City Research Online: City, University of London Institutional Repository
City Research Online: City, University of London Institutional Repository
City Research Online: City, University of London Institutional Repository
Citation: Sikora, I., Stanislav, P. and Bazijanac, E. (1999). The Example of Laptop Based
Performance Data Generating and Optimisation in Contemporary Commercial Aircraft
Operations. Paper presented at the ISEP '99, 7 - 8 Oct 1999, Ljubljana, Slovenia.
This version of the publication may differ from the final published
version.
Summary
Airframe and engine combination gives equal potential to every operator with such a hardware
combination. Operator's way of utilization makes its use to the maximum or less. Data related to
aircraft performance is one of the basic elements in daily aircraft operations and optimal utiliza-
tion of airframe and engine combination in real life environment. New technological solutions and
systems affected performance data calculation. Today's laptop computer technology has already
boarded the flight deck together with pilots.
This paper is to present possible structure and proposed application of one of the system together
with envisaged effects of its use in real life commercial aircraft operations. It will present the over-
view of the system with considerations taken into account when designing and developing it; it's
potentials and advantages compared to paper based performance data calculation and optimiza-
tion: and the most important how it is understood as a tool in very demanding, unpredictable air-
line operations of today
2
3. SYSTEM’S STRUCTURE whether it will be assigned to each pilot in com-
mand or just be a part of the documents and equip-
The whole laptop based performance data gen- ment required for each flight.
erating and optimization system consists of two dis- Minimum hardware characteristics for LAPTOP
tinct modules. They are: are Pentium CPU with system clock not less than
1. Raw data creation and system setting mod- 200 MHz and display of minimum 33 cm diagonal.
ule, The system is used for performance calculation
2. LAPTOP based performance data calcula- at the present stage of it’s development. In next
tion module. phases it is projected that other modules will be in-
Each of these modules requires particular hard- corporated as well. That will make system more
ware characteristics (in terms of Central Processor valuable in day-to-day operation. Aircraft weight
Unit (CPU) speed, and display resolution) that al- and balance calculation interacting with take off
low user to make system’s performance exploited at performance data calculation is the first step. Mod-
their best. ules that will adjust aircraft performance according
Today’s version of software is based on Win- to malfunctions present at a particular flight will be
dows 95 operating system. The issue of system sta- incorporated later on.
bility hasn’t been addressed during initial test Raw data created in the office is transferred to
phase. Decision whether it will be Windows 95 or LAPTOPs using some of the existing magnetic or
Windows NT together with the possible use of optic media for data transfer. It is of very high im-
Windows Server operating system, just because of portance to ensure full synchronization of data be-
convenient networking of laptops, are left for each tween originator of data and all LAPTOPs in use.
airline’s discretion. Comparing one ore more floppy disks or even one
3.1 Raw Data Creating and System Setting CD-ROM to more than hundred paper pages (that
are updated regularly for a just slight runway
Module
change) proves system’s advantages over paper
This module is basically office based work- based performance calculation. The gain is in re-
station. It requires at least Pentium CPU with clock duced workload and cost, together with improved
speed of not less than 200 MHz. The main func- efficiency and safety.
tions of this module are:
setting airline’s policy regarding: the use of 4. SYSTEM’S TASKS
units of measurement, regulatory set mini-
mal requirements and conventions, together LAPTOP based performance data calculations
with standard operating procedures adopt- and optimization employs the same algorithms that
ed by airline itself. are nowadays used for creating paper based tabu-
lated performance data. There are two distinct algo-
defining airline’s fleet for which program
rithms that differ in precision and time required for
will be used: setting aircraft registry, and
calculation:
setting aircraft design maximum weights.
1. Polynomial,
setting runway characteristics for all air- 2. 1st Principle,
ports, that airline is flying at, with specific
3. Neural. [5]
fleet. Runway characteristics are defined in
While polynomial advantage is it’s speed of
terms of specific runway lengths, airport el-
calculation, 1st Principle generates more precise re-
evation referred to mean sea level, and ob-
sults. It has always been a trade between those two
stacles in take off direction (their distance
categories.
and height). [2]
Polynomial algorithm is based on previously
All data related to aircraft and engine character-
calculated graphs. Each of them is generated for
istics are supplied by aircraft manufacturer. Pres-
specific values of input variables. No matter wheth-
ently, that is still done by CD-ROM. At the final
er these are weight, wind speed, and temperature,
stage in the future this will be done on line just to
all intermediate values are found using the basic set
avoid obsolete data and make this process less time
of graphs. The use of this procedure leads to the re-
sensitive.[3]
duce calculation time. It is almost always 50% fast-
er than 1st principle mode. Due to this advantage all
3.2 LAPTOP Based Performance Data figures are less precise – conservative [5].
Calculation Module
The main visible element of the whole system to
pilots is LAPTOP computer. It is still not decided
3
1st Principle method on the other hand improves teorological conditions results in distinct value of
precision at the price of time required for calcula- allowed aircraft weight.
tion. Calculation is performed based on basic aero-
dynamic and aircraft engine thermodynamic equa- Although all required pilot calculations are sim-
tions. They are supplied through three distinct da- ple, they give safe, but not always optimal values.
tabases pertinent to each aircraft type and model. In long term there are two potential loss generators:
Each value is generated knowing aircraft condition Unrealized revenue
data, meteorological conditions data, and other sig- unrealized revenue is the result of conservative
nificant influential conditions (aircraft malfunctions polynomial calculation used for producing tab-
and runway condition just to mention some of ulated data regularly. Lower than optimal air-
them). craft weight originates from conservative na-
New generation software for aircraft perfor- ture of polynomial calculation.
mance data calculation and optimization, that au- Increased maintenance costs
thors have evaluated and worked with, employs the increased maintenance costs can not be as-
third calculation mode – neural functions. They sessed directly. Annual balance sheet proves
modify themselves in order to give sought output that higher than required thrust settings result
upon certain input. [5] in aircraft engine deterioration. More precise
Long term experience using polynomial calcu- calculation of reference temperature for thrust
lations has built knowledge about output as the re- setting results in accurate thrust value for given
sult of known set of input values. For example: meteorological conditions thus eliminating ex-
what weight aircraft can have when taking off at cessive wear of engines.
