on its own, and performing engineering and scientific experiments. “Sojourner”
will transmit its information to the lander for re-transmission back to earth. The rover’s
prime power, 16 watts, is provided by a 0.2 square meter solar panel, backed up and
augmented by lithium sodium di-oxide “D” cells.
New autonomous control and data processing methodologies are being developed
which can be applied to underwater target detection, where transmitted pulse sequences
form a non-gaussian process in the presence of ambient/environmental noise. Using
these statisitical techniques with inherently efficient algorithms for emerging parallel
computational architectures (e.g. systolic arrays, neural networks) will result in effective
“near optimal’ algorithms for high performance, real-time underwater signal and target
detection, identification and tracking. Additionally, the autonomous control methods will
enable unmanned underwater manuevering with complete failure detection, identification
and recovery capability.
Because of the crippled 16 foot diameter high-gain antenna on the Galileo Jupiter
orbiting spacecraft, communications have had to rely on the much smaller low-gain
antenna, with a consequent decrease in transmission rates from an expected 135 kbs to
less than 2 kbs. To compensate, JPL communications researchers have devised
methods to extract the maximum amount of information from this abridged data stream.
The technique is described as a “feature driven data compression technique for
bandwidth critical applications”. Since underwater transmissions are typically bandwidth
critical, this approach may offer increased information transmission over a given
bandwidth.
Underway at JPL is a 3-year technology demonstration prog ram,“MU DSS” (Mobile
Underwater Debris Survey System) funded by the Strategic Environmental Research and
Development Program (SERDP) in the Cleanup thrust area. Its purpose is to demonstrate
technologies necessary to successfully survey underwater “formerly used defense
sites”(FUDS) for ordnance and explosive waste (OEW). The program is a joint
Department of the Navy and NASA effort being executed by the Naval Sea Systems
Command’s Naval Surface Warfare Center (NSWC), Dahlgren Division, and the Jet
Propulsion Laboratory. The first year of the effort was completed in 1995, to (1)
demonstrate that a prototype MUDSS sensor suite shows good promise against inert
OEW targets, and (2) provide a multi-sensor data base to be used during Phase II to refine
processing algorithms prior to at-sea testing at actual FUDS.
The foregoing provides a brief and necessarily limited introduction to several of the
existing technologies which it is felt have application to areas of your interest. Far from
satisfying your curiosity, it is hoped that this exposition will serve to arouse your interest
and provide motivation to explore further. From our standpoint, the preferred follow-up
would be to have you personally visit the Laboratory, which would serve two purposes:
(1) it would afford you the opportunity to see hardware and discuss these and other
technologies first hand with those directly involved, and (2) it would provide us your
5-123
assessment of these technologies and advice as to the direction for future developments.
It is a win - win situation!
The Consortium for Oceanographic Research and Education, CORE, undertook an
Interagency Partnership Initiative to “reexamine our Nation’s posture toward ocean
science and technology and establish a new and invigorated partnership concept”. The
Initiative produced “Oceans 2000: Bridging the Millenia -- Partnerships for Stakeholders
in the Oceans”. Among its recommendations were: (1) define specific research and
education partnership opportunities for academia, industry, and the Federal Government,
and (2) develop an integrated partnership management plan to provide effective and cost
efficient federally funded ocean science and technology programs.
The 1992 Ocean Studies Board Report, “Oceanography in the Next Decade: Building
New Partnerships”, in commenting on such partnerships, stated: “In general, partnerships
must be extended beyond financial relationships to include the sharing of intellect,
experience, data, instrument development, facilities, and labor”.
In his keynote address at the JPL Undersea Technology Symposium in May of this
year, RADM (Ret) Brad Mooney reviewed preliminary results of the Marine Board Study,
“Undersea Vehicles and National Needs”. He stated, “An increased role for AUV’s is
anticipated in all of the areas I’ve mentioned. AUV’s will require the most technological
advances for them to be competitive, efficient, and effective, but they promise great
payoffs in capabilities as sensors, communications, and control techniques are improved”.
The sentiments expressed in the previous three paragraphs are typical of current
thinking and reflect an awareness of the stringent budgetary constraints faced by all, and
the consequent need to conserve resources and use capabilities wherever they may
reside. We at the Jet Propulsion Laboratory invite and encourage your investigation of
our capabilities as potential resources for your use.
As Dr. Don Walsh commented in an article in the April issue of “Sea Technology”,
when referring to the JPL Undersea Technology Program: “A model that right now
produces bought-and-paid-for technologies that can be adapted to ocean research and
business. TAKE ADVANTAGE! THE PRICE IS RIGHT!”
The research described in this article was carried out by the Jet Propulsion
Laboratoiy, California Institute of Technology, under a contract with the National
Aeronautics and Space Administration.
5-124
Mission Planning for an Autonomous Undersea
Vehicle: Design and Results
Michael J. Ricard, Ph.D.
C. S. Draper Laboratory
555 Technology Square
Cambridge, MA 02139
Abstract - This paper describes the design of the targets, near real-time acoustic communications and an over-
on-line mission planning system implemented i n the-horizon approach to the region of interest.
the Autonomous Minehunting and Mapping The AMMT program extended the DARPA UUV’s
Technologies (AMMT) program sponsored by the autonomous capability by adding an on-line mission planner.
Defense Advanced Research Projects Agency
The mission planner is the focus of this paper. The planner
(DARPA). The AMMT vehicle was designed to
provides the flexibility to adapt the mission plan as dictated
survey undersea areas and map the terrain and any
by the circumstances that arise during the mission.
potentially mine-like objects. If the mission
planner determined it was safe, the vehicle would As input, an operator specifies high level mission
also image bottom mines. The vehicle i s objectives that are to be achieved during the mission. The
completely autonomous so the planner i s operator also specifies the utility of each objective. The
responsible for not only generating initial utility is a relative ranking of the objectives. If an objective
trajectories but modifying the plan to accommodate is twice as important to complete as another objective, then
unforeseen events. its utility would be twice as much as the other. Along with
As input, the planner is given a series of high- the objectives, the operator also specifies time and fuel limits
level activities to perform and their associated for the entire mission. It is the mission planner’s
utility. The planner must determine the sequence o f
responsibility to execute a plan that maximizes the expected
activities that optimizes the expected utility of the
utility of the mission using the time and fuel allotted.
mission while working within operational
Planning is a search through the infinite space of possible
constraints such as time and fuel constraints.
Typical activities include surveying specified decisions to find that sequence of decisions, or plan, that best
regions and imaging targets and getting a GPS achieves the given objective. A search algorithm generates
navigation fix. The vehicle must obey strict depth possible vehicle trajectories which are then scored using a
and altitude constraints so the execution of each cost function. The cost function is a weighted combination of
activity requires that both terrain and target the estimated resources (time and fuel) needed to perform each
obstacles be avoided. The planner does not have activity in the mission and the utility of each mission
any a priori knowledge of the terrain and must rely objective. The mission planner scales the utility by the
on data gathered by an ahead-look sonar. As new probability of completing the objective in the mission. In
information about the environment is learned, the
this manner, a near optimal vehicle trajectory is selected
planner must re-plan the mission to insure the
given the specified goals and constraints.
safety of the vehicle. The planner was successfully
demonstrated to work in the AMMT vehicle in both
simulation test and at-sea testing. n. Requirements
This paper describes the design of the mission Perhaps the most important requirement for an autonomous
planner, implementation issues and presents results planner is to be able to plan both quickly and effectively.
from both the simulation and at-sea tests. Possible Since the planner typically has no a priori knowledge of the
enhancements and areas for future research are also operating environment and the environment is inherently
presented.
stochastic, plans cannot be generated long before their
execution. The planning process must be flexible enough and
L Introduction
fast enough to be able to respond to stochastic events without
The Autonomous Minehunting and Mapping Technologies sacrificing the vehicle’s safety.
(AMMT) program is a DARPA response to the need for It is not sufficient to merely generate plans quickly; the
clandestine reconnaissance. AMMT combines precise mission planner is also responsible for generating vehicle
navigation, adaptive vehicle maneuvering, long-range acoustic trajectories that are safe and achievable. Safe trajectories are
communications, underwater imaging, bathymetric mapping, trajectories that avoid obstacles, maintain testbox constraints
and acoustic tracking and navigation to enable mine and and are robust to sensor dropouts. Obstacles include
obstacle detection and localization, imaging of underwater avoidance stovepipes, sonar targets determined to be tethered
5-125
mines or suspended objects and terrain obstacles (water depth trajectory. The mission planner will always attempt to
constraints). Avoidance stovepipes are locations at which complete all preplanned activities within resource constraints.
there are known obstacles or otherwise keep-out regions and The contingent activities are different from the preplanned
are specified by the operator. The testbox can be any convex activities in that the mission planner does not consider their
polygon in the horizontal plane and is defined by waypoints inclusion in the plan unless a predetermined set of conditions
at the boundaries. In addition, there are minimum depth, is satisfied and including the activity increases the overall
minimum altitude and maximum depth constraints. The value of the mission.
mission planner must generate plans that the vehicle can Each preplanned activity may have ordering constraints
achieve so that the vehicle maintains the safe trajectory which restrict its position in the plan relative to other
determined by the planner. The intent is for this to happen activities.
with the planner having minimal knowledge of the vehicle’s The number of times each contingent activity is considered
dynamics. It should be able to function using only for inclusion in the mission plan is limited to the specified
rudimentary factors such as turning radius, and allowable maximum number of times that it is allowed to occur during
speed ranges. the mission. In addition, contingent activities may also have
When the mission planner is in control of the vehicle, ordering constraints which restrict where they can be inserted
surfacing of the vehicle is only allowed at previously in the plan.
identified safe regions (surface stovepipes) and requires a Start Mission
permission signal from the host ship to prevent the vehicle The Start Mission activity is used to command the vehicle
from surfacing into heavy traffic areas. Surface stovepipes are to leave the surface, attain start up depth and power up any
locations where it is safe for the vehicle to surface and arc necessary equipment.
specified by the user. Transit
Naturally, the planner must operate in a real-time The Transit activity is used to command the vehicle to a
environment. That is, it must take information from the specified waypoint. If obstacles or terrain prevent the vehicle
ahead-look sonar, process the data, evaluate possible plans and from reaching the specified point, it will approach the point
select and implement a plan in time to maneuver the vehicle as closely as possible.
along a safe trajectory Survey
The planner framework needed to have the capability to The Survey activity is used to search a region defined by a
receive asynchronous input from both the sonar system and convex polygon. A mode parameter selects the survey search
the host ship. The sonar system sends terrain and target data pattern, either a comb search or a survey area perimeter
to the planner asynchronously but can also request that the search.
planner take an acoustic image of a specified target or that the Data Upload
vehicle perform a depth excursion to sample the water The Data Upload activity is used to maneuver the vehicle to
column. permit acoustic transmission of data files (image or map data)
If the operator needs to override the planner and terminate to a distant host ship.
the mission in a controlled manner, a request can be sent to Sensor Image Point
the planner. Upon receiving such a request, the planner will The Sensor Image Point activity is used to image a
smoothly transition to the end mission activity which will particular point using either acoustic or optical imaging. The
surface the vehicle in a pre-specified location that is assumed vehicle follows a track specified by a length and azimuth. If
to be free of boat traffic. the mode is optical, the specified track is centered about the
Any effective software system should have a simple, point. If mode is acoustic, the specified track ends at the
intuitive interface. In the case of the mission planner, this point.
applies to the specification of mission objectives as well as For optical imaging, the planner will command the vehicle
feedback during the mission. To simplify the input required to drive over the ground track at constant pitch if terrain
to execute a mission and to provide modularity, the planner following is specified. The planner will trigger the optical
input consisted of high level mission objectives. These imaging device when the vehicle is above the commanded
objectives were specified in the form of high level activities latitude and longitude.
and their associated mission utility. A brief description of the Depth Excursion
activities implemented for the AMMT mission planner is The Depth Excursion activity is used to sample the water
provided in the next subsection. column using any onboard environmental sensors. The
vehicle drives to the commanded latitude and longitude while
A. AMMT Mission Activities obtaining the minimum testbox depth and commanded speed.
The mission planner uses preplanned and contingent After having reached the commanded position and minimum
activity input specifications to determine the vehicle testbox depth, it will then drive to the minimum altitude. The
5-126
vehicle will use ballast and propulsion systems to maneuver A graphical user interface was developed to assist the
in the water column. operator in developing the planner's activity input file and to
GPS Navigation Fix provide an interface to the pre-mission planner. The Pre¬
The GPS Navigation Fix activity is used to obtain a GPS mission planner was used to ensure that the activity file,
position fix, which requires the vehicle to be on the surface. based on a priori environmental information, is feasible and
The vehicle drives to the selected surface stovepipe and desirable.
requests permission to surface. Upon receipt of permission or Most activities are to be performed at a commanded
passage of a time-out period to pass, the vehicle will surface altitude. The terrain following planner is used to generate a
under ballast control. Once on the surface, the vehicle will depth profile that will allow the vehicle to closely follow the
wait for either the standard deviation of the navigation error to desired altitude while maintaining vehicle safety. It must take
decrease below the specified completion accuracy or for a into account vehicle maneuverability in the presence of
time-out to pass. varying terrain.
End Mission If the operator needs to override the planner and terminate
The End Mission activity is used to terminate the mission the mission in a controlled manner, a request can be sent to
at the specified location. The vehicle maneuvers to the the planner. Upon receiving such a request, the planner will
commanded location at minimum depth, stops and requests smoothly transition to the end mission activity which will
permission to surface. Upon receipt of permission or passage surface the vehicle in a pre-specified location that is assumed
of a time-out period, the vehicle will surface under ballast to be free of boat traffic.
control-
Contingent Sensor Image Target A. Hierarchical Planner
The Contingent Sensor Image Target activity is used to The mission planner decomposes the problem hierarchically
optically or acoustically image discovered targets. In either into problems that can be solved quickly. The hierarchy has
case, the target must be within an allowable imaging region three levels. At the highest level are the mission activities as
(defined by a latitude, longitude, and radius) and there must be specified by the operator. The top level planner uses low
a feasible groundtrack (considering obstacles and maneuvering fidelity estimates of resource consumption and the vehicle
constraints) to obtain target imagery. The approach azimuth state to determine the sequence of activities that maximizes
to the target is determined by the mission planner. the objective in an expected value sense. At lower levels in
For acoustic imagery, the activity is triggered when the the hierarchy, the estimates become more refined and this data
sonar/mapping system requests it. For optical imagery, the is aggregated and fed back to the higher levels. For each level
activity is triggered when there is a target meeting the in the hierarchy, there exist a set of activities at that level and
threshold criteria of minelike probability level and signal to a pair of functions associated with each level. The first
noise ratio. function is an estimator of the resources needed to complete
Contingent Depth Excursion the activity and the vehicle state at the end of the activity.
The Contingent Depth Excursion activity is used to sample The second function is used to expand an activity into a
the water column using any available onboard environmental sequence of more detailed activities at the next lowest level.
sensors and is requested by vehicle subsystems. This activity This architecture provides for a plug-and-play capability
will be performed in the vicinity of the vehicle’s position at because the planner itself does not need to know the details of
the time of the request. a mission activity.
