Themes Case Studies and Implementation Suggestions
Themes Case Studies and Implementation Suggestions
Themes Case Studies and Implementation Suggestions
net/publication/277291385
CITATIONS READS
2 751
2 authors, including:
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Opportunity Gap and Women in the Energy Infrastructure Workforce View project
All content following this page was uploaded by Martin Arthur Fischer on 18 November 2015.
By
STANFORD UNIVERSITY
Copyright © 2008 by
Center for Integrated Facility Engineering
Abstract
Virtual Design and Construction (VDC1) is the use of integrated multi-disciplinary
performance models of design-construction projects to support explicit and public
business objectives. This paper describes the theory and methods of VDC, and it includes
specific examples of models and precise objectives as well as detailed suggestions on
how to implement VDC in practice. VDC models are virtual because they show
computer-based descriptions of the project. The VDC project model emphasizes those
aspects of the project that can be designed and managed, i.e., the product (typically a
building or plant), the organization that will define, design, construct and operate it, and
the process that the organization teams will follow. These models are logically integrated
in the sense that they all can access shared data, and if a user highlights or changes an
aspect of one, the integrated models can highlight or change the dependent aspects of
related models. The models are multi-disciplinary in the sense that they represent the
Architect, Engineering, Contractor (AEC) and Owner of the project, as well as relevant
sub disciplines. The models are performance models in the sense that they predict some
aspects of project performance, track many that are relevant, and can show predicted and
measured performance in relationship to stated project performance objectives. Some
companies now practice the first steps of VDC modeling, and they consistently find that
they improve business performance by doing so.
1
Italics indicates that the glossary defines the italicized term
10/9/2009 1
Virtual Design and Construction
10/9/2009 1
Virtual Design and Construction
Background
This section gives an overview of VDC and relates it to the broader use of technology in
AEC practice.
10/9/2009 2
V
Virtual Desiign and Consstruction
Figure 1: This photo was taaken in a consstruction plan nning office in n 1998. Some important
chharacteristics of this processs are unchanged in the passt thirty yearss, e.g., there arre multiple
paarticipants in
n the processs, although probably man ny other impoortant stakeh holders are
abbsent from thhe meeting. We W suggest thaat most such photos will loook differentlly in thirty
yeears, and somme today alreaady do -- thosee from constrruction managgement projeccts that are
siignificantly more
m efficient and effectivee than most. The big chan nge, which already has
sttarted to occu
ur, is the emerrging use of in mputer-based visual projectt models to
nteractive com
im
mprove comm munication of project information, reducce latency and increase coollaborative
owwnership of project plans.
DARRPA supporteed Concurreent Engineerring researchh that led too an integratted Product –
Organnization – Process
P moddel and desiign methodoology [Londdoño 89]. LondoñoL et al
a
used a blackboarrd for comm municating and for contrrol of inform mation flow.. The domaiin
conceerned engineeering parts,, and the blaackboard dattabase descriibed the Prooduct, Processs
10/9//2009 3
Virtual Design and Construction
Industry Foundation Class (IFC) standards work of the International Alliance for
Interoperability [IAI 05] and [Froese 02] both discuss semantic models for data
exchange. More recently, the IFC and web XML data language and IFC communities
collaborated to develop methods that provide web-based standards for sharing IFC data
[aecXML 05]. The large Building Lifecycle Interoperable Software (BLIS)
demonstration project defined hundreds of “views” and about 100 “concepts,” which
were a “practical” subset of the IFC standard at the time [BLIS 02, 04]. The BLIS project
was the first major demonstration by multiple software vendors to create an integrated set
of project design models based on a shared IFC-based architectural model. The project
had active participation from 1999 – 2002, and it demonstrated that real CAD and
analysis applications could usefully share and exchange at least some of the data that a
design team needed to create a design. While the BLIS project received lots of energy
and attention, it had only limited success in its original hope to stimulate broad software
developers’ support of IFCs and industry use of them. We have had mixed experience
with use of computer application interoperability: it is possible to some extent, but
difficult and limited [Kam 02], even in our own university teaching and research, as
modest as they are. We find anecdotally that many AEC companies share similar
concerns.
Our VDC work commits to making explicit the semantics of data that practitioners of
different perspectives and applications need to share, and attempting to facilitate
practitioners to define and use shared explicit representations. Like the IAI effort, we
encourage the project team to identify and commit to a standard vocabulary. Further, we
recommend being inspired by standards such as the IFC and pragmatic like BLIS, but
even more modest than both. However, unlike the IAI standards, we do not propose
standards for the semantic details of VDC models and method; we pragmatically assume
that the long-term solution will take a long time to emerge, and in the short term, we want
to support individual project teams to do as well as they can with modest incremental
effort. Further, we encourage a strict discipline on both the level of detail of VDC models
and the process for creating whatever detail the project team wants, which we describe
below in the section POP models have different levels of detail.
Example
10/9/2009 4
Virtual Design and Construction
proceed sequentially through the steps in this maturity model, but some of the third step,
Automation, often requires minimal and specialized, not general Integration.
1. Visualization and Metrics: In this first phase, project teams create models of the
Product in 3D, of the Organization that performs design, construction and operations and
the Process followed by organizational participants to do design, construction and
operations and management, based on performance metrics that are predicted from
models and tracked in the process. The results of the CIFE-CURT VDC use survey find
that this stage is in common (although not yet widespread) use within the global AEC
industry [Kunz and Gilligan 07]. For Visualization to work well, all stakeholder
organizations need to develop the competence to interpret the visual models, and many
need to develop core competence to develop them, which requires a strategic investment
in the methods and their use. Similarly, for Visualization to work well for multiple
stakeholders, multi-party collaboration contracts need at least to allow and ideally to
incentivize data sharing, which may require strategic change in partnering arrangements.