20°C and wind of 10 knots? Knowing the trait of
neural functions, aircraft performance data can be 4.2 Real Life Example
calculated and optimized at acceptable (shorter) Let us present three distinct advantages of
time. LAPTOP performance calculation compared to
Understanding the environment in which LAP- manual calculation.
TOP is used, neural functions are the best solution Assume that aircraft is taking off from a runway
so far. Time consuming 1st Principle calculation is 02W at generic airport with known set of obstacles
out of question in short time available for planning along the take off path. Runway is wet, and atmos-
the next flight, while results after polynomial calcu- pheric pressure is lower than standard value of
lation have payload trade offs. Therefore neural 1013.25 hPa. Outside air temperature is 5°C. Air-
functions are solely applied in LAPTOP environ- craft is using anti ice system, while air conditioning
ment. system is set OFF for take off. There is wind of 10
4.1 Real Life Benefits kts (1kt0.514 m/s) blowing from 45°.
Basic tabulated performance is computed for
In real life today, the main aircraft performance dry runway, standard meteorological conditions (air
calculation is based on tabulated data. Simple pressure of 1013.25 hPa, defined set of outside
mathematical manipulations are used to reach fig- temperature values and a set of discrete wind values
ures necessary for setting thrust of aircraft power- blowing down the runway) and standard aircraft
plants and reference speed values for certain pilot configuration (Anti Ice System and Air Condition-
actions. Each combination of thrust setting and me- ing System not engaged). (Figure 1.)
4
Figure 2. Tabulated Performance Correction Figures
These baseline values are modified by pilots in higher temperature setting resulting in substantial
order to meet actual status at the moment of calcu- savings in annual maintenance costs.
lation (Values in Figure 2.). Each calculation can
be considered valid at the time of calculation only.
Approximate time for that process is not less than 5. CONCLUSION
10 minutes assuming quiet cockpit atmosphere and
no distractions. After the initial calculation there is LAPTOP based performance data generating
another recalculation check needed. If the calcula- and optimization as the idea is recent achievement
tion case is simpler (closer to baseline conditions in airline industry. It has arisen as a logical step
with less modifications time can drop to 7 minutes forward in today’s aircraft cockpit layouts.
at best). Numerous computers in contemporary cockpits
LAPTOP calculation time for the same input have got another addition in the form of a LAP-
values takes 15 s at maximum. It is very important TOP. Although it is still not integrated with other
that this time is for the case with maximum number computers completely, it provides data to be insert-
of adjustments to standard values needed. Any sim- ed as inputs to them. Quality inputs can not lead to
pler case takes from 10s to 14s. This is the first ad- faulty or erroneous outputs.
vantage - shorter calculation time. As the system has been designed by an aircraft
Comparing aircraft weight permitted for take off manufacturer and tested in a group of different air-
LAPTOP calculation in this example gives lines, it is meant to be adapted for commercial fly-
207114,5 kg, while pilots calculation based on tab- ing. That is obvious from the intended and already
ulated data leads to approximately 201.5 t. Differ- incorporated characteristics (Figure 3.):
ence of 5614,5 kg is just because of improved cal- simple user interface (suited to different com-
culation mode applied by LAPTOP and its numeri- puter knowledge levels),
cal precision. That is second advantage – unreal- calculation initialization through runway des-
ized revenue. ignator,
direct meteorological and aircraft condition in-
The third advantage is gained when aircraft is to puts,
take off at some weight lower that one permitted by the selection between maximum payload and
actual meteorological conditions and aircraft status. flexible take-off modes. [5]
LAPTOP calculation in given example leads to 1°C
5
All listed characteristics should lead to easier revenue in case of maximum payload take-off or
and more efficient usage and results. decreasing expenses in the long-term use of flexible
User interface should bring more prompt reac- take-off).
tion to constantly changing operational situation at The system presented in the article is not
any large airport (i.e. switching from one runway to unique in airline industry today. Some airlines have
another, additional cargo or less passengers due to been using other systems for a while. Although sys-
lost flight connection, etc.). Actual meteorological tems do not correspond to each other completely,
and aircraft condition data usage often allows extra all users agree – computer performance data calcu-
weight loading, hence increasing commercial ef- lation in cockpits is quality step forward in airline
fects of each flight. The choice between two take- industry.
off modes reflects on increased profit (by adding
mance and Operations Conference, San
6. REFERENCES Francisco, 1998. Chapter 47.
[5] Laval Chan Kam Fai, “Octopus Improve-
[1] Alexander Wells, “Air Transportation, a ments”, 10th Performance and Operations
management perspective”, Bellmont, Conference, San Francisco, 1998. Chapter
1989. p. 55 48.
[2] “LPC Evaluation Package”, Airbus Indus-
trie REF 945.7867/ 98, Toulouse, Novem-
ber 1998.
[3] Docus Michael, “Less Paper In the Cock-
pit”, 10th Performance and Operations
Conference, San Francisco, 1998. Chapter
06.
[4] Laval Chan Kam Fai, “Less Paper In the
Cockpit Takeoff Module”, 10th Perfor-