Contingent GPS Navigation Fix As an example, consider the task of obtaining a GPS fix
The Contingent GPS Navigation Fix activity is used to which is illustrated in Figure 1. The top level planner uses
obtain a position fix when the navigation error has grown too an estimator that does not take into account details of the
large. activity. Using the estimated state of the vehicle at the start
of the activity, it computes a crude estimate of the resources
in. System Design needed to complete the task and estimates the vehicle state at
This section provides an introduction to the design of the the end of the activity. The top level planner uses this
mission planner. The design of the planner is a hierarchical information to insert the activity in the plan. This top level
one where the planning problem is decomposed into smaller activity is then broken up into five, more detailed, activities.
problems that can be solved with the time and information The vehicle must first transit to the stovepipe, request
available. This design also leads to a modular architecture so permission to surface, surface the vehicle, actually get the fix
the planner and plan management functions are separate from and ballast down. Each of these activities are able to make a
domain specifics such as the individual mission activities and more refined estimate of the resources needed. For example,
the mission database. The architecture was implemented in a the transit activity can use knowledge of the ocean current and
priority based, real-time UNIX environment. terrain to get a precise estimate of the time and fuel needed to
5-127
been met. This task was not aware of the details of the
activities or their trigger conditions.
The Mission Evaluator incorporates the output from the
Planner task into the currently executing plan. It also
controls the input to the planner. This input includes which
level the planner task should be working with and when it
should restart the planning process. Its most important job
was to verify the safety and the feasibility of the currently
executing plan. If it determined that the environmental
conditions had changed so much that the current plan was no
longer safe, it would invoke an immediate replan or provide
reactive measures for the near term.
The Guidance Interface operated at the highest priority and
was the planner's only interface with the fault tolerant
Figure 1: Decomposition of a GPS Fix processor. This task had to convert the "executing" activity
into a guidance command that the vehicle's guidance system
get to the stovepipe. The sum of each of the estimates at this could understand. It also monitored subsystem health,
level are then used to refine the top level estimate which may controlled subsystems and read environmental sensors.
cause the top level planner to resequence its activities and The Asynchronous Input Handler was responsible for
start the process over. processing terrain and target messages and updating the map.
At the highest two levels of the planner, the problem is Moreover, it would respond to requests from the sonar and the
one of sequencing the activities at that level. This is a host ship.
difficult combinatorial problem because the set of feasible
sequences is quite large even for missions of moderate size. C. GUI/Pre-mission planner
Instead of trying to find the optimal sequence of activities, a
A Graphical User Interface (GUI) was developed to support
heuristic, simulated annealing based approach was used.
the Mission Planner development and was used at sea by the
At the third and lowest level of the hierarchy, the problem
Test Director in defining the missions to be run. The GUI
becomes one of path planning. Path planning encompasses
was developed using TCL/TK, a platform-independent
planning in both the horizontal and vertical planes. The
scripting and graphic user interface development language .
planner first determines a plan in the horizontal plane. This
The GUI provides a front-end for generating activity lists
is accomplished with an A-Star search which takes into
which are then input to the pre-mission planner. The GUI
account the present state of the map and the estimated ocean
has a number of tools designed to simplify the design of an
current. The terrain following planner then finds the path in
activity list and to ensure that its output adheres to the
the vertical plane.
specification The use of a configuration file allows changes
to the specification (e.g., new activities, changes to parameter
B. Task Descriptions
ranges) to be quickly reflected in the operational GUI.
The mission planner was implemented in a real-time, Figures 2 and 3 show the two primary GUI windows for
priority based UNIX environment. The planner consisted of the mission run in the field on April 24th, 1996. Mission
four tasks. They were: specification. Figure 2, has four major components: 1) pull
1) Mission Planner down menus along the top (e.g., the preplanned activities
2) Mission Evaluator menu is shown as a tear off); 2) three boxes listing all the
3) Guidance Interface global, preplanned, and contingent activities, if any,
4) Asynchronous Input Handler respectively; 3) a data entry area for editing the parameters of
The Mission Planner task operated at the lowest priority one selected activity at a time; and 4) a set of buttons that
and was responsible for sequencing the activities of a single affect components 2 and 3 above. This window is used for
level in the hierarchy. It would consider permutations of the entering all non-graphical activity parameters (the graphical
cuiTent plan to see if it could do better. It would only parameters can be entered as well), viewing the parameters of
consider permutations which preserved the ordering all activities of a mission list, deleting and reordering
constraints for the mission. It would also attempt to ail activities, assigning ordering constraints, verifying activity
contingent activities to the plan if their trigger conditions had parameters.
5-128
Mission Display, Figure 3, presents a graphical view of the D. Terrain Following
mission list, which has 3 parts: 1) pull down menus along The terrain following (TF) process supports the mission
the top; 2) graphical display area; and 3) a set of displays and planning requirements of real-time plan generation, safe plan
buttons. This window is used for defining graphical activity generation, and the creation of achievable plans. TF is called
parameters (e.g., the figure shows the new seven sided survey with down track terrain depth information and outputs the
region ready for addition to the activity list); viewing all depth profile that will allow the UUV to best follow this
mission activities including the white Operating Area and terrain at a commanded height above bottom. For most
green Testbox; and displaying the trajectory output from the useful missions, this process is more complicated than
pre-mission planner as an overlay simply adding the commanded altitude to each of the terrain
After the activity file is created with the GUI, it is sent to depths, because the vehicle dynamics need to be considered in
the pre-mission planner. The pre-mission planner is used to
ensure that the activity file, based on a priori environmental
information, is feasible and desirable. It is a replica of the
code executing in the planner and evaluator tasks and
generates a baseline plan based on any avoidance stovepipes
in the activity file. Based on this, a trajectory generator
function outputs vehicle operating speed, operating depth or
altitude, heading and fuel consumption that is overlaid on the
GUI output. The operator can also specify an a priori map
and ocean current estimate that can also be included in the
baseline plan. The combination of the GUI and the pre¬
mission planner was an invaluable tool in the field test. This
allowed for positioning of the host ship during the missions
and accurate estimates of mission duration.
PragZZiQQm j ^ Draw
U ... r •• /' ■.?
■ ......
■ Survey! |
Delete Last Dot J Gear Dots j Qear Figures j j
. .•
h.•:.•' . ..vwJs/f : Figure 3: Mission Display
1 . . '.sorveyAi
.Piwrwfit. order to produce a flyable trajectory. Figure 4, which
illustrates the role TF planning plays, shows terrain in gray,
Cmd.SPteWkU) Oep<lijMI<n> $mdhjSPC<n> j
.. ...'"'p-gg-'..j a dashed gray trajectory exactly offset by the commanded
altitude, and a solid black TF plan that includes vehicle
’ ■ —^1' .
dynamics. The figure highlights some of the tradeoffs that
^26.051502 -80 0B3S34
the TF planner makes in trying to remain at a commanded
altitude, while accounting for vehicle capabilities, as follows:
m. •The TF plan smoothes through the high frequency terrain
at A
JSSSiSfJ I PBiete j '!T-''SrV TE \ •
•Positive pitch limit requires an early climb in order to
Figure 2: Mission Specification clear the peak at B
5-129
if this is a partial plan). If no plan was returned, it would try
again in hopes that more map data had been collected. To
decrease the chance of this happening, path planning
problems were broken up into small pieces. The design of
the planning algorithm makes it more amenable to solving
many small problems than one large one.
Emergency trajectories. In an attempt to limit the number
of situations where the path planner would time-out, the
planner would attempt to insure that there was sufficient time
to plan. If a plan was needed in less than 30 seconds, a
reactionary planner would be called to find an interim plan.
•Vehicle cannot follow altitude too closely from B-C in This planner would only be concerned with finding a 30
order to properly clear pinnacle obstacle at C. second trajectory that avoided obstacles, so it was very fast.
•Vehicle cannot dive too deeply from C-D in order to clear The results of the reactionary planner would be appended to
peak at E. the current plan and then the path planner would be called.
FTP time-out. As a final safety net, the fault tolerant
The Terrain Following software is organized into two processor would monitor the activity of the mission planner.
major pieces: Plan Generation and Plan Access. Plan If the planner did not send a guidance command within 10
generation is called at a relatively low rate with the goal of seconds after completion of the last guidance command, the
providing a vertical trajectory for the next segments of the FTP would assume control of the vehicle and surface it.
horizontal plan. Plan generation accepts the current vehicle While this feature was successfully tested in the simulation
state and arrays describing the terrain over which TF plan is lab, it was never needed at sea.
desired, and the routine outputs the depth and pitch commands While the above features enabled the planner to generate
(i.e., the TF plan) to properly traverse the terrain. TF plan plans quickly, .they did sacrifice optimality. For example,
access is called at a high rate to provide the appropriate depth since the plans for the path planner were decomposed, optimal
and pitch command from the active TF plan given the trajectories were not always produced. Consider the ground
vehicle's current state. track in Figure 5. To plan for the eastern track of the survey,
the planner first found a path to the from the current position
rV. Implementation Issues and Results to the southern end of the track. This path planning problem
did not have knowledge of the future track, so the vehicle was
This section reviews implementation issues that were not properly aligned. The solution from the next call to the
faced in implementing the planner as well as results from the path planner had to begin with a loop to align the heading. If
at sea test runs of the planner. During the test program, the the planner did not decompose the problem for the sake of fast
planner successfully controlled the vehicle for missions up to run times, this loop could have been avoided. Additional
six hours in length. planner design beyond the scope of AMMT would find
This section is divided into subsections that address issues trajectories that are closer to optimal while not sacrificing
in 1) real-time performance 2)The ability to be robust to low reliability.
quality sensor data 3) the ability to generate safe plans and 4)
the ability to create plans that were achievable with respect to
the vehicle’s control system.
A. Real-time Performance
Generating plans in real-time can cause difficulty in two
areas. First, when the planner must expand a node in its
hierarchy (e.g., when it calls the path planner to generate a
detailed trajectory), this function must complete before the
trajectory is to be executed. Second, if the current plan
suddenly becomes unsafe, the planner must be able to replan
quickly so as to keep the vehicle in tact. These problems
were addressed at three levels:
Time bounds on the path planner. Time bounds were based
on the time remaining in the plan. The planner needed to
return before this time, with any safe plan it had found (even
5-130
Figure 6: Map Built on April 24, 1996
5-131
For each grid, the planner maintains an estimate of the mean
terrain depth and variance. The planner assumes a normal
distribution of the terrain which would imply that with 95%
probability, the depth will be deeper than the mean estimate
minus 1.6 standard deviations. This is the terrain estimate
that is used for the grid. The planner would then allow travel
through that grid cell only if this depth were at least as large
as the commanded depth. Any cell that had a terrain estimate
less than the commanded depth was considered a terrain
obstacle. To account for uncertainty in the terrain estimates
and the navigation system, the planner would not get within
three map cells of any terrain obstacle. The planner's map is
initialized with every cell having a very large variance thus
unsurveyed cells are considered obstacles.
While this conservative approach worked well in
simulation, it was not functional at sea. The actual sonar
data did not always provide complete coverage or consistent
estimates so the planner perceived there to be far more terrain
obstacles than actually existed. As an example, Figure 7 Figure 8: Depth Estimates of smoothed data
shows one row of the map built during a diagnostic tow test
As the sonar system identifies possible targets, it notifies
on April 12. This figure shows the depth estimates for
the mission planner. If the target is indicated as a tethered
adjacent 30 foot map cells along an east-west direction.
mine or some other obstacle that the vehicle should not pass
Since the test environment was a fairly level terrain, the
through, the respective cell on the planner's map should be
planner used a technique that minimized the amount of
flagged as an obstacle. This applies to pre-planned avoidance
unsurveyed areas. When the planner received an estimate for a
stovepipes also. A grid cell is flagged as an obstacle by
given map cell, it would use that estimate for any unsurveyed
negating its depth estimate value; this preserves the depth
cells within 100 feet of the estimate. This, along with
estimate at that point. Future depth estimates are correctly
additional smoothing, provided much more usable maps.
incorporated but the path planner always views this cell as an
When this technique was applied to the same data from April
12, the result is shown in Figure 8. obstacle. Once again, sonar, navigation and guidance
uncertainty must be factored in, so the planner does not plan
to get within 3 map cells of an unsafe. An example of a
ground track that goes around a target was shown in Figure 5.
The figure shows the effect of the padding around the target.
In all of the at-sea trials, the vehicle avoided all terrain and
target obstacles.
The mission planner is required to keep the vehicle within a
three dimensional testbox at all times. The only exceptions
are when the vehicle surfaces for a GPS fix. The testbox is
specified as a East and West longitude boundaries, North and
South latitude, minimum and maximum depth and a
minimum altitude. If the planner should command the
vehicle to leave the testbox, the FTP will take control of the
vehicle.
The boundaries of the testbox in the horizontal plane are
Ueated the same way as terrain obstacles so the planner will
not exceed the latitude or longitude boundaries. Additionally,
the planner does not consider any activities that are within
1000 feet of the edge of the testbox to prevent it from having
problems turning near the edge of the testbox. As an
additional safety margin, the mission planner's testbox is
typically specified as being smaller than the FTP testbox for
testing of the mission planner without risking mission abort.
5-132
The mission planner’s depth and altitude limits are also trajectories it created would not be achievable because they
more restrictive than the FTP limits to test the mission dictate a waterspeed outside of the vehicle’s range. By
planner. This is needed to account for uncertainty in the explicitly planning with the ocean current, the planner was
sensors and because it is more difficult to control the vehicle able to generate trajectories that could actually be achieved.
vertically. Typically, the difference in the limits used in the This was critical in generating safe plans
mission planner and the FTP were on the order of 25-50 feet. The mission planner must generate plans that the vehicle
After the test program, these values would be minimized can achieve so that the vehicle maintains the safe trajectory
When comparing the actual depth of the vehicle to the determined by the planner. At the same time, the intent was
mission planner's depth limit, the difference was never more to build a planner that could easily be transferred to other
than 8 feet. The minimum altitude limit was never violated vehicles. Moreover, in order to meet the real-time goals, the
except for one data point during dive 6 of April 25 which was path planner could not model all details of the vehicle’s
apparently a result of noise in the altimeter. capability and guidance functions. The planner only hM
The planner is highly dependent on its map which is created knowledge of the vehicle's turning diameter, water speed
from the terrain and target data that it receives from the sonar limits, pitch limits and the ocean current estimate.
system. If this data should cease to become available, say, To estimate the ocean current, the planner used a two stage
either because of a sonar malfunction or a data process. The planner filtered the ground speed estimates from
communication problem, the planner must take steps to the navigation system to have a smooth estimate of the
preserve the safety of the vehicle. The planner’s guidance current. At the time of path planning, this estimate would be
task has been implemented to monitor the time since the last used to get a safe trajectory and the estimate would be stored
valid input from the sonar system. This threshold may be with the trajectory. When a part of that trajectory was to be
changed at runtime, but was typically set at 30 seconds. If executed, the planner would compare the present current
the time since the last input exceeds this threshold, the estimate with the estimate used in planning and modify the
planner assumes that there is a malfunction somewhere on the water speed or yaw rate appropriately. This second correction
vehicle and does not trust its own map data any longer. If the proved to be useful as the current changed quickly. Figure 9
planner was executing a depth command, it transitions to a compares the planned versus actual ocean current for a 4500
safe depth in hopes of avoiding terrain obstacles. It continues second period on May 3. Figure 10 compares, for that same
to avoid any targets that it has been notified of If the planner time period, the planned water speed with the commanded
is executing an altitude command, it transitions from terrain waterspeed after the change in current was taken into account.
following to altitude following. Monitoring the time since Without this process, the generated plans would not always
the last data input only ensures that some data is being be achievable without violating the vehicle’s water speed
received; it is not sufficient to test for valid terrain data. The limits.
planner actually monitors the time since the last high quality
North
data message. Quality is defined as the ratio of valid data
points to the total data points in the message; a high quality
data message must be above this threshold.