In the Visualization phase, projects:
• Routinely model and visualize the most expensive elements of the Product,
Organization and Process (POP);
• Use a social process among project stakeholders to integrate multiple VDC
models and model versions;
• Justify investment in VDC tools, methods and human resources based on the
value proposition to the project, since this phase is (relatively) inexpensive and
individual projects receive can significant benefit;
• Clarify project objectives, values, responsibilities, designs and expectations
because good visualization enables many more stakeholders to participate in
project review far more meaningfully than in routine practice.
• Share data meaningfully among Product, Organization & Process models and
analysis programs using interoperation, i.e., reliable computer-based data
exchange.
• Cannot justify investment in VDC tools, methods and human resources based on
their project value proposition. Rather, the value proposition must support the
10/9/2009 5
Virtual Design and Construction
firm, since this phase is (relatively) expensive and multiple projects must use the
same methods for the investment to produce significant benefit.
• IFCs are designed to enable this process, but there is little evidence that they are
in significant use.
• Various vendors provide families of software applications that interoperate, often
using proprietary exchange methods, which still limit exchange with other
applications that might be useful to a project.
• Derive incremental value from integration per se because it can reduce modeling
effort and time.
3. Automation In this phase, projects use automated methods to perform routine design
tasks or to help build subassemblies in a factory. For Automation to improve design,
project organizations normally need to dramatically change their processes to enable or
perform more high-value design and analysis and spend much less time and billable
effort for routine design. To support fabrication, the project needs to change from
Design-Build or Design – Bid – Build to Design - Fabricate – Assemble, which takes
strategic commitment to support a new partnering arrangement. Automation requires
Integration, and good visualization helps make it work well. In the Integration phase,
projects:
Themes
VDC models are virtual
The current practice of AEC design and construction, as shown in Figure 1, obviously
works for developing value-adding projects today. Computer applications generated most
of the paper documents shown in this photo, such as the project schedule on the wall,
which is the output of a scheduling program, and the drawings, which are the output of a
CAD program. Paper documents today provide high-resolution descriptions of project
elements including architectural designs and plans, and the vast majority of AEC projects
in use today were created using these paper-based methods. However, the discrete paper
based documents do not help integration of different disciplines and making even simple
changes requires hours to days to make the initial change, print and review the updated
documents and do even simple updates to related documents of functionally related
disciplines. In addition, the format of today’s paper documents is often difficult for
10/9/2009 6
V
Virtual Desiign and Consstruction
diverrse stakeholders to undderstand: forr example, users can rarely makee meaningfuul
commments about 2D architecttural drawings or Gantt charts.
c
Like their forbeaarers in pracctice today, VDC modeels are compputer based. In additionn,
howeever, the use of the VD DC project models is flexible, vissual and intteractive, noot
docum ment or pap per-based. Thhe engineer who generaates the VDC C schedule can project it
and show
s it to other
o stakehoolders who have
h responnsibility for the
t CAD model or som me
area of
o the design n or construuction. They in turn can project theirr CAD moddels or projecct
digitaal photos, id
deally simulltaneously with
w the scheedule, each on a separaate projectioon
screeen such as thaat shown in Figure 2, a photo
p of a prroject meetinng taken in our
o CIFE labb.
Inexppensive commputers and large, highh definition and inexpensive projecction devicees
enablle social shharing of VDC
V compuuter models, and the modeling
m annd simulatioon
applications are now
n powerfuul and afforddable.
Figure 2: Photto taken in th he CIFE lab of a construcction planningg meeting usiing Virtual
D
Design and Co onstruction methods.
m As in
i the traditional method of Figure 1,, there are
m
multiple stakeh
holders in the meeting. Moddels of the product, organizaation and process can be
diisplayed, expllained and uppdated simulttaneously on the separate displays. Wee find that
deesign team perrformance immproves dramaatically compaared with in th he traditionall method of
Figure 1, and our
o goal is to simultaneoussly improve prroject team performance in n schedule,
coost and quallity dramatically. Interacttive computerr models rep place traditioonal paper
doocuments.
The VDC
V modell supports usse by multipple stakeholdders, as Figuure 1 showss occurring in i
curreent practice and Figure 2 shows inn the interaactive VDC process. Siince VDC is i
desiggned to suppport a multi--disciplinaryy project teaam, appropriiate stakehoolders includde
the multiple
m arch
hitects, enginneers and geeneral and multiple
m speccialty contracctors of AEC C
as well
w as own ner represenntatives, useers, supplieers, communnity represeentatives annd
goverrnment jurisdiction officcials.
VDC C creates ann integrated frameworkk and set of o methods to manage the projecct,
including those aspects
a of thhe project thhat must andd can be dessigned and managed,
m i.ee.,
10/9//2009 7
Virtual Design and Construction
the building, the design-construction process and the organizations that follow the
processes to design, build and use the building. Building Information Modeling (BIM)
focuses on the building elements of the VDC model, which we find useful but limiting
because management issues usually involve building – organization – process
interactions. BIM definitely appears to hold promise in practice [Bedrick, 05], [Haymaker
05]. BIM today is enabling many AEC professionals to improve performance. However,
even using best BIM practices, projects do not normally model, visualize or analyze the
organization and process accurately and effectively, and methods to manage and
communicate multidisciplinary information and processes remain ad-hoc.
Interactive VDC enables a very big change in the behavior of the design-construction
process: dramatic reduction in decision latency, or the time between posing a question
and having information with sufficient quality that it can be used to make a design
decision. Questions can be formal “Requests for Information” or informal inquiries of
fellow stakeholders. We repeatedly see latency change from days to hours and even
minutes in integrated design sessions (see the section below on Integrated Concurrent
Engineering).
Natural visual VDC models make the content of each model much more accessible than
they are in traditional static paper descriptions. Specifically, most stakeholders find that
interactive 3D models are vastly more understandable than static 2D plan and section
drawings, and 4D product-construction process animations are similarly much more
understandable than traditional project schedule Gantt charts. Our interactive project
models have started to become mutually parametric in the sense that change or
highlighting any one will lead to very rapid or even instant change or highlighting in all
others that are dependent. Because models can be examined with respect to each other,
each grows to support the issues of others; time to get explanations and make decisions
drops from days to seconds, likelihood of both design and construction rework drops
because relevant stakeholders have increased ownership and timely participation in
project decision-making. Since VDC models are visual, project team members who have
different native languages can all reference the same graphic models, providing some
support for the multi-cultural teams that are now common on many construction projects
10/9/2009 8
Virtual Design and Construction
worldwide as well as some larger design teams. Some organizations have started to use
multi-disciplinary VDC models as the focus of daily, weekly and major milestone design,
planning and review sessions.
Suggestion: Hold project kickoff, major review, weekly and daily design and
construction project meetings in a room with multiple computers, ideally
with at least three projected screens that all participants can see
simultaneously. Plan the agenda around description, analysis and evaluation
of product, organization and process issues as shown explicitly in models.
Invite all relevant stakeholders to the project meetings.
The POP model is object-oriented in the sense that each P, O and P element has defined
meaning (or semantics) to the stakeholders. For example, the Product model defines
building elements such as Floors, Walls and Beams; the Organization model defines
organizational groups, and the Process model defines activities and milestones.
We define two related types of POP models: generic and instance. Generic models
describe the conceptual vocabulary and thus can be very useful to define shared
vocabulary for project stakeholders at the time of launching a project. More generally,
generic POP models can define the vocabulary that a company or partnership uses to do a
kind of work, allowing a community of organization professionals to define shared
vocabulary that individual projects can customize as needed. Generic POP models define
entity names, such as column, Design Team and activity. They may define
associated attributes such as height, team responsibilities, and planned
duration. However, generic models lack specific detail, i.e., have no values for their
attributes, or names of individual instance elements. Instance models specialize the
vocabulary as their generic relatives, naming individual elements, such as Design
Team A. POP instances refer to corresponding objects in individual modeling or analysis
applications, such as entities in a CAD, organization or process modeling application.
They may refer explicitly to corresponding objects in modeling applications, or they may
refer implicitly if POP model users understand the model naming correspondence. They
may contain values of design variables such as planned dimensions when it is useful to
share those values across multiple models.
The POP model specifies information that is shared among models, not a complete
project model with which individual modeling applications send and retrieve information.
10/9/2009 9
Virtual Design and Construction
Thus, the POP model describes the content of the individual P, O and P models, each
which represents the details of designed conceptual elements with their attributes,
attribute values and relationships. The individual P, O and P modeling tools have user
interfaces that present the models using natural visual idioms that are appropriate for each
relevant discipline perspective. The purpose of the shared POP model is to define
conceptual elements that are shared and help the stakeholders to assure that the product,
organization and process specifications are appropriate and mutually consistent. For
example, the product model defines the physical elements to be designed and built, at
some selected but necessarily incomplete level of detail. The Organization model defines
the groups that design and build each defined physical element, and the process model
defines the activities and milestones that stakeholders follow to do their work.
The product model, and hence the Product segment of the POP model, should represent
the components and systems of the building. A well-designed POP model then also
includes organization entities to design and build the systems and components, as well as
the activities to do design and construction. The Industry Foundation Classes (IFC)
define a large set of product components and systems -- such as Floors, Walls, Beams,
and Equipment – as well as some definition of organizational and process entities [IAI
05, Tarandi 03]. The Scope section of coarse level of detail POP models will include only
a small subset of the total IFC specification. Projects that find it useful to define POP
models at several levels of detail may represent more of the IFC specification at the finer
levels of detail of their generic POP models.
At a high level, POP models represent the function, designed form or scope, and behavior
of the project product, organization and process. Figure 3, for example, shows a generic
POP model in the sense that it specifies the vocabulary used to describe a line of
business. Related instance models represent specific projects, such as that to create an
individual building.
10/9/2009 10
Virtual Design and Construction
Figure 3: Level-A generic POP model for design and construction of buildings. This most-
generic level of detail specifies that POP models represent the Product, Organization and Process
in terms of functional requirements or design intent, Form or Scope or design choices, and
behaviors or predictions and observations. Level “A” models (shown) have about one form/scope
element each for product, organization and process; Level “B” have about ten, level “C” about a
hundred, etc.
As shown in Figure 3, we define the content of the POP models using the classic
function – form – behavior taxonomy of design theory [Gero 90, Clayton 96]:
• Function, or design intent, represents the intent of the owner in making a
requirement or the requirement of a critical stakeholder such as the code
jurisdiction. Examples include that an auditorium seats 100; that an organization
include a licensed structural engineer; and that the design process include certain
specified review milestones.
• Form, or design choice – or designed scope -- represents the decision of a
designer in response to a functional requirement or the designer preference.
Examples include choice of specific spaces, the choice of a particular contractual
relationship among the architect and contractors, and the construction plan.