By the time the planner starts receiving valid data messages
again, the vehicle might have traveled to a previously
unmapped area. To provide for a smooth transition back to
terrain obstacle avoidance, the operator can specify the
number of good data messages that must be received before
the planner starts working with the map again. This value
was typically set to 30 seconds.
D. Achievable Plans
An additional artifact of the requirement to operate in real¬
time is that the planner must plan with the dynamics of the
ocean current. The planner maintained an estimate of the
ocean current based on filtered measurements of groundspeed
and estimates of waterspeed. This estimate was used both to
estimate the time and energy needed for an activity as well as
in path planning. The path planner would often need to find a
path that maintained a constant ground speed. If the planner
ignored the effects from the ocean current, many of the
5-133
50 feet during turns. For the same time period on May 3, the
errors are plotted in Figure 11.
*50
-100 h
-150
-200
2.8 2.85 2.9 2.95 3 3.05 3.1 3.15 3.2 3.25
sec
^ X10
Perpendicular Error
60
40
20
i 0
-20
-40
-60 ^
2.8 2.85 2.9 2.95 3 3.05 3.1 3.15 3.2 3.25
sec , 4
5-134
An Integrated Ground and Aerial Robot System
for UXO/Mine Detection
Yutaka J. Kanayama *
Isaac Kaminer ^
Xiaoping Yun ^
Xavier Mamyama ^
Nelson Ludlow ^
Naval Postgraduate School
Monterey, California 93943
Abstract
This paper proposes a semi-autonomous robot system for land mine/UXO search¬
ing/processing tasks. The proposed robot system consists of a land vehicle (called
a rotary vehicle), an aerial vehicle and ground equipments, where the two ve¬
hicles complementing each other to solve the difficulty of the mine processing tasks.
The unique feature of this system is a perfect coordinated tasks done by
the two autonomous vehicles cannot do: The land vehicle wiU do (1) detecting
mines and UXOs in a small area, (2) processing a mine or marking the place if one is
found, and (3) confirming absence of mines in an area if they do not exist. The aerial
vehicle will do (1) global surveying, (2) evaluating the situation by observing the global
situation, (3) guaranteeing a communication path between the ground system and the
land vehicle. The coordination at this level is not attained by a system which has
only one of these. Another major ad^ntage of this proposal is that the rotary vehicle
to be used here has a complete rotational degree of freedom, which will be extremely
useful for the mine processing task. Use of the semi-autonomous robots for the task
of eliminating the 100 million land mines planted aU over is a far better concept than
using human engineers and workers in a long run.
5-135
Figure 1: Rotary Vehicle Trajectory ( u = 20, uj ■= 0.6)
5-138
5-139
Fonboctywth
r»(fo raotiver
oonboioompulot
•tabOyMntocL
ftjof Op6c (rtncduotfi,
nwo^ilomtiAraflcI
poMvr ooc^wrtyi
was not easily adaptable to anything other than the narrow range of conditions planned for
AROD. The Sandia Labs papers also pointed out several types of coupling in the AROD.
The most prominent of the coupling effects is the gyroscopic coupling between the pitch and
yaw axes resulting from the large amoimt of angular momentum contributed to the aircraft
by the propeller. Another dynamic coupling exists between the altitude-rate and the vehicle
attitude, since a loss of lift due to thrust will occur when the vehicle is tilted to generate
horizontal motion. Yet a third dynamic coupling exists between the altitude and roll control
loops, since the reactive torques applied to the roll axis vary as the engine speed is varied.
Sandia Labs also provided data for modeling both the engine and the servos as second order
transfer functions which were used in this task.
Additional information was obtained by Weir [We 88] in wind tunnel testing. This in¬
formation included non-dimensional derivatives for vane effectiveness and non-dimensional
stability derivatives. The report also stated that the control-vane effectiveness is constant
out to at least 25 deg of deflection. Wind tunnel data were also presented to show that
The Coastal Battlefield Reconnaissance and Analysis (COBRA)
Program for Minefield Detection
Ned H. Witherspoon, Bob Muise
James A. Wright
Key>vords: multispectral, mine detection, minefield In FY 93, the SMDG program transitioned to the COBRA
detection, unmanned aerial vehicle (UAV) AID program. InFY 93 a fundamental multispectral video test
bed was flown in a Pioneer UAV at Pt. Mugu California to
1.0 PROGRAM OVERVIEW demonstrate the feasibility of performing beach reconnaissance
The COBRA program is a United States Marine Corps in a UAV for minefield detection. With the UAV
Advanced Technology Demonstration (ATD). The COBRA demonstration successfully completed, a complete redesign of
ATD objective is to design, develop, and demonstrate, in a the airborne subsystem was begun, along with developing a
Pioneer Unmanned Aerial Vehicle (UAV), a passive ground processing demonstration station. The system has now
5-141
m
UAV AIRBORNE CAMERA SYSTEM
DUAL MULTISPECTRAL VIDEO
ADJACENT FIELD-OF-VIEW W/OVERLAP
FORWARD LOOKING SURVEILLANCE CAMERA
GROUND POST-PROCESSING
MINEFIELD LOCATION
t
FIGURE 1. COBRA OPERATIONAL CONCEPT
undergone both preliminary developmental and operational power. Additionally, to minimize the costs, maximum use of
testing. The remainder of this paper briefly describes the system, nondevelopmental items will be made for concept
and presents recent testing results. demonstration. The use of a fast shuttered camera to eliminate
ground motion blur removes the requirement for a stabilized
2.0 COBRA SYSTEM platform, conserving weight and power. By using video cameras
with standard RS-170 video output, commercial video products
The COBRA ATD system test bed includes an airborne such as recorders, digitizers, and video links are readily
passive multispectral video imaging subsystem for data available. Commercial lenses are not optimal for multispectral
collection and a ground station subsystem for real-time system imaging which utilizes the full spectral range of the sensors;
tracking and post-mission processing. The COBRA airborne however, if properly selected, while stHl imposing performance
subsystem uses advanced multispectral video technology to limitations, commercial lenses are adequate to demonstrate
image the ground scene for purposes of mine, minefield and capability. The multispectral bands, based on target/background
obstacle detection along the beach from the surf zone to inland. characteristic, solar illumination, as well as system hardware
Information on the aircraft heading, altitude, roU, and pitch will limitations, are carefully selected to optimize performance.
be combined with the aircraft Global Positioning System (GPS) Figure 1, depicting the COBRA operational concept, shows the
position for location of all ground images. This information is use of tw'o multispectral cameras to increase the coverage swath
encoded on the system video. During testing of the COBRA and a forward looking surveillance camera to assist with the
ATD system, as the aircraft is flying over the test site, limited navigation and obstacle detection.
real-time video is down linked to the ground station for real-time
system assessment and tracking. AH system data is recorded in The COBRA airborne subsystem performs the collection,
the aircraft for automated minefield detection and location after storage, and transmission of video data which includes the
tapes are returned from the mission. The COBRA functional aircraft heading, altitude, attitude and position data which is
system concept broken do^vn by subsystems is shown in encoded on each video stream. Data from each video sensor is
Figure 2. recorded onto Hi-8 tapes for post-mission processing. During
flight tests, selected data is video linked to the COBRA ground
2.7 COBRA AIRBORNESUBSYSTFM station subsystem for real-time tracking and system assessment.
As the COBRA ATD system is being developed for UAV The airborne subsystem consists of passive multispectral
deployment for testing, it is critical to limit size, weight and imaging components configured to output standard RS-170
5-142
AIRBORNE SUBSYSTEM GROUND STATION
• ACQUIRE VIDEO • CONDITION DATA • RECEIVE SENSOR • MINE DETECTION • DISPLAY IMAGES
IMAGE DATA
•RECORD DATA • MINEFIELD • MANUAL BARRIER
• PROVIDE AIRCRAFT • DECODE SENSOR DETECTION IDENTIFICATION
• TELEMETER DATA
NAVIGATION DATA DATA
• MINEFIELD • DISPLAY MINEFIELD/
• PROVIDE AIRCRAFT • DIGITIZE SENSOR DETECTION NON MINEFIELD
ATTITUDE DATA LOCATION LOCATIONS
• TRACKING
video. The video outputs are recorded on a Hi-8 mm, triple-deck multispectral detection but limit the across aU bands focus and
recorder and simultaneously sent through a video switch to the optical throughput.
Pioneer’s downlink video transmitter. The m^or components
are as follows: two multispectral video cameras, lenses, filter In order to locate any detections, the aircraft position,
wheels, a surveillance video camera, Hi-8 triple-deck tape heading, altitude, roll and pitch are encoded on within each video
recorder, GPS receiver, attitude sensor package, altimeter, analog image. In addition, camera information such as exposure time
video link transmitter and other ancillary equipment including and gain settings, filter number, etc. are also encoded in each
power conditioners and custom controller. The airborne imaging image. The aircraft GPS position information is encoded in
subsystem assembhes are mounted on a vibration isolation Vertical Interval Time Code (VTTC) in each video signal. All
platform. Figure 3 shows the current COBRA ATD airborne other information is encoded in bars along the left side of each
subsystem concept. image. The VITC and bar code data can be decoded in the
ground station either during the flight or during the digitizing
The multispectral video sensor function is provided by two process after the tapes are received for processing. Using these
specially configured Xybion Model IMC-201 multispectral encoding methods, a large amount of data is added to each image
video cameras which are aligned to provide a double width with minimal impact on the active area of the image.
swath as previously shown in Figure 1. Figure 4 is a diagram of
a IMC-201 camera. The IMC-201 is intensified and gated for 2.2 GROUND STATION SUBSYSTEM
automatic exposure control. A spinning filter wheel is located
between the camera lens and the imaging plane. The filter The COBRA airborne subsystem supplies video data to the
wheels are interchangeable and each contains six filters. The ground station subsystem, referred to as the COBRA Tactical
filter wheel rotation places a different filter in front of the camera Information Display System (CTIDS). CTEDS was developed
imaging plane every 1/3 0th of a second which is the camera's by the Environmental Research Institute of Michigan (ERIM)
frame rate. In this mode of operation, every video frame is a under contract to CSS, CTIDS performs two basic categories of
separate spectral band. The spectral range of the camera is from operations: real-time functions during flight to provide an
400 nm to 900 nm. The intensifier, which allows for short operator with sufficient information to validate proper sensor
exposure times through narrow spectral filters, does however, operation and to ensure required data is being collected, and
limit the spatial resolution. The camera functions are micro¬ post-mission functions using recorded mission data to detect
processor controlled. The output from the camera is standard minefields and report their location. The COBRA processing
RS-170 interlaced video, which will be recorded on a Hi-8 video chain implements image analysis functions which currently
recorder. Select commercially available lenses used with this operate on a near real-time (NRT) processor. The NRT
camera provide spatial and spectral resolution adequate for processor wOl eventually be replaced with a processor capable
5-143
FIGURE 4. XYBIONIMC-201
5 -
of producing results in real-time. While there is no ATD has been demonstrated, a high performance Sun UltraSPARC
requirement for the COBRA to perform automated minefield computer was used for CTIDS NRT processing. However, the
detection in real-time, the NRT processor was implemented so RTF uses high-end personal computers making operation and
results can be achieved in a timely fashion to aid performance maintenance of the system easier for a wider range of people.
assessment during the formal testing phases of the program. The COBRA ground station subsystem concept is shown in
Figure 5. Figure 6 is a photograph of CTIDS. The CTIDS
The CTIDS is organized according to these operations into ground station subsystem includes a data link receiver, a
the RTF subsystem and the NRT subsystem. The RTF provides ruggedized 486DX2-66 computer CTIDS controller, a
CTIDS with the capability to record COBRA multispectral 486DX-50 laptop computer and docking station for vehicle
imagery from a real-time video downlink, digitize recorded tracking, a Sun UltraSPARC processor computer for NRT
imagery and archi\^e images with required ancillary information. processing, a computer controlled S-VHS recorder player, a
The RTF also provides video image review capability and video monitor, two color monitors and an Uninterruptible Power
analysis functions useful both during real-time operations for Supply (UPS).
assessing flight operational effectiveness and sensor
performance issues, as weU as during post-mission operations 3.0 COBRA PROGRAM STATUS
for manual location of obstacles and fortifications. The RTF also
displays real-time aircraft position updates overlayed on a The COBRA Preliminary Design Review was held in
variety of maps, satellite images, and other mission data for September 1994 and the Critical Design Review was held in
monitoring search patterns along with downlinked video
January 1995. Preliminary developmental testing (DT-0) was
imagery. The software can also be used to display and print final
conducted at Eglin AFB, Florida and Camp Lejeime, North
post-mission minefield detection results along the platform flight
Carolina from May through August 1995. Preliminary
path. operational testing (OT-0) is currently being conducted at Camp
Lejeune during November 1996. The COBRA system is
The CTIDS NRT subsystem provides automated minefield
meeting or exceeding ATD goals in all background
detection using COBRA sensor imagery. Each spinning filter
environments tested to date. COBRA will demonstrate is
wheel multispectral video camera collects six bands of
operational utility, along with other countermine systems, as part
multispectral imagery sequentially at a camera frame rate of 30
Hz while in flight. Each of the six images are automatically of the Joint Countermine Advanced Concept Technology
registered for multispectral processing for minefield detection. Demonstration (ACTD) in FY 97 AND FY 98.
ERIM developed an image-to-image registration algorithm^ for
COBRA which is capable of determining and performing 4.0 DEVELOPMENTAL TESTING
complex coordinate transformation between images and
subsequent image resampling to achieve subpixel registration Preliminary developmental testing (DT-0) was performed
accuracies. ERIM also implemented an adaptive multispectral according to an official COBRA DT-0 Test Plan developed by
Constant False Alarm Rate (CFAR) mine detection algorithm^ CSS and ^proved by the Marine Corps. This plan specified that
which exploits spectral and spatial target signatures for automatic testing would be performed using several different minefield
target detection. Once mine-like targets have been detected and test arrays set up in various coastal environments. The test sites
located using data from the airborne video tapes, a linear density included a benign coastal environment at EgUn Air Force Base,
algorithm for patterned minefield detection^ developed at CSS a very cluttered coastal environment at Eglin, a cluttered grassy
performs minefield detections. Capability is also incorporated field also at Eglin, a moderately cluttered coastal environment at
for unpattemed (scattered) minefield detections using an Marine Corps Base Camp Lejeune, and a homogenous grass
algorithm originally developed under contract by The MITRE field also at Camp Lejeune. These test sites were selected to be
Corporation.^ Besides automated detection, the operator also representative of a range of backgrounds from very easy to very
can use the video from the surveillance camera as well as the difficult. For each background environment, several missions
multispectral cameras for manual identification of obstacles. All were flown under varying conditions including time-of-day,
detections, whether manual or automatic, are tagged with altitude, airspeed, type of minefield, and other conditions. In
ancillary position, attitude and other information so that
total, over 200 flight hours were flown during approximately 40
detections can be located and tapes automatically repositioned to
DT-0 test flights.
raw imagery corresponding to the detections, if desired by the
operator. Two types of minefields were used during DT-0 testing: a
Because of the level of complexity of the COBRA processing staggered row patterned minefield and randomly scattered
minefield. The former was used for the majority of test flights.
chain, and the desire to not use specialized image processing
The density and distributions of these minefields were chosen
hardware until the automated detection processing technology
5-145
• PLAY BACK • normalize and MONITORING
FLIGHT DATA CO-REGISTER
► MANUAL BARRIER
IMAGES
IDENTIFICATION
5^146
to be tj^ieal of deployed minefields. The COBRA sensors were The nex^ test site analyzed was an extremely cluttered coastal
flown over these minefields in a Cessna 172 aircraft acting as a area at Eglin. This test site presented many technical challenges.
surrogate to a Pioneer UAV. This afforded more flexibility and The water was stagnant and dark right up to the shore. The
less cost in testing since the Pioneer is a fleet asset. The Cessna sandy region at the water’s edge was only 1-2 feet wide. The
was flown at an altitude and airspeed corresponding to those water/land interface had areas which were rocky, as well as areas
used in our previous Pioneer testing^ and for the current OT-0 which contained metal trash, concrete, and other manmade
testing in the Pioneer. The COBRA automatic minefield debris. The inland area of this test region was no less cluttered.
detection processing chain^ was exercised on data from these There were areas of patchy grass, trees 10-20 feet high, small
flights to estimate a probability of detection versus probability of scrub bushes and a couple of areas with just benign sand. The
false alarm cur\^e, called a Receiver Operating Characteristic inland areas changed backgroimd characteristics very rapidly,
(ROC) cur^^e, for each type of environment. To estimate the highlighting the need for localized measurement of background
ROC curv^es for each test condition, at least 25 minefield statistics for the mine detection algorithm to be successful. Also,
decision regions were collected and processed at each test field because of the very diverse backgrounds, the need for a variety
along with at least 25 non-minefield decision regions over the of spectral bands was apparent since the optimal bands for mine
same test field. This provided a statistically significant number detection are different for each type of background and there
of samples from which to accurately estimate the ROC curves. may be two to three very different backgrounds appearing in
After these 50 decision regions were collected and processed, the each multispectral image. Typical performance results from this
ROC cur\^e was estimated by varying the threshold values for test site showed that the patterned algorithm successfully passed
each minefield detection algorithm. Computing these values for the detection and false alarm goals, but just barely. The scattered
a given threshold results in a single point on the ROC curve.
algorithm, however, was even more successful.