10/9/2009 11
Virtual Design and Construction
Figure 4 shows an example of a POP model that both specializes the generic model of
Figure 3 for a university dormitory project and elaborates its level of detail. For example,
the generic Level-A generic Product functions concern capacity and sustainability. At a
greater level of detail, the instance Level-B Product functions include requirements to
house 100 students, energy and water use that are stated fractions of 2002 comparables,
specific noise, air quality open space, recycling and height objectives, the assumption of
two-person dormitory apartments and a specific project budget. The generic Product form
or scope represents systems and physical building elements, which remain unstated. As
shown in
Figure 4, the Level-B instances include ten two-bedroom apartments per floor at 400
square feet each and elements that still need to be sized including solar panels, a
foundation pad, laundry, corridor, lab spaces, etc. In the judgment of the designers at this
(still early) design stage, these physical elements represented the top ten physical
elements in terms of cost. Their management goal is to design, procure and build these
elements with predictable and acceptable costs, thereby minimizing the overall project
cost risks. The team will elaborate the level of design detail once the design team has
significant confidence that the design of these physical elements, the responsible design-
construction team and the associated activities are all consistently specified and
acceptable. The generic Product behaviors include the predicted and observed (by post
occupancy measurement and evaluation) Product functions – those listed in the function
segment of the POP model – and periodic assessment during the design and construction
process of the conformance of the design to the design functional requirements. Many
researchers use model-based computational methods to predict behavior, such as [Dym
88], [Shea and Cagan 99], Flemming and Woodbury 95] [Stiny 80]. Additional product
behaviors include design team assessment of the conformance of the prediction of design
performance to each stated design requirement and measured observations of the final
building performance for each stated requirement, by appropriate post occupancy
evaluation (POE).
10/9/2009 12
Function Form/Scope Behavior
Project
Element Attribute
Virtual
ship
DesignObjective
Relation-
and Construction
Choice Predicted Observed Assessed
Product
Relation- Functional Product Scope
Product Scope ship Requirement (Space, System)
Product Building Spaces include Offices Offices
Product Building Spaces include conference rooms conference rooms
Product Building Spaces include public areas public areas
Product Building Systems include HVAC HVAC
Product Building Systems include telecom/network telecom/network
Product Building Physical Elements include foundation foundation
Product Building Physical Elements include above-ground steel above-ground steel
Product Building Physical Elements include drywall drywall
Product Building Physical Elements include skin skin
Product Building Physical Elements include windows windows
Product Building Physical Elements include roof roof
Product Objectives
Figure 4: an instance Level-B POP model represents the Function (intent), Form/Scope (design
choices) and Behavior (properties) of a project Product, Organization and Process. The broad
goals of the POP model are to help the stakeholder team to identify major requirements, the
most expensive design choices made by the design team to meet those requirements, and the
predictable and observable project behaviors early in the design process. The hope, and our
experience, is that the POP model helps enable the most valuable possible modeling and analysis
of a project during its entire lifecycle.
10/9/2009 13
Virtual Design and Construction
Figure 4 is an example. This initial Level-B POP model shows the P, O and P design
elements that, in the judgment of the project team, represented the elements that will
require the greatest cost, effort or schedule at this level of detail. The broad objectives of
the POP model are to help the stakeholder team to identify these resource users explicitly
early in the design process and to enable consistent modeling of those elements in the
associated product, organization and process models. Having developed such a model
and understanding its significance for managing the project, we can then elaborate its
detail to a Level-C POP model, which represents those POP elements with about 1% of
the cost, effort or duration. While AEC projects often define Level-D or greater level of
detail, we focus our research and this paper on Level-A to –C models.
We borrow the convention of “ABC” from the method of Pareto analysis, or ABC
analysis. With respect to the total cost, schedule or quality, we try to guild POP models
that identify the ten most important factors in the “A” category. We try to identify the ten
or so most important sub factors of each “A” category as the hundred factors in the “B”
category, and in turn we will identify the ten sub factors of each B category as members
of the relatively unimportant “C” category.
The greatest value of the POP model comes when the project influence is greatest, i.e., at
the schematic and early design development phases. At these early phases, by definition,
only Level-A or –B and possibly -C details
Figure 5 shows a set of guidelines for how to create individual functions (design intent),
form/scope (design choices) and behaviors (parameters) of a project. The modeling
purpose might be to support understanding of architectural concepts, space management
(i.e., how much space is planned and actually available for different spatial functions),
10/9/2009 14
Virtual Design and Construction
POP models represent three facets of the project, and the models start with generic
Product, organization and Process (Work) Breakdown structures. A goal of a project POP
model is to define the PBS, OBS and WBS so that they represent the important
characteristic of their respective project models and are mutually consistent in both
naming and references. The scopes of each row of a POP model should be consistent with
the corresponding PBS, OBS or WBS.
Suggestion: Create a Level-B instance POP model very early in the design
process, at least by the end of the first day of a kickoff meeting. Start to
elaborate the level of detail to Level-C only after the Level-B elements are
defined, modeled, mutually consistent, acceptable and well understood.
The POP model represents breakdown structures hierarchically. Figure 6 - Figure 8 show
a generic PBS, OBS and WBS, which represent a hierarchical decomposition of types of
product elements, organization elements and work respectively. The individual product,
organization and process models will use the names defined within the breakdown
structures and shown together in the Form segments of the POP model. The relationships
in each breakdown structure represent a class-subclass specialization hierarchy. That is,
Product, Organization or Work elements at the top of each BS are abstract; lower levels
become increasingly more specific and specialized.