Thus, vary ing the threshold from the minimum encountered
value through the maximum value results in a complete ROC
A variety of other test cases were processed from the DT-0
curv’^e for each minefield detection algorithm.
data^, including processing a 40 second long pass covering a
minefield and many different background environments
The first test site anal}^ed was a benign coastal area at Eglin. occurring along the aircraft’s track. A preliminary pass of OT-0
This is a very clean beach with no significant environmental data which is over a minute-and-a-half long has also been
clutter. The beach is on the Gulf of Mexico with extremely processed. In aU cases, the COBRA system is meeting or
white sand and t}^ically relatively little wave activity. On the
exceeding the ATD goals for the program.
sand, the miues were clearly visible in all bands of the
multispectral imagery and appeared as dark objects in an
5.0 ACKNOWLEDGMENTS
otherwise highly reflective scene. In the water, the mines were
generally visible through the surf foam as well as at a depth of
The Coastal Systems Station, Naval Surface Warfare Center-
about 5-6 feet of water. In this type of non-cluttered
environment one would expect highly successful minefield Dahlgren Division serves as the Technical Direction Agent for
detection results. The ROC curve for this environment is COBRA. ERJM is the prime contractor for the program. Mr.
presented iu Figure 7 and shows both the patterned and Ned Witherspoon is the Project Engineer and Mr. Bob Muise is
scatterable algorithms have excellent performance. Selecting a the Image Processing Lead for the program. Dr. Jim Wright is
typical operating point on the curve results in the patterned the lead engineer for the ground station development. The
algorithm achieving Pd = 0.86 for a Pfa = 0.02 while the program sponsor is Mr. David Vaughn, Director, Amphibious
scattered algorithm achieves a Pd = 0.94 for a Pfa = 0.07. Warfare Technology, Marine Corps Systems Command.
5-147
5/2/95 first run -- initial results
for declaration of target area
3. Robert R. Muise and Cheryl M, Smith, "A Linear Density 6. Robert R Muise, James A. Wright, and Quentin A. Holmes,
Algorithm for Patterned Minefield Detection." Proceedings “Coastal Mine Detection Using the COBRA Multispectral
of the SPEE Land Combat and Law Enforcement Sensor.” Proceedings of the SPIE Mine Detection
Technologies, Orlando, Florida, April 1995. Technologies, Orlando, Florida, April 1996.
5-148
Clandestine Mine Reconnaissance -
Unmanned Undersea Vehicles
5-149
'/-i
E
CD 0
o w
q: =L
o d)
c c
W-i^l^$-’r'\
i o
CD >"
£ CQ
00 0)
S’ o5
o
S cn € s
(D <
CO S S ^ H O
S S
S p s *^* T3 "O
O O
•S
I ^ 4-*
C
0
</>
4-»
C
O
</>
U Cs^ 3 > O
^
0.
O
V.
a.
'*^^<^p'-|jr;,,
'■-'"^.1-,‘ '"'■.?\^^ "^'v%\y
1®lffe-’
y%'^l^*'<''"\^^ ’>^1^
'" ■t<A'
A-UUV-U-0463- 2 - 11/21/96
HKtc:"j^.V, .■,i!f'JjAigiii^iSa^sk£jii- _
0 o
P c
^ C8
C (A
“ W
cu
CT)
O
*■*
/i«
0)
Q. C
<D ■=
a
Gh
O E
Q> O =
'X
« JR 0
O <A c 4^
s I
■S =
0 Q.
■o
<» I
<0 (0 iS > 2
■> O) ^ o Jr» ®
o o q.;l ® >
0. a o o O "D
5-151
Navy UUV Program Plan
5-152
Both programs are on track
5-153
A-UUV-U-0463- 4 - 11/21/96
Mine Warfare Plan
5-154
A-UUV-U-0463- 5 - 11/21/96
Near-term Mine Reconnaissance
prototype capability
NMRS Concept of Operations
Deployment Concept
A-UUV-U-0463- 8 -11/21/96
NMRS System Architecture
CO
3 __
O CD
<0 O
S; >
g) -
0
LL .2
i~ 0
g ^ o
C O S ^ 8
o ^
^ I ^ CD
05 M= CO c: w O ^
c W S^ S 0^0)
O w O) od
2 o
CO i5 .E 0
> ^ 3
J5^ :3
o o 2 g.^
if FODL breaks
^
fi
05
W
§1
§ ^
-2 SCD t” 0 C
O C ?-5 ^ CD 0
o
5 -i
.b
^
o
1J=
E - ® ^ c
i
O w S §.9
"2 o o 0 C5) >
05
05 0 _l CO C 3 Q. «
•D </)
-^ vs i ^
0 O E => o z §. 3
o :j- o • :2 • o • 0> CO §■ E
^ C\J U. X < CO <0 E
o I
o '
5-159
Back-up A-UUV-U-0463-11 -11/21/96
Control and Display Consoles
A-UUV-U-0463-12 - 11/21/96
A-UUV-U-0463- 13 - 11/21/96
0
0
0
o
0
>
o CD
■Q
o C
o *i_
0 o 0
1-
0 ”0
> 0
0
D X
c D 0
-»-*
O ^ CD
w O 0 O
c O “0
0
0 S2 Q.
^ o
0 _0
ig s 0
XI
0
O
C
O
O)
o c 0
0
c li<;
o 0)
0
-I—»
o 0 o
c ■0 .£ :0
0 “D g> z >
0 0
E i 0 0 C
O D) c p_ o
CD 0
c v_
^■i
o
X ■— 0
O o 0
o
0 Q.
Q_ O
u.
c >.
jh-
§,1
T3
C _0 HO
— 0
= O 0
0 ”0 >
Q- C 0 0 O
0 O £ ■0 0 O
Q O o
c
< ZD 0
I I I
5-162
A-UUV-U-0463- 14 - 11/21/96
A P'.. * „JL
|MS,v^%
t#4i ^
^ $ '3s« < i ^ '/' ■
I U/'t 'A^'d M
iW<* ^.,y ■* I
. .
■ X ;r/ /:x
fe# Jfy ""^r/ ^7y&A
m'^ i// ^ /y't yy."^ > S ;<»>',< /"yfy ,, /' /' y/y'y/f// '‘1"'/^,'''V'y^y^Ay
^'7x07 '<}
f'*n‘tl!v"5’4'/*;.T^:f^. 'A';T Xysv
5-163
^•UUV-U-0463-
LMRS Performance Specification
Scenarios and TACSITs for
LMRS COEA
11/21/96
11/21/96
Adecjuacy of Total Area Coverage Capabiliti
(per SSN) for Various TACSIT Conditions
A-UUV-U-0463- 22 - 11/21/96
Applicable Science and
Technology Efforts
O 0)
w -'
> o
fWS!
nl). W«'
Q
sJ*
CO w
o
< ;o..2...-. -
e- —' V. . ^
-J O iW ^
?! z
o #€'''■>?§
(/) ■•'wrv-''r
OC
3
O
0> X
01° 2
t2l:sr»Vi«;;s:G
c ^
lOJ.P#-'
I5SW/:
;«}>.'■ 1% , CQ cf:|^
K
'-
5-175
CD
CD
JZ a.
^ § 0>
^ o
0 O ^
0 05 w
3 0
3 ■Q 0 o c O
O £ x: N ■:::
s O
0 E c ■U c
O o d E 3 <D
1— 3
L. M— 3 «»
■D O
XD c c
0
“o :x
=: O
c •♦-«* CO ^ § « CB ^ •c .S' >.
0 2 If-- CO © = >.
o CO c I*-
(0 Q. O
Q. P
o) .2 ©
s-■ c
■ e ^
^ = <B ■E ©
C E
CD
w ^ 0) ”* o
.2 z
E .52 O TJ "O
•a ■ii«
Q. W E CD 0
c - ^2 ■*■* fl) I S 2 . <0
o ©
CO
C O) ^ W/ -Q s> o S
S o m
3 3 0 0) ^ C
S: CM ^o 2 i E E
o ^ Si
■S o
(0
c u. it: a a .1? < IM
LU ^^ (3
X ©
11
^ o
c c
O o . 43 0)
c ••i. LU gl ? = P
■■■
D) O _-
ns ;a < !C « o o
3 £ =
■= O 0)
O d) x: o =5 e o o
•£i S < CO O
<s -S
i-^ +i
^ s
o
» &
CO 2 Ql :; %=
in -5^ O) 3 *5 TJ
x -I .£ -^
3 ^ ^ ? <D ®
< ^ i ?
S O^ E'-1 ® 0) w « C «
K KT'- O i 5 TJ fc. J. (0 <0
c-= o o
uq □ So i 3.S
o 2 © JS
v>^S
3
E
(5 .^ © = o
0) g
o X C ±i
o © n © o S s
5so * '*- £ CC s:
o O o ^
t: 3 © > ©
O) £
< c
^ c
O 2= X ^
|a 2 i 5= 3 .ti c I Q
>' 52
© 0) © o
m »» ^ "Ja
0) vi" o .5 ® 8
.E E c § ^ <0 8
O) 3 5 %
c: V- C lo >, 2 0-^
© £ O O © CT lo 0) £
■ mtmm
Ills: >.1.1 • O ^
^ ^ JQ O) ^ HI <D
© o E © ts
fc o X S o C © I
ic 111 ui I* H c o © .a
ofrSoc X
Mbc
Q. iS
Olh UIO
”i s*
P i
irirv Technologies
A-UUV-U-0463- 28 -11/21/96
A-UUV-U-0463- 29 - 11/21/96
J5 E
a. 0
-t—*
(/)
E >» CO
(/) o
E o
CM
o
§ O)
P ^ 00 o
o> o
t o >
TJ 0
0 c 0
>. ^ TJ CD 0
:l.
JQ 0 ■D 0
0 0
0 O
Ql JC 2 I i 75
c o
O CL
0
o 0 c g-
0 o 0 ‘■f—• 0 E I—
O O 0 o
"D c ^ O
H—*
-Q ^
o 0 >. O 0 f CO
(/> 0
0
O 0 "S -S
3 0 3 0 ■0
O 0 0 O E « O- ■■■
o C Q. C 0 <u u X
c/^
O 0
0 hr I? 0
i>5
O O = D O) 5 3 ^ ^ <i>
o ■u ™ -O 0
0
0 "D 0
0 0 3
0 >
tr JO f i2 1
(0 C O o o >
0 CL 0 E o 2^ 0
I O Q.
D
Q. .E C 0 0
c 4-'
o
0
0
0 ^
^ E Q
CD ^ = O
VW 'w'
0 0 o cc O)
> 0 CL 15 ? Q o 0 0
O 0 ■O O) 0 C
•4—' ©a <0 c (/> c■MH c o
^ o 0) O CO CC cc o a: O o LJJ CO
(/> z s S 1
clT co" 1 •l-i«
1 !
D a z (/>
5-178
The Iguana:
A Mobile Substitute for Landmines
Summary:
There are a number of sites where WWII war gas deposits have been placed on the sea bottom. In the
Baltic Sea, there are two such areas -- one situated east of Bornholm and the second west of Gotland ~
together containing 35,000 to 50,000 tons of WWII war gas ammunition of various types [1], [2], Due to a
number of reasons, there has been no systematic evaluation of the condition of or any regular monitoring of
these deposits. One of the main obstacles to long-term monitoring of environmental impacts of war ps
deposits is their unknown three-dimensionl distribution in the sea bed. Another major obstacle to setting
assesment procedures is the cost of information gathering. Although the utilization of surface vessels is very
costly, it can bring detailed results and such application is justified in a large-scale projects given sufficient
financial support.
Development of new technologies for ocean floor documentation, particularly AUVs and their work
packages, creates new possibilities for cost-effective reconnaissance of targeted areas which are known
locations of war gas deposits. Application of precise surface and underwater navigation allows for
periodical survey of selected sites. New search and classification tools available on the commercial
market can produce more data of better quality. The safest way of cost-effective information gathering is
through the application of robotic systems for step-by-step data collection. Following a survey of selected
points, areas for more detailed investigation can be selected. The type of information required includes
target mesh coordinates, ammunition type, corrosion status, and danger status, and can be obtained from
data collected by dedicated AUVs performing search and data collection missions.
Problems confronting the investigation of war gas deposits are not fully
recognized to define the best monitoring procedures. From historical data, it is known
that such deposits contain metal containers, barrels, grenades, artillery shells, and
bombs. The grenades, shells and bombs contain explosives, and sometimes detonators
and chemical agents, so they are very dangerous to handle due to the potential risk of
leakage or explosion. The fact that it is not known which kind of the chemical agent
has been placed inside (mustard gas, adamsite, LOST, Zyklon B or other type) makes
all handling procedures even more dangerous. The largest deposits are east of
Bornholm, at least 35,000 t (Bornholm Bassin); and South East of Gotland, about
2,000 t (Gotland Deep). Both areas are valuable fishing grounds. Due to this fact, a
majority of incidents involving chemical war agents took place during recovery of
fishing gear, where accidentaly caught ammunition carnet into contact with humans.
According to information obtained at experts’ meetings, Danish fisheman have been
catching 1 to 3 tons of war gas ammunition each year [3]. In a majority of these
incidents, the dangerous catch was returned back into the sea immediately after
discovery. In certain cases, the ammunition was again dumped into specially selected
areas. [Source: Cmdr Soetofte Rep, Danish Navy, 1992). This latter approach agrees
with expert and scientific opinion that refraining from recovering war gas ammunition
from the ocean is the best and most harmless way to keep this war heritage relatively
safe [3]. 5.181
Having in mind all the security precautions, the survey of war gas deposits could be
done in different ways. A typical scenario should consist of preliminary investigation by
surface survey vessel. Mapping using a ship mounted or towed sonar system could
collect information for AUV survey route planning. A bathymetric survey could bring
valuable data about bottom three-dimensional coordinates which could help to plan
AUV paths optimal from a data collection point of view. After analysis of collected
data, potential targets could be selected for detailed AUV mission planning. Very
important is to learn as much as possible about bathymetry, geology, geophisics, and
oceanography of the investigated area. Precise mission schedule and motion energy
consumption could be optimized, bearing in mind environmental conditions and specific
data collection requirements. Moreover, the knowledge of AUV performance
characteristics is crucial to mission planning procedures.
Task definition
From a practical point of view, data collection methods are similar to those used in
mine countermeasures procedures. The difference is that bottom laid mines are often
bigger and amagnetic. All information about those dangerous deposits could be
classified into two groups:
The most critical factors for AUV task definition and mission planning are
mission energetical requirements. State of the art allows for application of newly
developed power sources with optimal energy density. Given a vehicle with certain
dimensional restrictions and resulting weight/buoyancy figures, the overall AUV
hydrodynamic and energetical data could be evaluated. Estimating potential work
package requirements, like weight/ power characteristics, against available payload and
power, the detailed mission definition could be evaluated and optimized.
The basic characteristics of the sea environment in the planned area of AUV
operations is given below:
Underwater robotic systems are able to perform a variety of different tasks. Sea
environment monitoring and suitable data collection are possible due to the large
variety of equipment available and well known procedures (like STD, CTD).
Application of tethered vehicles in sea environment monitoring is growing steadily.
Particularly, towed undulating vehicles are used more and more frequently. On the
other hand, tethered ROV’s are used mainly by offshore industry. Their application for
on line sampling is recognized for cases where other systems and methods fail (dunking
bottom and water samplers, divers). Chemical analysis of water is a complex task and
requires specialized instrumentation and clean laboratory conditions. Tracing traces of
specific products of decomposition of chemical agents from gas ammunnition and
containers underwater is a challanging task that cannot be done remotely. Obstacles
include lack of instrumentation and recognized procedures. Water quality monitoring
using autonomous vehicles could be conducted off line with the help of special water
sampling devces preprogrammed and controlled by an AUV control algorithm.
Selection of the optimal data collection procedures during war gas deposits
monitoring requires different criteria to be analysed. The most crucial of these are:
1. Data quality
2. Data amount
3. Real time data availability
4. Sampling system energy consumption
5. Risk of data loss
6. Risk of loss sensor carrying platform
7. Mission duration time
8 Operational costs
9. Sampling feasibility
1. Collection of visual and sonar images during circular cruise done as close as
possible to the seabed in a predefined area where possibly high surface
concentration of the war gas deposits has been detected by surface survey
vessels.
2. Collection of bottom images and water samples from a distance closest to the
located dumps using AUV water sampler fi-om predefined locations (nodal
points) lying on preprogrammed survey route. Water samples will by analyzed
off line by a specialized chemical laboratory for traces of chemical agents
resulting fi-om chemical degradation of war gas chemicals.
The general arrangement of AUVs is given in the Fig.l., and general characteristics in
Table 4.
Range 1 10-20 nm
Options Fiber optic data comms link for surface vehicle control
Two potential missions scenarios are given in Fig 2. and Fig. 3. In both cases,
a reference Ultra Short Baseline u/w transponder with acoustic release is placed on the
bottom in the selected area. U/w navigation system position data output is related to
surface based DGPS system on board the support ship. 5-185
Other operations could be conducted using AUVs, but this requires more
specific analysis. An example of such a task could be collection of bottom samples in
close proximity to war gas shells or containers. While there is an evidence that
mustard gas could be found as separated lumps, there is a quite high potential risk that
very dangerous samples could be collected in the automatedl sampling mode. The
vehicle could be contaminated with toxic agents and suitable decontamination
procedures must be carried out following the completion of each mission on the deck of
the surface support vessel. Ground sampling using a tethered ROV manipulative
system are reccomended for such operations, but this is a separate and very complex
problem from a safety pont of view. Some chemical shells contain detonators and
explosives, and no one knows the true content of objects of interest.
There are diffrent ways to conduct the task. Data amount and quality are
determined by mission requirements and the tool set specified. Potential mission types
can be identified as follows;
The instruments specified and integrated into the AUV system should have an
energy consumption as low as possible and data recording and storage capabilit as high
as possible. The following data collection formats are envisaged. Digital data storage
is optimal for post mission processing and data fusion. In certain cases (continuous TV
path recording, still pictures), analog data format provides some advantage due to
resolution and amount of information content. All data should have suitable position
and vehicle status references.
While war gas dumping sites contain explosives and dangerous chemical
compounds, suitable means must be envisaged in the following emergency situations;
1. Recovery of bottom sediments due to accidental collision with war gas lumps
2. Detection of contaminated water samples
References;
5-187
The Lemmings/BUGS System
Amis Mangolds
Foster Miller, Inc.
Because a full paper was not received by publication date, the above Abstract appears in this
Proceedings. Mr. Mangolds can be contacted at Foster Miller, Inc., 350 Second Avenue,
Waltham, MA 02154-1796; telephone 617-684-437; e-mai <amangolds@foster-miller.com>.
5-189
Application of the Explosive Ordnance Disposal
Robotic Work Package to the
Clearance of Terrestrial Improved
Conventional Munitions (AutoRECORM)
Gary M. Trimble
Lockheed-Martin/Ocean, Radar and Sensor Systems Division
Because a full paper was not received by publication date, the above Abstract appears in this
Proceedings. The author can be reached at Lockheed-Martin, telephone 408-742-7596.
5-191
5-192
CHAPTER 6: COUNTERING
MINES ON LAND
The papers grouped in this Chapter deal with technologies and systems that seem most
relevant to the military problem of Land Countermine operations. Many of these applictions could
also be relevant to Humanitarian Demining, or to the companion military problem of countering mines
at sea.
Technologies and systems for countering mines on land must advance the objective of
reducing the risks from mines to levels commensurate with the risks to a military force from other
types of enemy action. Furthermore, the approaches must help meet the stringent time parameters
that attend minefield breaching under conditions of land combat. Even in what is called
"administrative countermine operations" (clearing of rear areas for depots, etc.), field operators often
face severe time constraints.
In general, systems for the land countermine operations will be operated by trained military
personnel. While administrative operations may be in areas relatively immune from hostile fire, the
breaching operations are assault operations. The expectation that trained military personnel will be
the users helps the developer know what kind of logistics support and maintainability criteria must
be met.
Candidate technologies and systems for the Land Countermine operations must demonstrate
capability in most, if not all, physical land environments. What works in the desert may be totally
inappropriate for application in Bosnia’s forested regions. Sensors are particularly sensitive to the
physical environments in which they are employed. The effects of the environments will be
demonstrated during the exercises that constitute the Advanced Concept Technology Demonstrations
(ACTDs) slated for '97 and '98.
The physical environments of Mine Warfare so dominate the technological solutions to the
Problem that the 1998 Symposium on Technology and the Mine Problem will call for papers on the
land and sea operational environments.
As in medicine, there is no approach that meets all requirements — no "Silver Bullet". Rather,
there must be sets of tools that, in the hands of the military engineer, can be used to carry out the
necessary operational tasks.
6-1
Opening Remarks by Chair of Session XV,
“Systems and Technologies
for Countering Mines on Land”
6-3
- ''Let Us Try
o
a
CZ3
c/3
<D
ESSAYONS
‘S)
•1-H
o
c/3 o
"o C/) o
o
<D c/3
a
H <D (D
+->
a (D
<D
bJO cd
C3
UNITED STATES ARMY ENGINEER CENTER
§ Ph
13
p
c/3 o
psi c3
<u <D
c/3
Ph
O
6-4
UNITED STATES ARMY ENGINEER CENTER
STATUS OF COUNTERMINE
MATERIEL SYSTEMS
• CURRENT SYSTEMS
PROBE
METAL DETECTOR
ROLLER
FLAIL
ESSAYONS ‘IHUtTo’
6-5
UNITED STATES ARMY ENGINEER CENTER
• GAMBLE
6-7
- UNITED STATES AHMYENGnOIEP-CEKTEH
i'iSjf
HAZARD IDENTIFICATION
ENVIRONMENTAL CONDITIONS
GROUND SUBFACE
MOimjRE CONTENT
SOLARLOADING
CLUTTER
TEMPERATURE
MINE TYPES
AN’TI-TANK
anti-personnel
FUSE
METAL content
anti-handling devices
TRIPWIRES
6-9
Objective in this process is to evaluate the hazard
to determine the probability of it occurring and the
severity/effects if it does occur.
6-10
An artificial intelligence system is needed to
assist commanders in process the large guantity of
information needed to optimize risk management.
6-11
HOLISTIC APPROACH TO
RISK MANAGEMENT
• INDUSTRY
• TACTICAL COMMANDERS
• SOLDIERS
6-12
, -. UNITED STATES ARMY BNOa^ER CEJ.’^ER
• ARTIFICIAL INTEL/EXPERTSYSTEM
b-13
UNITED STATES ARMY ENGINEER CENTER
,C/3
<D
Oh
6-15
ESSAYONS - •■UtUsTry"
UNITED STATES ARMY ENGINEER CENTER
6-16
ESSAYONS - "Let Us Try"
UNITED STATES ARMY ENGINEER CENTER
6-17
ESSAYONS - "UtUsTry"
UNITED STATES ARMY ENGINEER CENTER
6-19
Combat breachers
UNITED STATES ARMY ENGINEER CENTER
<D
00
6-21
First aid procedures/supplies
UNITED STATES ARMY ENGINEER CENTER
C3
CZ3
c:S
c
6-22
II
2: CO
(D
<D
T3
Oh
6-23
45,000/7.5 = 1 FA/6000m2
UNITED STATES ARMY ENGINEER CENTER
6-24
1500 X 5 minutes = 125 manhours/hoiir
UNITED STATES ARMY ENGINEER CENTER
II'•
’T3
8 N
^
(D
a
<D
§ 5
c/3
cn
in
cd
o
6-25
Vh
O
o
Abstract
practically feasible, as they will also be
DeTeC (Demining Technology Center) is sensitive to ferrous soils, leading to the
developing a sensor system for detection of smaller debris and augmenting
humanitarian demining, which reduce the the false alarms rate. Nowadays, once an
number of false alarms and can be carried alarm is given by the metal detector, the
by a man or an autonomous lightweight soil is prodded at a shallow angle using
robot. rigid sticks of metal to determine the
shape of an object; this is an intrinsically
The objective is to reliably recognize dangerous operation.
minimum metal antipersonnel mines. A
metal detector is used to recognize the The need for new, efficient and affordable
location of objects with some metal demining technologies and sensor systems
content. A GPR then provides an image is therefore obvious. An overview of the
that allows to differentiate a mine from current research status is given in
metallic debris. The efficiency of deminers CMaechler95] and [Gros96]. Past
using the combined detector should conferences dealing with this problem are
increase significantly, and the database that listed in [Nicoud96b].
can be built at the same time is essential
for further steps in automating the search
process. 2. The DeTeC test system
Initially, deminers will look at the GPR Extensive tests in a "sand box" are
images as an optional information, not required to develop the filtering and
changing their SOP (Standard Operation recognition algorithms, under repetitive
Procedures). They should progressively get conditions. Two containers have bieen
confidence in the displayed information, built, one filled with sand and the other
which they can relate in real time with with loamy soil; they are 1 metre deep
the result of their prodding. and 3.5 by 3.5 metres wide (Fig 1). A
cartesian gantry positioning system allows
1. Introduction to move the sensor above one of the
containers at a time. Vertical motion is not
The metal detectors currently used by controlled: the sensor is set at a fixed
demining team cannot differentiate a mine height, or a spring adjusts the pressure on
from metallic debris, which sometimes the ground. The step mo :tors control
leads to more than 1000 false alarms for box receives displacement orders from a
every real mine found. Although the serial line, and the acquired data is stored
detectors can be tuned to be sensitive on a PC's disk and transferred later to
enough to detect the small amount of some server. Most of these measured data
metal in modern mines, this is not files are available on our Web site.
6-27
APs with a diameter of 8-10 cm. Smaller
mines might require correspondingly shorter
wavelengths, which will shorten the usable
depth range too, but they are also buried
closer to the surface.
A. Hardware
B. Software
2. GPR selection
Software embedded in the radar is limited
Current GPR systems are still way too to some basic functions, mainly designed
expensive to be used in large number for to improve the image quality and it is not
humanitarian demining, such as it is now sufficient for antipersonnel mine image
done with metal detectors. But we hope analysis. Affordable GPR software for
that prices will fall when the efficiency real-time applications seems not to be
for mine detection will be proven and available on the market. Systems developed
when the manufacturers will realize the for military use are often mentioned, but
potential market available. are usually either classified or prototypes.
A GPR for antipersonnel mine detection To start with, we have selected the
must have a wide frequency band to Reflex seismic off-line processing package.
achieve a good resolution, but since higher Several modules are available for data
frequencies do not propagate well, the analysis. Algorithms not included in Reflex
chosen range is always a tradeoff between are developed using the Matlab
resolution and penetration depth. For environment. Now, all the required modules
antipersonnel mines (AP), a center are rewritten in Matlab, to allow us to
frequency of 1 to 2 GHz, and a bandwith evaluate all the modules of the data
of the same magnitude, seem to be a processing chain. The next step will be to
good choice for most types of soil and for rewrite all the routines for a fast DSP.
6-28
C. Data Visualization respect to this approach, the GDIS project
at DASA-Dornier [Borgwardt95], in
Different visualization techniques are being cooperation with the Foerster company, has
evaluated to find the most suitable one, demonstrated encouraging results.
from a practical and computational point of
view. One has also to bear in mind that The Foerster Minex 2000SL metal detector
in the demining case, GPR data will have generates two continuous wave
ultimately to be interpreted by non expert frequencies, f1 and f2, at 2.4 kHz (for
personnel. The most common GPR data ferromagnetic objects) and 19.2 kHz (for
visualization consists in displaying the data stainless steel and alloys) respectively. To
as a vertical slice (Line or B-scan), whilst fully exploit the detector's capabilities we
moving the antenna along a line on the intercept, at the output of the
surface. receiver-transmitter module, four signals
corresponding (in the complex plane) to
If the real size of the buried target is the real and imaginary parts of the analog
needed by the recognition process, pulse signals f1 and f2 induced in the receiving
deconvolution and migration algorithms will coils.
be necessary to transform the target
response into a more compact one. We
are still looking for a robust and fast 4. Results
algorithm which must be able to work on
cluttered images. As soil characteristics The response to the minimum metal mine
play an important role in the migration (containing only a striker pin of 0.1 g!),
aperture, it will also be useful to develop a metallic debris of about 2 g and a stone
an adaptive algorithm. (Fig 2) have been compared (Fig 3).
Results are convincing, but at time of
In order to distinguish an object's shape it publication, the data acquisitions have been
might be necessary to display horizontal made in the sand box only, that is in a
views of the ground at different depths very clean environment.
(Area or C-scan). In this case it is
necessary to combine data from several
parallel scans. The distance between two
parallel scans is an important parameter, in
order to reconstruct the real shape of the
buried object. Parallel scans are performed
each 20 mm, with an acquisition each 10
mm. In order to improve the resolution we
Rock
take a second set of measurements
orthogonally to the first one. The area of
a minimum metal AP mine of diameter 8 Fig 2: The 3 objects used for initial
cm is therefore covered by about 40 comparative tests
A-scans.