The PBS represents the physical and abstract components that together represent the
physical and functional facility being built. The PBS represents both physical
components to be designed and built, such as columns and slabs, as well as abstract
systems such as egress and ventilation systems. The PBS shown in Figure 6 is based on
the Industry Foundation Classes (IFC) specification [IAI 05, Tarandi 03]. In the simplest
generic POP models, entities have only name; there is no explicit representation of their
attributes. Generic POP models can also define the names of the most important attributes
10/9/2009 15
Virtual Design and Construction
of each entity type. The relationships among P, O and P entities should be clear to the
stakeholders, but they are normally implicit in simple POP models, as shown in Figure 5.
10/9/2009 16
Virtual Design and Construction
Figure 5: This table shows a set of guidelines for how to create individual Functions,
Form/Scopes and Behaviors and how they appropriately relate to each other. Analysts can
predict or measure behaviors by day, week, month, major milestone, etc. as appropriate. R:
indicates relationships among POP elements, which normally are required for a consistent
POP and project model, although they are implicit in the model itself. For example,
individual Product Form or Scopes specify spaces, components and systems, which the
design team chose individually and collectively to satisfy product Functional objectives.
TCE2 is the total predicted Time, Cost, Effort or life cycle Energy use.
As shown in Figure 6, the generic PBS has multiple levels of detail, each of which can
have one or multiple corresponding project instances.
Suggestion: Create the generic Product Form segments of the POP model to
be consistent with the generic PBS at corresponding levels of detail. Design
similar consistency between Organization and Process Form elements of the
generic POP model and the OBS and WBS.
The OBS represents the vocabulary to describe the organization design, specifying the
organizational elements that do the work of the WBS to create the building of the PBS.
Nodes higher in the OBS hierarchy have responsibility for management, oversight, and
10/9/2009 17
Virtual Design and Construction
resolving the exceptions identified by lower-level organizational teams. The OBS should
describe all the groups with responsibility for significant activities in the WBS; normally
they will appear as lower nodes in the OBS. The relationships in the OBS specify
information flow: lower-level (rightmost in Figure 4) organization elements pass issues
that require executive resolution “up” the (or left and up) hierarchy to the next-higher
level or supervising element, rising if necessary to the ultimate decision making element,
which is at the top of the hierarchy. The generic OBS defines positions, not individual
people, although an instance OBS might name individuals. Each position in the OBS
might include one or more than one individual, each working part or full time on the
project. The project management system records the separate assignment of people and
other resources to organizational groups.
Suggestion: The generic OBS elements at each Level of Detail (LOD) have
responsibility to design or construct the deliverable Product elements in the
PBS at a corresponding LOD. POP Organizational Form elements have
similar responsibility for POP Product Form elements.
The WBS represents the work design, i.e., the activities that the organization performs to
design, build and manage the project of the PBS. The generic Work Breakdown Structure
defines the types of project design and construction deliverables. The deliverables may be
physical products such as built elements of the product, abstract products such as designs
or reports, or services such as continuing supervision. The WBS describes the work to be
done to create the product, not the functions or attributes such as cost of those product
elements.
10/9/2009 18
Virtual Design and Construction
Figure 8 shows a generic WBS and some representative instances of those generic
elements. The WBS elements at Levels 1-3 have corresponding elements in the process
Form/Scope segment of a typical POP model, and individual activities and operations
specified generically by WBS elements 5–7 constitute the typical construction master
schedule.
Modern 3D CAD product models look to a viewer the same as their traditional
counterparts, but they are built using “objects” that the computer modeling system
recognizes as physical elements such as columns, doors and windows and that appear in
the user interface as meaningful visual representations of the modeled element. The
modeling tool understands the number, location and properties of each such object, and
the tool can export a project model in a computer readable format that other computer
applications can interpret meaningfully. VDC models of products, organizations,
processes and the integrated POP models all are object-oriented in the sense that each can
represent a set of project elements using a vocabulary that the modeler specifies. Using a
VDC methodology, the project will build these product, organization and process models
10/9/2009 19
Virtual Design and Construction
using consistent vocabulary and mutually consistently with each other, to enable
concurrent management of the people, the work, and the unfolding project itself.
VDC models show the physical elements of the product and the abstract elements of the
organization, i.e., the teams or “actors” and the abstract elements of the work process,
i.e., the activities. The POP model lists these physical and abstract project elements, and
the individual VDC models show them in a visually meaningful way, describe their
attributes and attribute values, and describe the dependencies among them.
Based on organization theory, the “Virtual Design Team” method creates a computational
model of a project organization and the process followed by the organization to build the
project [Jin et al. 95, Kunz 98, Levitt 02]. The VDT organizational models consistently
describe and predict the behavior of both organizations and processes, including task and
project durations and the volume and distribution of direct work by actors and of
“hidden” work, which is the sum of coordination effort, rework and wait time for
information or decisions. It also predicts the time-varying actor backlog due to the
cumulative time demands on an actor of direct and hidden work in excess of available
time. Finally, it also predicts risks that task durations will exceed the nominal
(conservative) Critical Path Method (CPM) predicted task durations due to the impact of
“hidden” work and actor backlogs.
10/9/2009 20
Virtual Design and Construction
Most issues of latency involve organization design and management. One actor asks
information or decisions of another, and the dependent actor lacks time, knowledge,
information, authority or motivation to reply promptly. Meanwhile, the first actor must
wait on the issue at hand before proceeding. A project can reduce latency dramatically by
making the requirements for coordination and the objectives on timeliness both explicit.