5. Hand-held device
6-29
Mine
Metallic
scrap
6-30
In a first step, it is not required to engagement in this project and the very
identify precisely the mine. All mines must good results already obtained.
be signalled, with a safety better than
99.6%. They may then be prodded, or
destroyed immediately. False alarms must References
be minimized, but a factor of 2, against
[Borgwart95] C.Borgwardt, "GDIS - Ordnance
the current 100 to 1000, of false alarms
Detection and Identification System",
is probably acceptable. WAPM'95 Workshop on Antipersonnel Mine
Detection and Removal, Lausanne, June
30-July 1st, 1995, pp 37-43
6-31
6-32
MINE DETECTION WITH MODERN DAY METAL DETECTORS
INTRODUCTION APPLICATION
With the end of the cold war, the threat of nuclear To remediate and render safe an area which is
conflict has been substantially reduced. Countries contaminated with mines, ammunition, and OEW (or the
around the world view this situation with both combination of these hazards), the surface must first be
celebration and opportunity. One negative result has cleared from these explosives. This can be accomplished
been an increase in regional conflicts over national and by two basic methods:
political sovereignty.
1. MINES: First, mines must be detected with a hand¬
The world has seen a dramatic increase in the use of held mine detector and immediately removed /
landmine warfere in many regions. Today, there is an deactivated. This work can be very fetiguing for the
estimated 100+ million mines which will require EOD operator and can only be done manually.
detection and disposal work. One reason for this increase
is due to the relatively inexpensive cost of mine
deployment with relationship to a highly effective
strategic effect.
6-33
So-called efficiency methods wliich use heavy machinery world" conditions can be overlooked by novice customers
or explosives are not recommended because it carmot be during the detector selection process,
ensured 100% that the mines wiU be destroyed. Further,
a lot of metal fragments will be scattered over the area GROUND / SHALLOW WATER CONDITIONS
rendering it impossible to perform a repeat survey scan.
The following field conditions are typical considerations
which are commonly encountered in mined areas;
2. OEW: For the detection of UXO and Explosive This means a metal deteaor (mine detector), should work
Waste (mines excluded), both metal detectors and to it's optimum level in all conditions to assure reliable
magnetometers may be used together to clear surfece and and safe operation.
sub-surface targets. Advanced detection systems are
available which will produce target lists and maps to Moreover, the detection sensitivity must be very high
assist with the removal process of these items. level to detect both small metal items such as firing pins
in plastic mines and larger metal targets at a greater
DETECTOR SELECTION distance below the surface. In wartime scenarios small
AP mines may be placed in close proximity to larger AT
For both of the above noted methods a variety of mines. The detector should be able to discriminate these
detectors are available on the market. However, only a different targets to avoid detonation.
selea few^ models will meet the safety and detection
needs for mine detection requirements.
o
o o
6-34
These requirements can only be fulfilled by a “modem metal debris or other targets outside the detected target
day” mine detector uith highly sophisticated electronics must be discriminated out or the operator will fatigue
combined with an optimum physical working design. quickly and reduce the safety level of the operation.
For this purpose, Vallon GmbH produces their model
ML1620B along with several variations for special user
requirements.
TARGET RESPONSE
6-35
Therefore, it is absolutely necessary that the metal
detector uses a measuring principle which does not
produce false signal indications under the full variety of
ambient conditions (the detector must also adapt
instantaneously to changing ground conditions without
the need for operator adjustments). For this purpose,
either a single coil design (which serves as both
transmitter and receiver), or a multi coil design (one
transmitter coil and one or more receiver coils), may be
selected.
_r _r _r _TL
directly or fiilly transferred, to real field applications.
During laboratory measurements the preparation consists
of an ideal relationship between metal test samples and
the detector's search coils.
6-36
The electronically induced current impulsed through the functions are continuously monitored and checked for
detector coil produces an electromagnetic field which 100 % reliability during use. The detectors can operate
contains a high quantity of frequency points. in a synchronous fashion allowing for side-by-side
sweeping operations.
UNDERWATER CONDITIONS
6-37
TECHNOLOGY ASSESSMENT OF PASSIVE MILLIMETER WAVE IMAGING
SENSOR FOR STANDOFF AIRBORNE MINE DETECTION
Brad Blume, email: Brad.Blume@sfo.nichols.com
Sam Taylor, email: Sam.Taylor@sfo.nichols.com
Jack Albers, email: Jack_Albers@netqm.nichols.com
Nichols Research Corporation
P.O. Box 9474
Panama City, FL
Telephone: (904) 233-0667
Fax: (904)233-1262
ABSTRACT - Several mine detection systems are currently background clutter in the MMW regime. A signature code
under development which will provide airborne mine detection provided insight into the effects of various parameters on the
capabilities. For example, the COBRA program utilizes a overall system’s performance. Finally, a physics based
multispectral video camera which will provide an interim clear image generation software package simulated the
weather daylight capability when deployed in a Pioneer performance of an airborne PMMW mine detection system
unmanned aerial vehicle (UAV). The ASTAMmS program has versus both visible and infirared systems.
a dual approach, one contains an active polarized source as well
as a passive IR camera and the other only a passive IR imager. II. PMMW Technology Overview
Either ASTAMIDS system will provide day/night and limited
visibility operation. The use of a PMMW Imaging sensor
The millimeter wave regime has been exploited for both
promises to provide day/night and all weather mine detection
performance. Attenuation in the MMW regime is not
active and passive sensor systems since the 1930’s. Due to
dramatically effected by adverse weather. In addition there is a the long wavelengths, MMWs are capable of penetrating
large contrast between metal targets and the background for air clouds and to some extent rain, and are not dependent on the
to ground scenarios. Furthermore due to the long wavelengths sun as a source of illumination. The initial use of MMW
vegetation and soils are not completely opaque in the MMW radiometers was for extraterrestrial observations. It wasn’t
regime, offering the possibility to detect buried targets under until the 1960’s, that MMW radiometers began being used
specific conditions. This paper will describe the assessment of for terrestrial applications. Currently, passive microwave
an imaging PMMW sensor for mine detection. The results of sensors are used for meteorological, hydrological,
data collection and modeling analysis will be presented as oceanographic, and military applications. This report
evidence to the utility and capabilities of the technology to
addresses the application of a passive millimeter wave
perform under adverse weather conditions.
imaging sensor to mine detection from an unmanned aerial
vehicle.
I. Introduction
6-39
temperature, according to Rayleigh-Jeans law for blackbody radiance quantities. The aperture received radiances will also
radiation. Metal objects are highly reflective and highly include atmospheric path radiance and atmospheric
specular, like a mirror. The radiance received from metal attenuation effects. Imaging in a millimeter wave system is
objects will not be indicative of their temperature, while that performed either by sampling the image in the aperture plane
received from plastics and composites are a combination of (aperture plane imaging) or in the focal plane (focal plane
transmissive, emissive and reflective. For example, mylar imaging). A scanning lens or antenna is an example of an
plastic covers are transmissive and imaging through a 1.4” aperture plane imager, while a focal plane array placed
gypsum garage door has been demonstrated.* Water and wet behind a refractive lens is an example of a focal plane
snow are highly absorptive, while ice and dry snow are imager. A typical millimeter wave radiometer consists of a
highly transmissive. Imaging through 4.0 inches of snow has calibration source, a low noise amplifier or an intermediate
also been demonstrated in this effort and is presented in frequency (IF) mixer, and IF amplifier, a detection stage and
Section IV. an integrator. The calibration source can also provide a gain
drift compensation, will allow the calculation of radiometric
A. Radiometric, Apparent, or Brightness Temperature temperatures from output receiver voltage levels and ensure
high quality data. The sensitivity of a well designed
The term used to describe the radiometric levels of an radiometer is set by the gain and noise characteristics of the
object in a scene in the millimeter wave regime is radiometric first receiver element. This makes the design of low noise
temperature (or apparent temperature or brightness amplifiers (LNAs) an important development area for the
temperature). This term stems from the fact that at millimeter advancement of millimeter wave receivers. The use of IF
wavelengths, blackbody radiation can be described using mixers (super-heterodyne receivers) has historically been
Rayleigh-Jeans law, which is linearly proportional to required due to low performance high frequency amplifiers.
temperature. Therefore, at a given operating frequency, a But, now current state-of-the-art LNAs have noise
temperature value in Kelvin can be used to completely characteristics competitive with super-heterodyne receivers.
describe the received radiation levels. This temperature is
then termed the radiometric temperature. Because the
downwelling sky radiance is at such a low level of radiance,
it is termed “cold.” The sky radiometric temperature is
around 60 K at 94 GHz on a clear day. As the operating
frequency increases and as the weather degrades the sky
temperature increases. Table 1 gives typical values of
radiometric sky temperature and atmospheric attenuation for
various weather conditions and operating frequencies.
Figure 1. PMMW Radiometer System Components
Table 1
Typical MMW Atmospheric Quantities. C The Millimeter Wave Advantage
6-40
required to get improved resolution. In addition the Currently, the state-of-the-art solid state RF devices are based
technology of MMW arrays is immature. Single element on the pseudomorphic HEMT (PHEMT) technology, which
radiometers have been around since the 1930’s, but it is only uses a combination of InGaAs, AlGaAs and GaAs to provide
recently that many receivers have begun to be integrated in enhanced electron mobility over GaAs based HEMTs.^
arrays. It is the development of monolithic millimeter-wave LowNoLso
Detector
Ajnplirier
integrated circuits (MMIC) technology and other
microfabrication techniques which has made array —H—
technologies feasible. Local Oscillator
TRF
6-41
approach to reducing the thermal mass is to separate the Microbolometers have been successfully applied to
radiation absorber and the bolometer element, which results infrared detection, and to terahertz detectors. Several sources
in the antenna coupled microbolometer. Antenna coupled have also demonstrated fabrication of arrays of antenna
microbolometers have been shown to have higher coupled microbolometers.’’* Millimeter wavelength
responsivity, better sensitivity, and much faster response than microbolometers have been used for astrological
conventional bolometers.^ Figure 3 is an illustration of a observations, and plasma fusion reaction monitoring.’
antenna coupled bismuth microbolometer and typical readout
circuitry. The key to the development of sensitive millimeter B, Imaging Technology
wave microbolometers is the design of small responsive load
elements. The sensitivity of microbolometers is determined There are several approaches to doing passive imaging in
by three main noise sources, the shot noise, the Johnson the MMW regime. These basic techniques are outlined in
noise, and the phonon noise. Through the choice of Table 3 along with advantages and disadvantages of each.
appropriate load element materials and dimensions, the Single element scanning radiometers have been in existence
thermal responsivity can be adjusted such that only the since the 1950’s and are the most basic form of imaging
phonon noise is the limiting noise mechanism. Figure 4 system. This type of device can use conventional MMW
below illustrates the noise equivalent temperature difference components or the state-of-the-art MMIC receivers or
(NETD) of a well designed antenna coupled microbolometer microbolometers. It has been the advent of MMIC receivers,
as a function of operating temperature. Notice that the which has led to the development of linear and two
NETD improves as the operating temperature in decreased dimensional arrays of MMW receivers. In general, the
and as the operating frequency is increased. Through cooling MMIC receivers can be one of two basic configurations:
of the microbolometer elements or arrays, sensitivities super-heterodyne or direct detection.
approaching that of MMICs can be achieved. For example a
220 GHz antenna coupled microbolometer operated at 77 K, The interferometer type antenna can achieve the same
and a 14 msec integration time has an NETD of 1.2 K. resolution as real aperture systems with smaller elemental
antennas, but the widest separation of the elemental antennas
must equal the diameter of the real aperture system. The
decrease in weight and convenience in placing the elemental
antennas, costs signal to noise ratio and additional
processing. There are two basic types of Interferometric
imaging. One is the “Michelson” interferometer, in which
the signals received by different antennas are always added
or subtracted after being transmitted to the amplifiers and
processors through cables. Larger antenna separations or
baselines pushed the Michelson type interferometers to limits
Figure 3. Antenna Coupled Microbolometer and Readout Circuitry. imposed by phase distortions due to transmitting the signals
from the individual antennas to the signal processors. To
overcome this limitation, Brown and Twiss^° proposed a new
type of interferometer based on correlation processing, that
is, time-averaging the multiplied signals. Figure 5
graphically depicts the interferometer receiver.
-^95 GHz
-»■ 140 GHz
• 220 GHz
6-42
Table 3
PMMW Imaging Techniques.
^^1
Scanning
focal plane to generate image. . Well calibrated and stable • Mechanical design complex
• Classical, mature imaging system . Overall sensitivity reduced by
design number of pixel samples
Arrays Multiple elements configured in a focal plane • Starring image acquisition • Difficult to calibrate
I
imaging system, where each element represents a . Best possible integration times • High power consumption
single MMIC receiver. • High cost
Cl • Above two linnits number of
possible pixel elements
Use of sparsely placed multiple receivers or slot • Improved spatial resolution • Significantly reduced
antennas and coherent processing to synthesize an • No moving parts sensitivity
aperture larger than any single receiver element. • Tight tolerances on antenna
spacing and design
MMW antenna coupled to a thermo resistive load • Starring image acquisition • Reduced sensitivity
fabricated on a substrate with many antenna • Dense array of pixel elements
coupled elements. • Cheap batch processing
6-43
the Smart Tactical Autonomous Guidance (STAG) program,
is developing several radiometers and imaging concepts.
One of these devices is the Millimeter Wave Analysis of
Passive Signatures (MAPS) trailer. The MAPS is under
development by Millitech Corporation in South Deerfield,
MA. The MAPS consist of four radiometers and a video
camera mounted on a trailer. The data collection and
radiometers are controlled from inside the trailer. The
National Institute of Justice is sponsoring the development of
an 8x8 MMIC based focal plane array, which is being
developed by Millitech Corporation.*® This device will be
capable of imaging 30 frames per second with ~2 K
temperature sensitivities. The 8x8 array is dithered to get a
32x32 pixel image. A super resolution algorithm is applied
to the images that provides a 4 times improvement in the
resolution. TRW, under the DARPA Technology
Reinvestment Program (TRP), is developing an 40x26 MMIC
based focal plane array.This device will provide imagery at
a 17 Hz frame rate with ~2 K temperature sensitivity. A Figure 7. ThermoTrex Real-Time PMMW Imager.'"
conceptual drawing is given in Figure 6.
nATKMM MOTION
6-44
The metal plates are oriented such that the coldest part of the
sky (i.e. zero degrees zenith) is reflected off the plates into The data collection demonstrated the capabilities to image
the radiometer. The inert mines are laid flat on the ice and mines at 94 GHz under clear conditions. But would a
are reflecting a warmer part of the sky than the metal plates. PMMW radiometer have similar performance under adverse
Therefore, the contrast between the mines and background is weather conditions? A system performance model was
reduced. In an operational scenario, the mines would have developed by NRC to address this question and execute
contrast similar to the metal plates. In Figure 11, the entire performance trade off analyses. Figure 12 below depicts the
test set up is under 4.0” of new snow. There is no difference performance of several PMMW system configurations. The
between the data with snow cover than without snow cover. system configuration assumed capabilities compatible with
This demonstrates the penetration capabilities of PMMW current detectors used in a scanning mode and operating at an
imaging sensors. Notice that the plastic and the fiberglass altitude of 300 feet. As the operating frequency of a PMMW
mines are both observed as warmer than the background. radiometer is increased, the atmospheric attenuation is
increased as well as the sky temperature. This decreases the
This is as the theory predicts.
contrast between a metal mine and the surrounding
background, thus decreasing the signal to clutter ratio. As the
operating frequency increases, the resolution of the system
improves, resulting in more of the target filling an image
pixel. This will provide improved signal to clutter ratios
under conditions where the target is smaller than the pixel’s
LEGEND:
0 20” Mumlniim OI»e ground projection. From analysis of Figure 12, it can be seen
• lO^MurntnimOtoe
• 3” AhMnlnMii Otoe
that the effect of improved resolution can provide enough
”1” Metal Mint
"2” PtotticMIrM
signal to overcome the reduction in contrast. The size of
"T MttalMkit
n” MttalMbit possible apertures for a UAV are between 0.3 and 0.5 meters.
”5” FU»trgtoM Mint
S’ltS* AL Shtta For an aperture of 0.4 meters, the best overall performance is
S'JtS* Roly Ovtf RAM
provided by a 94 GHz. It is also important to notice the
signal to clutter ratios can be directly related to the
Figure 9. Data Collection Test Set Up. performance of a mine detection algorithm. These results
indicate improved performance through the use of a PMMW
mine detection system.
6-45
assumes a starring focal plane array radiometer under fog
conditions. These two cases illustrate some interesting
points. One is that the 220 GHz system suffers from poor
performance under short integration times, which is due to
the larger noise figure. If the integration time is increased to
drive the sensitivity well below the clutter, the fill factor
dominates the system performance as evidenced by the Figure 15. PMMW, IR, Visible Imagery Of Minefield From UAV Under
improved performance of the 220 GHz system in Figure 13b. Fog Weather Conditions.
As the altitude continues to increase, the atmospheric effects
versus frequency begin to dominate the system performance.
It can be seen in Figure 13b that the 140 and 94 GHz system
performance surpasses the 220 GHz at 900 and 1200 meters
altitude respectively.
Figure 16. PMMW, IR, Visible Imagery Of Minefield From UAV Under
Clear Night Weather Conditions.
6-46
10000
-.-Dry 35GHz
-.1-Moist 35GHz
-»-Wet 3SGHZ
-.-Dry 94GHz
Moist 94QKZ
-*-Wat 94GHz
6-47
required operating frequency between 94 and 220 GHz. This TABLE 4
operating frequency range is required to achieve sufficient PLASTIC Mine Contrast Versus Soil Background.
Dry 35GHz
35GHz
-*-Wel 35GHz
(b) 35 GHz
-•-Diy94GH2
-*-Mdst MGHz
-^Wat 94GHz
(c) 94 GHz
E. Effects of Water
6-48
A PMMW system will lend its self well to the detection of
buried objects, but not from an airborne platform. The
apertures required to get only target pixel information at the
soil penetrating frequencies are prohibitive. The detection of
plastic, wood and metal mines is feasible. The plastic and
wood mines will be warmer than the background, and the
metal mines will be much cooler than the background. The
design of a buried mine detection system will require a
systems engineering approach, as there are many variables
which affect the performance of a PMMW buried mine
detection system.
6-49
Geoscience and Remote Sensing, VoL 26, No. 5, pp.
597-611, Sept. 1988.
David M. Le Vine: ‘The sensitivity of synthetic aperture
radiometers for remote sensing applications from
space,” Radio Science, Vol. 25, No. 4, pp. 441-453,
July 1990.
Peter J. Napier, et al: “Very large array: Design and
performance of a modern synthesis radio telescope,”
Proceedings of the IEEE, Vol. 71, No. 11, pp. 1295-
1320, Nov. 1983.
J. G. Good: An Overview of Microwave Interferometry and
Aperture Synthesis for Observing the Earth from
Space, CSC/TM-83/6080, Computer Sciences
Corporation, prepared for NASA under contract
NAS 5-24350, Aug. 1983.
G. R. Huguenin, S. Bandla, and E. L. Moore, “Millimeter
wave focal plane array imaging sensors for
concealed weapon detection,” Proceeding of Signal
Processing, Sensor Fusion, and Target Recognition
V, SPIE AeroSense ‘96, April 1996.
G. S. Dow, D. C. W. Lo, et al, “Large scale W-band focal
plane array for passive radiometric imaging,” IEEE
MTT-S International Symposium Digest, 1996.
J. Lovberg, R. Chou, C. Martin, and J. Galliano, “Advances
in real-time millimeter-wave imaging radiometers
for avionic synthetic vision,” SPIE Proceedings of
Synthetic Vision for Vehicle Guidance and Control,
AeroSense ‘95, April 1995.
Christopher S. Ruf, et al: “Interferometric synthetic
aperture microwave radiometry for remote sensing
of the Earth,” IEEE Transactions on Geoscience and
Remote Sensing, Vol. 26, No. 5, pp. 597-611, Sept.
1988.
J. Watson, and D. S. Flynn, .’The Irma Multisensor
Predictive Signature Model,” Proceedings of the
Sixth Annual Ground Target Modeling and
Validation Conference, August 1995.
L. Yuijiri, S. Fornaca, B. Hauss, M. Shoucri, S. Talmadge,
“Detection of metal and plastic mines using passive
millimeter waves,” Proceedings of Detection and
Remediation Technologies for Mines and Minelike
Targets, SPIE Vol. 2765, AeroSense ‘96, April
1996.
J. Groot, R. Dekker, L. van Ewijk, “Landmine detection
with an imaging 94 GHz radiometer,” Proceedings
of Detection and Remediation Technologies for
Mines and Minelike Targets, SPIE Vol. 2765,
AeroSense ‘96, April 1996.
DARPA’S HYPERSPECTRAL MINE DETECTION PROGRAM
C. J. Sayre
NCCOSC R T D & E Division, San Diego, CA
D. J. Fields
Defense Advanced Research Projects Agency, Arlington VA
A. P. Bowman, A. L. Giles
Space Applications Corporation, Arlington, VA
6-51
standoff neutralization of all the mines in a mine field. measurements in the long wave infrared have shown that
The DARPA Hyperspectral Mine Detection program is the particle size is an important determinant of the
identifying and developing technology to find mines magnitude or strength of the spectral signature[l]. Very
quickly and affordably. Hyperspectral Mine Detection small particles (comparable in size to the wavelength of
sensors can be employed from a helicopter or a low flying the infrared radiation) generally exhibit less spectral
aircraft to detect mines on roads and in offroad areas. variation (the spectral signature or color) than larger
The output of the HMD sensors will be a map of the particles of the same mineral. Thus, even for soil where
positions of individual mines along the road or in the off¬ there is no compositional difference between the top layer
road area. and the subsoil, the different sorting of sizes will result in a
different spectral signature.
2.0 Phenomena
There are also spectral differences between disturbed and
The DARPA Hyperspectral Mine Detection program has undisturbed soil in the visible and short-wave infrared
made measurements over the full optical spectral region region. In particular, wet soil has an overall reflectance
from visible through long wave infrared. Every effort was difference from dry soil. Thus, a fresh mine can be
made to explore all possible observables and specific immediately detected with a visible light sensor (or human
sensors were deployed to cover each spectral region. The eye) in many cases because the subsoil now over the mine
emphasis of this program has been on detection concepts is wetter than the surrounding areas. The disturbed soil
using midwave infrared (3-5 urn) and longwave infrared will appear darker than the surrounding undisturbed soil.
(8-12 pm) because of observed long persisting phenomena This difference will only last until the water evaporates or
in these spectral regions. a rain shower wets everything. Other potential spectral
differences in the short-wave infrared can arise from soil
The principal phenomena identified for hyperspectral mine clay content but have been determined to be unreliable.
detection is based upon detecting localized differences in Also, the spectral region around 2.2 pm is a region where
the scene created by the mines. The placement or presence the signature of the moisture content of soil is particularly
of the buried mine will change the observables of a small strong. This particular signature may last longer than the
area above the mine. Initially, this observable may be broad reflection difference due to wet soil. It will not last
simple to detect by other techniques. For example, as long as the signature due to a true compositional or
immediately after placement there will likely be a texture particle size difference.
or moisture difference that can be detected with a broad
band infrared instrument or even visible light sensors. 3.0 Sensor Experiments
Shortly, after drying and weathering, the obvious texture
and moisture differences may disappear. What remains as Hyperspectral data were acquired over many regions of the
an observable are long term effects of the soil disturbance. United States using both non-imaging and imaging
spectrometers. Three major imaging sensor deployments
The hyperspectral mine detection concept is based upon were conducted using both a mid-wave infrared imaging
the existence of some compositional or particle size spectrometer and a long wave infrared imaging
difference between the soil above the mine and the spectrometer. During these imaging deployments large
surrounding area. This difference will result in a localized data sets on disturbed and undisturbed soil and on buried
difference in the spectral signature of the surface of the soil mines were acquired. The imaging data sets were used
that can be observed by a hyperspectral or multi-spectral extensively to study the phenomena and develop detection
sensor. It is also possible that the existence of the mine techniques. These large imaging data collections were
itself can be an observable since it can introduce localized supplemented by many geologically and geographically
differences in the temperature of the surface and in the dispersed non-imaging collections. By referring to Figure
behavior of vegetation growth in the near surface area. 1, the broad base of the measurement program can be seen.
Data has been taken in the Desert Southwest, the Eastern
The disturbed surface phenomenology and possible buried Seaboard and in the Hawaiian Islands. Special emphasis
mine observables were investigated for the infrared and was made to acquire a wide variety of geologically
reflection band spectral regions. All soils show some different soil types using field teams with trained
spectral structure in the 3-5 pm and 8-12 pm regions. geologists. For the vast majority of these locations, the
These spectral features are characteristic of the soil mineral same spectral observable was found in the long wave
content. The surface layer can be different from the infrared. In most of the remaining locations, a secondary
subsoil either because of compositional differences or spectral observable in the long wave infrared, useful for
because of particle size differences. Many laboratory detection, was found.
6-52
Figure 1 Map of the United States Showing Locations Where HMD Experiments Have been Performed. Imaging experiments have been
performed in Southwest (Arizona and Nevada), while non-imaging experiments have been performed in Hawaii, California, New Mexico and
along the Eastern Seaboard
mine is buried, soil from below the surface is placed on the WNelen^(iiicronB
surface and this new soil contains a mix of small particles
and large particles. The spectral signature is then reduced
because the spectral signature of small particles is much
Figure 2. Field measured emissivity on the same soil showing a signature
lower than the spectral signature of the larger particles
difference at 9.2 pm between the surface soil (the bottom curve) and the
which normally reside on the surface. subsoil brought to the surface by digging (the upper curve).
6-53
During the course of the HMD Measurements program, infrared imaging array is used as the focal plane and is
other observables for mine detection have been measured. cooled to liquid nitrogen temperatures.
In particular at some of the locations a compositional
difference has been seen between the surface layer and the The AHI sensor will be installed into a helicopter platform
subsurface layer. This compositional difference can be and will acquire hyperspectral data from an altitude of 100
caused by rocks or vegetation differences in the area. meters. From this altitude, a mine will be seen on multiple
Measurements were also made on silicate free soils at sensor pixels. The output of the sensor will be digitized
locations in Hawaii. and calibrated in real time on-board the helicopter. The
calibrated data will be presented to the operator in real
Imaging spectrometer measurements proved to be essential time.
to the understanding of the phenomena behind mine
detection. Not only do imaging measurements give orders In addition, mine detection processing will be performed
of magnitude more data than non-imaging measurements, in real time on-board the helicopter. The mine detections
but they provide data to support the understanding of the will be geo-referenced using a Global Positioning System
spatial aspects of the mine signature. receiver. All calibrated data will be recorded and will be
available for further processing at the ground station.
An example from the imaging data collections is included
here. This data was acquired using the Lawrence 5.0 Summary
Livermore National Laboratory LIFTIRS hyperspectral
LWIR Sensor [2]. A further discussion of the sensors and The disturbed soil signature due to the placement of a
experiment may be found in [3}. buried mine can be seen as a spectral difference from the
neighboring undisturbed area. While the strength of this
The example in Figure 3 shows the application of a signature is strongest immediately after mine
detection algorithm to the problem of detecting mines emplacement, it will remain strong for a period of days to
buried in a road. This data was taken from a range of weeks and is difficult to suppress. Experiments have
approximately 1100 feet with the sensor on a cliff side shown the mine detection observable to be detectable after
location. Broad band infrared data is shown first, with the a period of weeks to months depending on the degree of
road running through the center. No mines are apparent in weathering. Since the weathering process is what sorts the
this broad band image. The second image shows a small particles from the large particles and creates the
composite color image made from the first three Principal observable, large amounts of rainfall will tend to wash the
Components. The mines can be seen as patches along the small particles from the surface layer. Nevertheless,
side of the road. The mines could be detected with a disturbed soil signatures have been measured after months
spectral matched filter if the off-road area were excluded of weathering, including rainstorms. A residual disturbed
from consideration. Since it is the aim of the HMD soil signature has also been seen in mine fields that had
program to detect targets both on the road and in the off¬ been flooded.
road area, a quadratic detector was then applied to the data.
This detector, using training information from one mine, The detection of buried land mines by the signature of the
was able to detect all the mines and reject false alarms not disturbed soil has limitations. It cannot be used, in
only on the road but in the off-road area. general, for the detection of long buried mines. There are
other observables that may be applicable to the detection
4.0 Sensor Development of mines buried for long periods of time (from several
months to years). Preliminary results show that the long
As part of the DARPA HMD program, a new buried mines can be seen at certain times of the day as a
hyperspectral long wave infrared imaging sensor is being thermal anomaly. Immediately after sunrise, the thermal
developed by the University of Hawaii. This sensor, the mass of the mine retards the solar heating of the soil above
Airborne Hyperspectral Imager (AHI) will fly on a the mine. At least in the sparsely vegetated desert areas
helicopter platform and be used for mine detection where these experiments were conducted, there was a
experiments in 1997. distinct lack of vegetation over the mine due to the
restrictions in root growth. Mines buried over two years
The AHI sensor is a grating spectrometer and is designed were used for these experiments. The coupling of a
to operate from an airborne platform. AHI will acquire vegetative anomaly (lack of vegetation) and a temperature
spectra for a row of 256 pixels on the ground anomaly could be a good indicator of a long buried mine.
simultaneously and build up the second spatial dimension
by pushbroom scanning . A HgCdTe 256 x 256 longwave
6=54
6.0 Acknowledgments
7.0 References
Figure 3. Detection of Road Mines in Hyperspectral image Data. Broad band LWIR image of road (top) does not show the mines. The spectral false color
image (center) shows the mines but with off-road false alarms. The thresholded output of a quadratic detector algorithm (on bottom) shows only the
detected mines, with no false alarms.
6-55
6-56
On the feasibility of microwave imaging of buried land mines
at modest stand-off distance
Abstract - This paper describes feasibility tests of a has been utilized for short range and vertical incidence, at
concept for a vehicle carried, microwave system for oblique incidence [1]. The motivation for imaging is
imaging land mines at distances 5 to 10 meters ahead of a identification, which can accelerate searches.
vehicle. The system would have antenna arrays and Section II describes the system concept, and Section III
multiple discrete frequencies to acquire data. The paper summarizes the image formation theory. Sections IV and V
presents images from data measured with a transmitting describe initial measurements and image calculations that
antenna and a translated receiving antenna. were done to test feasibility. The measurements did not use
an antenna array; instead, an antenna scanned a linear path,
I. Introduction and another, fixed antenna transmitted. Mine targets were at
distances approximately 3-1/2 meters from the antennas
Section IV describes results for a 12-inch diameter, non-
Land mine detection is a difficult technical problem
metallic mine on a paved surface. It presents an image which
because it involves many variables including mine
is a range profile generated from reflections for 26
configuration, size, composition, and depth as well as diverse
frequencies in the band 5 to 6 GHz. This image suggests
soil properties. Many sensors have been developed for
wave mechanisms, reflections from the mine’s front and back
detection. These include electromagnetic induction detectors,
surfaces.
ground penetrating radars, and infrared cameras. These
Section V describes results for the 12-inch mine buried
sensors are useful but have deficiencies. For example,
in damp soil. It shows a plan view image computed from
induction methods detect metallic but not dielectric mines.
data for frequency 2 GHz. Section V also presents a range
Radars can detect both metallic and dielectric mines but to
profile generated from 101 frequencies in the band 2 to
our knowledge do not detect small, non-metallic mines. For
6 GHz. These images also suggest reflection mechanisms.
infrared systems, mine depth and sunlight variation can limit
For comparison, an image was formed of a metal plate, from
detection. These problems have stimulated the development
specular returns.
of multi-sensor systems and data fusion, but work continues
on improving sensors.