From the perspective of a field manager, however, the project objectives are like the
sound coming from an AM radio; they are the final observed outputs, not the decisions a
line manager can make or the “knobs” that the operator might turn. Thus, we identify a
set of factors that a manager controls day by day, as well as a set of process performance
metrics that the project team can measure and use to judge how well the management
choices are moving toward the final project outcome objectives.
10/9/2009 21
Virtual Design and Construction
Operationally, each organization must identify the factors to control, process metrics to
monitor and use in management and outcomes by which to evaluate project success.
From this story, which is difficult but extremely useful to develop, companies can then
identify a strategy and annual plans for making the changes that it judges will help realize
its future breakthrough objectives. Each company is completely in control of its VDC
vision, measurable objectives and implementation strategy.
Suggestion: Each project should set, track and manage against a small (2 – 3)
set of explicit objectives of each type:
• Controllable factors, including the VDC modeling and analysis
strategy, process objectives to measure and one or two additional
factors (see Table 1).
• Measurable process performance parameters, such as schedule
conformance and response latency (See Table 2).
• Measurable project outcome objectives, such as safety, schedule, cost
and functional quality as assessed by post occupancy evaluation (See
Table 3).
10/9/2009 22
Virtual Design and Construction
Table 1: This table shows controllable project factors that can be made strategically by
organizations and line managers on a daily basis. Performance of these factors can be
measured, reported to the project team weekly or bi-weekly, and used in management. The
theory of VDC is that attention to the controllable factors leads to improved process
performance, which is measurable, and in turn to improved project performance that can be
reported to the owner and senior management.
VDC strategy and The plan and strategy concern:
plan • Visualization involves showing elements of the product, organization
and process in a way that different stakeholders can understand and
relate to them. A project can choose modeling level of detail and
focus considering contribution of the elements to total estimated
project Time, Cost, Effort or life cycle Energy use (TCE2).
• Integration includes definition and support of relationships among
modeled product, organization and process elements that enable the
computer to update values of dependent elements when an
independent value changes as well as make parametric change,
cross reference and appropriate highlighting of related elements in
different models. Considering their respective capabilities and
limitations, a project team can choose the tools to use and methods
to use them to enable different levels of social and technical, i.e.,
computer-supported, integration.
• Automation is support by the computer for elaborating design details,
checking consistency, doing analysis, moving and processing
materials as part of prefabrication and assembly at the work face. A
project can choose the amount of automation to perform given
business objectives and the capabilities and limitations of different
tools.
Process objectives Of the candidate process objectives, as shown in Table 2, select about
to measure, track two to measure, track and use for management.
and use for
management
Decisions and Objective is to record descriptions of and decision choice rationale for
rationale recorded 100% of POP items with > 10 (or 2)% of the budged time, cost, effort or
energy TCE2
Coordination Objective is that >= 90% of all actual coordination activity among project
requests participants is planned (weekly), explicit, informed, public and tracked
Coordination support Objective is that 90% of all planned coordination activity is reported
(weekly) by intended recipients to have been timely and suitable
Prediction basis Objective is that >= 80% of all predictions by project designers are made
by theoretically founded and automated methods
Design versions Objective is 2 or more for >= 80% of all decisions that affect more than
10% (or 2%) of cost, effort or schedule
10/9/2009 23
Virtual Design and Construction
Risk management There is an explicitly defined risk assessment and management strategy
strategy that is followed on 100% of POP items with > 10 (or 2)% of TCE2
Globalization Objective is that >= 50% of project purchased components and services
strategy and plan can be acquired from global suppliers
Lifecycle cost factors Objective is that project lifecycle costs explicitly model financial costs and
considered value returned, natural resources consumed, and emissions generated
Table 2: This table shows quantitative project process performance measures that can be
measured, reported to the project team weekly or bi-weekly, and used in management.
Achieving these process objectives makes it more likely that projects will reach aggressive
overall project objectives.
10/9/2009 24
Virtual Design and Construction
The breakthrough objectives are “reach” objectives in the sense that they will be
achievable only given success in practice of VDC visualization and metrics, integration
10/9/2009 25
Virtual Design and Construction
and automation, plus significant but necessary enabling changes in the processes of
design and construction.
10/9/2009 26
Virtual Design and Construction
Table 3 suggests our view that there are opportunities to achieve breakthrough
performance in the measured effectiveness of AEC practice as provided to clients.
Achieving breakthrough objective performance, however, requires commitment to the
objectives by clients and the providers of AEC services and products. We suggest that
breakthrough performance objectives – such as these we give for 2015 – are possible in
practice with effective use of VDC methods but will not be possible without effective use
of such methods.
Normally, as discussed above in the VDC maturity model section, we find that VDC
emerges in three stages: visualization, which is easily justified and implemented by a
project; integration of multiple models and segments of the business; and automation to
perform some significant portion of design or construction far more rapidly and reliably
than in traditional practice. The latter two stages require corporate commitment, as does
implementing the visualization stage consistently throughout an organization.
10/9/2009 27
Virtual Design and Construction
Integration: automated methods can relate the product, organization and process models
VDC project models are, by design, multi-disciplinary. For example, they include
product, organization and process models, and the product model often includes physical
component models and systems such as structures, egress and energy. The organization
model represents all the parties with significant project responsibilities, and the process
model represents the milestones and tasks of the organizational entities to develop the
project. In the Visualization phase, designers manually create consistency among
different models. The costs include effort and calendar time to build models and often
reenter data from one model into another, time and effort to check for consistency, to
analyze and use in management, and time and the design, construction or operational
costs of the inevitable failures to maintain consistency. Although automated systems
integration remains elusive in practice, some software vendors provide reliable and useful
automated data exchange within the family of models they sell and high level of
consistency among models. In addition, some software vendors support the Industry
Foundation Class (IFC) data standard, which is designed to enable automated integration
and interoperability. Ideally, automated Integration assures that the content of one model
is propagated appropriately to other models, including both the choices of designers and
the predictions of models.