In addition to the influence of sensors, the effectiveness II. System CONFIGURATION
6-57
where Uj, is reflectance at frequency 4, r is slant range, and c
is the speed of light in air. Note that propagation in soil is
omitted so that Equation 1 applies to mines on the surface or
at depths small relative to wavelength.
The second processing method utilizes data over antenna
positions for each frequency. The first step is to form the
angular spectrum.
(2)
*=0
III. Theory
IV. Results for a surface mine
6-58
network analyzer was the transmitter and receiver, and a soil surface. Reflectance was measured for distinct
personal computer controlled the analyzer and digitized data. frequencies between 2 and 6 GHz, at intervals of 0.04 GHz
during antenna motion.
o
X
60
position (inch)
Antenna Position (inch)
Fig. 3. Bistatic arrangement for measurements with a mine stimulant on a Fig. 5. Arrangement for measurements on buried mine simulant. The
paved surface. The receiving antenna is labeled R; the transmitting, T. transmitting antenna T was fixed in position. The receiving antenna R was
Vertical polarization. Antenna heights were 52 inches. (BPL2; translated. Antenna heights were 52 inches above the ground surface.
Polarization was vertical. (229GG)
Fig. 4 shows a range profile calculated by evaluating Figure 6 shows a plan view generated by evaluating
Equation 1 for frequencies from 5 GHz to 6 GHz in Equation 4 with data from the interval 24 to 48 inches in
0.04 GHz steps. The image in Fig. 4 suggests reflections Fig. 5. Frequency was 2 GHz. Image values were calculated
from the mine’s front and back surfaces. for three values of slant range, 132, 136, and 140 inches. In
this image, the shaded regions show where amplitude
exceeded 0.7 times the peak amplitude, which occurred for
ranges 132 and 140 inches.
6-59
Although the calculations are sparse, they do suggest the VI. Summary
object’s shape. The image also suggests that reflections
occur at the object’s boundaries as well as at the center.
The paper presented an approach to imaging land mines
Range profiles were calculated using frequencies from 2
at distances of 5 to 10 meters ahead of a vehicle. The
to 6 GHz in .04 GHz steps. As a preliminary, to test
approach uses a band of discrete microwave frequencies, a
accuracy, an image was formed of a 6-inch square metal plate
transmitting antenna, and an array of receiving antennas. The
arranged for specular return to the 24-inch position of Fig. 5;
system synthesizes a line array of reflectance data, which are
the mine was absent. The image, shown in Fig. 7, was
digitally processed to form images which are range profiles
computed for data from the 24-inch position. A range profile
or plan views.
of the buried (1-1/2 inch deep) mine was computed from data
The paper described measurements for a 12-inch
for frequencies 2 to 6 GHz in .04 GHz steps. Again, the data
diameter, non-metallic mine simulant. The measurements
were for the 24-inch position in Fig, 5. The profile is in
were made with a fixed transmitting antenna and with a
Fig. 8.
receiving antenna scanned on a linear path. Images were
The range profile for the plate shows a sharp peak at the
formed for the mine on a paved surface and buried 1-1/2
plate’s position. The range profile for the mine suggests
inches in damp soil, at distances approximately 3.4 meters.
multiple reflections.
The images suggest multiple wave reflection mechanisms.
REFERENCES
6-60
CHEMICAL SYSTEMS FOR IN-SITU NEUTRALIZATION OF LANDMINES IN
PEACETIME
Divyakant L. Patel and Beverly D. Briggs
U.S. Army CECOM R&D Center
Night Vision & Electronic Sensors Directorate
Countermine Division, Environmental Systems Branch
Fort Belvoir, Virginia 22060-5806
and
Allen J. Tulis, James L. Austing, Remon Dihu and Alan Snelson
IIT Research Institute
Chicago, Illinois 60616-3799
6-61
NVESD, to improve the world’s capability in humanitarian landmines, leading to the non-detonative autocatalytic decomp¬
demining operations. osition/self-consumption of the entire explosive charge [5,6].
Ignition is achieved at even sub-zero temperatures with some
Background hypergols; e.g., diethylzinc (DEZ) reacts hypergolically with
TNT powder even at -30^C within a matter of seconds.
The main explosive charge in almost all foreign and domestic
landmines is typically TNT and TNT-based explosives such as Current investigation
Composition B (Comp. B), amatol, picratol, etc, [2]. Explo-
sives contain considerable oxygen within their metastable Two prototype remotely-operated chemical delivery systems
molecules; hence, they do not need air in order to detonate, have been developed for the in-situ chemical neutralization of
deflagrate, or dissociate by autocatalytic decomposition. Most the main explosive charge in landmines [7,8]. The first, System
explosives can dissociate by alternative mechanisms, and No 1, referred to as “bullet with chemical capsule” (BCC), uses
dissociation by detonation generally involves an entirely a small quantity of an amine or metal alkyl in a plastic capsule
different mechanism than by autocatalytic that is placed just above the landmine using a simple tripod. A
dissociation/decomposition. TNT will generally burn fiercely bullet, shot through the capsule and into the mine, ruptures the
but without transition to detonation if simply ignited; i.e., capsule, penetrates the overburden and the mine casing, and
without use of a detonator and explosive booster charge to enters into the explosive charge, carrying the dispersed chemical
shock-initiate the TNT. Hence, if a stimulus means such as a hypergolic reagent into the explosive charge inside the mine.
chemical hypergol or radiant-energy laser is capable of “direct¬ Within seconds a highly exothermic, hypergolic autocatalytic
ing” the dissociation mechanism into autocatalytic decomposi¬ self-destruction of the explosive charge takes place. Within
tion in lieu of detonation, it is likely that detonation will be minutes, depending on the size of the explosive charge, the mine
precluded. The chemical transformation of TNT, as well as is chemically neutralized. The second. System No. 2, referred
most other organic secondary explosives, can proceed by four to as “chemical-filled projectile” (CFP), shoots a cartridge-case
general mechanisms as follows: projectile into the mine in such manner that the projectile
penetrates the mine casing and enters the explosive, rupturing
1. Burning; simple combustion in air (oxygen), the cartridge case to release a metal alkyl into the penetrated
explosive. This similarly causes a hypergolic, highly exother¬
2. Heterogeneous (stoichiometric) chemical reaction, mic autocatalytic complete destruction of the explosive. Both
systems are effective against TNT and Comp. B, the major
3. Detonation, and explosives in landmines, as well as other explosives such as
tetryl. The major advantage of this chemical neutralization
4. Autocatalytic decomposition. methodology is that complete, non-detonative neutralization of
the explosive component in the mines is achieved, without
Open-pit burning had been a common practice for the disposal detonation damage to the area or contamination by mine-casing
of propellents and explosives; the materials were placed in open debris, especially in the case of metal-encased landmines.
trenches, covered with straw and drenched in kerosene or fuel
oil, then ignited. Exposed to adequate air, combustion pro¬ Major effort was placed into design and development of these
ceeded to completion. Because of its dependence upon oxygen, prototype remote-operated delivery systems, which are suffi¬
burning of confined, often buried ordnance is not feasible. ciently robust to be reusable. For expediency, an over-kill
Heterogeneous chemical reaction of explosives with suitable scenario was adopted for delivery System No. 1 in that hypergol
chemical reagents is effective [3,4] but requires excessive (stoi¬ reagent loss due to dispersal by the bullet and absorption into
chiometric quantities of such reagents, and there is no practical, the overburden would limit the amount actually entering the
effective delivery system for in-situ neutralization, especially in mine and reacting with the explosive. The amine diethyl-
the case of buried mines. Detonation of explosive ordnance is enetriamine (DETA) was selected for this system. It was
a viable option that is in practice, but as discussed earlier has demonstrated that 60 mL of DETA should suffice for mines
considerable drawbacks. Autocatalytic decomposition is the buried under 305 mm of overburden. For the second delivery
simplest, cheapest, and most effective option for chemical system, hypergol reagent loss was not of concern since the
neutralization of landmine explosives. This type of chemical cartridge delivered the reagent directly into the explosive within
neutralization is most readily achieved by using suitable the mine casing. Hence, in the case of this system nominally 5
chemicals that are hypergolic with the explosives; e.g., metal mL DEZ was effective.
alkyls and aliphatic amines. Very small amounts, even several
drops in laboratory tests, cause nearly instantaneous hypergolic
ignition of TNT, Comp. B, and most other explosives used in
6-62
Delivery systems cartridge cases to the extent required to achieve penetration of
the overburden (if any) and the mine casing.
Both delivery systems are operated remotely using an electric
squib, and a tripod for positioning the delivery devices above Fort a. p. hill demonstration tests
6-63
As indicated in Table 1, all tests were conducted with 60 mL of As indicated in Table 2, in some tests the DEZ hypergol was
DETA regardless of the type of mine or mine casing. Since in diluted with 20 percent toluene to mitigate its pyrophoricity.
well-controlled laboratory experiments only a few drops of Because of this pyrophoricity (spontaneous ignition in air),
DETA are adequate to hypergolically initiate the autocatalytic unless the projectile penetrated the mine casing and discharged
decomposition of TNT and Comp. B, the 60 mL was believed the DEZ directly into the explosive, it was consumed by nearly
to be a considerable excess. However, for the purposes of these instantaneous combustion in air. In the case of Test Nos. 12 and
qualitative tests, this was a calculated condition to overcome 13, inadequate propellant charge prevented the projectile from
unknown influences of parameters such as temperature, type and penetrating the heavy steel casing of these AT mines; hence, the
form of the explosive, confinement, presence of foreign hypergol did not come in contact with the explosive.
materials, and the actual dynamics of the delivery process. In
tests with Comp. B the explosive charge was completely This delivery system is very design intensive, but is believed
consumed. In tests with TNT the explosive was also completely to have much greater applicability once design aspects are
consumed, except that a small amount of carbonaceous residue resolved. Its potential advantages include the following: (1) It
remained. In all tests the initial hypergolic response occurred requires a much smaller amount of hypergol because it delivers
within a few seconds and intense smoke and flame appeared it directly into the explosive; (2) It is capable of penetrating soil
throughout the neutralization process which persisted for about to neutralize buried mines (the BCC System has already been
5 to 45 minutes depending on the amount of explosive and other demonstrated to neutralize buried mines up to 305 mm); (3) It
factors. In tests with fuzed mines neutralization occurred should be effective against mines under water; and (4) It should
without initiation of detonation. It is interesting that for the AT be capable of neutralizing exposed landmines from a distance.
wooden mine, which had 7.0 kg of TNT wrapped in heavy wax
paper, all the TNT was completely neutralized without the mine Applicability for humanitarian demining
casing being burned; i.e., no flames were observed throughout
the neutralization process. The chemical neutralization systems described here are in the
prototype stage and are under further design and development
to meet the requirements of the U.S. Congress that they be
TABLE 2. FORT A.P. HILL TEST RESULTS OF CHEMICAL NEUTRALIZATION SYSTEM NO. shared in an international environment. These in-situ chemical
2, CHEMICAL FILLED PROJECTILE (CFP), AGAINST VARIOUS AP AND AT LANDMINES:
CHEMICAL HYPERGOL, 5-15 ML. DIETHYLZINC (DEZ). neutralization systems meet the requirements of humanitarian
demining such as low cost, simple ease of operation, safe,
Test Chemical effective, environmentally safe, and capable of being shared in
Nos. Mine tvoe Casing Exolosive Fuzed Neutralization an international environment. These are described separately as
J *3!c:fc AP PMD-6 wood
follows:
TNT no partial
6-64
DETA to be effective against tetryl It is also believed that (1) Conclusions
DISCUSSION: Both of these chemical neutralization systems [3] . Application of the Unreacted Core Model to the Chemical Neutralization of
will neutralize surface buried or buried mines to acceptable TNT in Solvent/Amine System, A.J. Tulis, J.N. Keith, W.K. Sumida, and
D.C. Heberlein. Sixth Int. Congress of Chem. Engineering, Caracas,
depths; however, the chemical-filled projectile system should
Venezuela, July 13-16, 1975.
neutralize mines under water, and its application to neutraliza¬
tion of surface mines remotely from a tank or helicopter is [4] . Chemical Neutralization of Explosives, J.N. Keith, A.J. Tulis, W.K.
feasible. These chemical systems can be applicable for clearing Sumida, and D.C. Heberlein. Proc. Eighth Symp. On Explosives and
Pyrotechnics, pp. 35.1-35.6, Feb. 5-7, 1974, Los Angeles.
unexploded ordnance (UXO), both exposed and buried. The
chemical neutralization of mines without detonations is highly [5] . Non-Explosive Destmction of TNT with Hypergols, A.J. Tulis, J.N. Keith,
desirable, but application to humanitarian demining needs W.K. Sumida, and D.C. Heberlein. Proc. Fourth (Int) Pyrotechnics Sem.,
pp. 17.1-17.32, July 1974, Steamboat Springs, CO. Also available NTIS AD
further investigation, especially to develop an appropriate single
A057599.
chemical reagent composition to be effective against all types of
main charge explosives in landmines.
6-65
[6] . Chemical Neutralization of Landmines, T.C. Beverage, D.C. Heberlein, A.J. D.C. Heberlein. Proc. 27th Int. Annual Conf. ICT, Energetic Materials
Tulis, J.N. Keith, and W.K. Sumida. Proc. Fifth Int. Pyrotechnics Seminar, Tech., Manuf., and Process., pp. 41.1-41.14, 1996, Karlsruhe, Germany.
pp.21-38, July 12-16, 1976, Vail, CO.
[8]. Chemical Neutralization of Explosives as a Viable Option in Humanitarian
[7] . Hypergolic Chemical Initiation of Non-Detonative Autocatalytic Self- Demining Operations, D.C. Heberlein, B.D. Briggs, and D.L. Patel. Proc.
Consumption/Destruction of TNT, RDX, and TNT-Based Explosives, A.J. 22nd Int. Pyrotechnics Sem., pp. 828-840, July 15-19, 1996, Fort Collins,
Tulis, A. Snelson, J.L. Austing, R.J. Dihu, D.L. Patel, B.D. Briggs, and CO.
6-66
Radar Imaging Experiments
for Landmine Detection
Because a full paper was not received by publication date, the above Abstract appears m this
Proceedings. The lead author can be reached at P.O. Box 808, L-437, Livermore, CA 94551,
telephone, 510-422-8538; e-mail, <azevedo3@llnl.gov>.
6-67
Ultra-Wideband, Short Pulse
Ground-Penetrating Radar:
Theory and Measurement
Because a full paper was not received by publication date, the above Abstract appears in this
Proceedings. The lead author can be reached at Duke University, Dept, of Electrical and
Computer Engineering, P.O. Box 90291, Durham, NC 27708-0291.