10/9/2009 28
Virtual Design and Construction
Today, for example, the Heathrow Airport project detailed and pre-assembled rebar cages
on a cycle time of one week or less, including detailing, fabrication, assembly, delivery,
installation, and concrete pour [Kunz A and Ballard 04]. Many projects now do rapid
design and pre-assembly of common large systems such as bathrooms for offices,
kitchens for homes, and equipment spools for process plants. The job site installs these
pre-assembled systems far more rapidly than it ever does field construction, with gain in
schedule performance, final product quality, cost reliability, and often of base cost.
Since the middle 1990s, a Jet Propulsion Laboratory (JPL) design group called “TeamX”
has created conceptual stage designs of space mission within a few weeks, down from the
year or so in its traditional practice [Mark 02]. Their method defines the functional
objective, design response and makes many predictions about the cost, schedule and
performance of the proposed mission project. The project design explicitly considers the
physical and systems design of a physical vehicle, the organization to do the design,
manufacturing and operations, and the processes of the design – manufacturing –
operations team. The cost for a design study also dropped markedly; and the reported
10/9/2009 29
Virtual Design and Construction
The JPL TeamX has evolved a culture and set of methods to design space missions at a
vastly accelerated pace in comparison with traditional design methods. Researchers now
call this technology-mediated method Integrated Concurrent Engineering (ICE)
[Chachere 04]. While the engineering details of space missions for JPL and building
projects for AEC are different, both use projects to build new capabilities, and the
projects involve collaboration among multiple disciplines and multiple stakeholders with
a mixture of shared and competing objectives and methods.
We had the goals to develop multi-discipline project models collaboratively and very
quickly that support highly effective multi-stakeholder design review and approval.
Based on careful observation of the JPL method, we formalized, extended, specialized
and then implemented the ICE method in our VDC work and practices. We find that it
provides an exceptionally effective methodology for teaching and applying VDC
methods.
The ICE method attempts to remove most non-value adding diversions from the attention
of the design staff as they participate in an ICE session, such as clarifications of goals,
methods or vocabulary, secondary responsibilities and waiting for responses to questions
from fellow stakeholders. In the absence of diversions for designers and with technology,
methods and skills to do very fast design and analysis, the design team achieves response
latency of about a minute in greater than 99/100 inquiries by all members of the design
team, which in comparison with routine practice is both very fast and very reliable.
Independent observers report that CIFE ICE teams achieve rapid design project
completion and generally high quality design product.
We found that a set of complex, interrelated factors enable high level ICE performance as
shown in Table 4.
Table 4: Factors that enable Integrated Concurrent Engineering. In our
experience, each factor must be well managed to achieve high performance
ICE. Coordination latency is an observable but non-specific indicator of
10/9/2009 30
Virtual Design and Construction
10/9/2009 31
Virtual Design and Construction
Based on theoretical analysis and observation, we find that ICE teams at JPL manage ten
enabling factors that lead to exceptionally low information response latency, and
consequently to a dramatic improvement in project duration over traditional methods. A
carefully designed network of knowledgeable participants who have the skills and culture
to work independently of much central management direction, along with rapid, precise,
and semantically rich communication of design intent, choices and predictions, are two
features of the ICE approach that shrink response latency to near zero. We present
2
Scale-free networks have an exponential distribution of network links density, i.e., a few highly
centralized nodes and most with low connectivity, while Scaled networks have a normal and generally
more even link density distribution.
10/9/2009 32
Virtual Design and Construction
Giga Consultants developed the Total Economic ImpactTM (TEI) to allow analysis of both
the value and the cost of software system investments [Cormier 02, Gliedman 03].
Macomber described its use in AEC [Macomber 03]. The method makes the value
proposition explicit and quantitative, not just the costs, and it establishes quantitative
assumptions and specific measurable objectives for business performance that
responsible parties can track and manage to achieve. The method considers the
anticipated effects on owner value and costs of investments. Values include revenues due
to increase in market size as well as reduction in costs due to increased operating
efficiency. Costs include the immediate and support costs of the investment.
We apply the total economic impact method to elucidate the value proposition of
technology investments, specifically for investment in use of VDC for individual projects
and for companies. The method requires estimating any change in revenue or cost of
subcontracted work, plus cost of an investment and service of it. Thus, the method
requires making assumptions about both how a new investment will be used and its
impact on the business. While many engineers and managers resist making such
assumptions, doing so creates a coalition in favor of an investment, which is good.
We use the method not to say what the revenue and cost changes will be but rather as a
mechanism to set explicit, specific and public objectives. The investment advocate must
identify individuals within the team who will “sign up” for specific revenue and cost
10/9/2009 33
Virtual Design and Construction
numbers assuming a technology investment. Using the TEI method, the advocate and the
management team can then identify the total economic impact of making those numbers.
The method will predict the pay back time for the investment, given the assumptions. The
development, business, and management teams can then make a collective decision to
invest because the payback period and risks in the estimates are acceptable, or choose not
to invest.
Consider the example of Figure 9. This model assumes that IT costs are amortized and
paid in the three years after the investment. The model further assumes that cost of
contracted work drops due to the investment, because of better designs and cost
management, but that the cost of self-performed work increases because of using the
system. The TEI modeler adjusts the revenue change so that the payback period is
acceptable for the company. If the marketing team will commit to the revenue change
number (2% in this case), and the contract management and self-performed work teams
also commit to their numbers, then management can predict that the investment will pay
for itself in about two years. If these commitments have acceptable risk in the eyes of the
team and management, and the payback period is sufficiently short, the investment is
justified. The method identifies the team that must take responsibility for the business
success of the investment, the commitment each subteam must make, and specific
quantitative business objectives for each subteam. In addition, the model suggests some
of the measurements the team should make and use in management, i.e., its costs of
contracted and self-performed work.
Baseline Year-1
Rate ($K) Change (K$)
Revenue 100,000 2% 102,000
Cost of contracted work 85% 85,000 -2.0% 84,660
Cost of self-performed work 10% 10,000 2.0% 12,240
Gross Margin 5,000 5,100
Sales, G&A 2% 2,000 2,040
IT investment 70
Amortized costs of IT/yr 33% 23
Net income 3,000 3,037
Time to payback (years) 1.9
Net Income change (%) 1.2
Figure 9: Simple Total Economic ImpactTM model applied to a simple pro forma financial
summary of a representative General Contractor (GC). This model assumes typical revenue,
10/9/2009 34
Virtual Design and Construction
costs of business and a relatively comfortable baseline net income for a typical GC. With the
assumptions given, following the investment, the company must both predict and commit to
increasing revenue by at least 2%, reducing cost of contracted work at least 2% and
increasing cost of self-performed work no more than 2%. The team making this
commitment constitutes the coalition in favor of the innovation. If the team makes its
numbers, the investment will pay for itself in slightly less than two years.
Simple analysis of TEI models for AEC leads to a number of important conclusions.
First, assuming that the investment level is modest (e.g., $70K as we assumed), VDC
technology investment can be a feasible goal for a technology leading company in a
relatively good market. In fact the revenue increases we see due to VDC can be dramatic,
and the savings of contracted work can also significantly exceed the usual GC profit.
However, as the model indicates, an investing company requires superb execution to
make its commitments and simultaneously to preserve its margins on other work.
Assigning the best staff may make the investment successful but make the rest of the
business falter. Second, VDC investment is unlikely to make business sense in a bad
market. Third, if a new VDC method becomes established in a market, low-cost
companies will need to invest to stay in business as competitors grow. While their costs
of investment will be marginally lower because they are late in acquisition, their market
growth might actually be negative due to the earlier success of their competition.
iRoom
hardware
PC 2,500 4 10,000
Projector 3,000 3 9,000
Smart Board 2,000 3 6,000
installation 2,000 1 2,000
Software
4D 10,000 1 10,000
CIFE infrastructure 0 1 0
MS Project 1,000 1 1,000
Organization model 10,000 1 10,000
Budget 70,000
Figure 10: Representative costs to implement basic VDC capabilities in a company. These
costs assume purchase of a three-screen iRoom, such as that shown in Figure 2, purchase of
4D and organizational modeling software at certain assumed prices, some staff time, and
availability of good CAD modeling software. The TEI analysis of Figure 9 assumes this
bottom line number as the investment cost.
The most important conclusion of analysis of TEI models for AEC concerns the crucial
role of slack resources. The numbers of Figure 10 highlight the issue. If a good engineer
is available for assignment to support a discretionary investment, the staff cost is the sum
of direct plus indirect cost. In this example, at $62.50/hour loaded cost, the internal staff
cost represents a significant but not a dominating fraction of the total investment cost.
However, many AEC companies work with virtually no slack resources. In this case, the
10/9/2009 35
Virtual Design and Construction
cost of an internal staff member becomes the marginal cost of taking that engineer off a
critical project. Those costs can easily be $10K per day on projects with liquidated
damage risk. If the cost of the staff engineer changes from loaded to opportunity in a
fully committed company, almost no investment will ever reach a reasonable payback
time hurdle rate. The AEC industry is nearly universal in its culture of running with
negligible resource slack. This simple TEI analysis suggests that it is indeed rational for
companies to resist almost all investment, in the absence of slack. This simple example
also suggests the opportunity for companies that will make small pools of slack resources
available to support promising investments.
Summary
The theoretical basis of VDC includes several major components, which we discussed
above in the Themes section:
• Engineering modeling methods to represent the product, organization, and
process;
• Model-based analysis methods to predict the project schedule, cost, effort, hidden
work, organization, process, and schedule risks, 3D and 4D interferences;
• Visualization methods to present views of the product, organization and process in
ways that are clear for professionals and a broad class of interested stakeholders;
• Business metrics and methods to manage project processes using measured
performance; and
• Economic Impact, i.e., quantitative models of both cost and value of capital
investments, including the project as a whole, individual project elements, and
incremental investments to change the process.
Discussion
VDC strategy can enable companies to achieve significant
breakthrough objectives
Some organizations now find significant competitive advantage from their facilities or
their timeliness of developing new facilities. Other companies find that unbudgeted
changes to new facility development significantly impact corporate profitability. For all
these reasons, senior management has started in some organizations to place dramatic
“breakthrough” objectives on the capital development process. Some companies now
have senior management objectives to lower cost (say per square foot) by 20% while
improving quality and schedule; others want to reduce unbudgeted change to a few
percent. Others want to develop new facilities – from project approval through to high
value occupancy – in dramatically less time while preserving or improving quality,
schedule and cost performance. Uniformly, project delivery organizations accept that
10/9/2009 36
Virtual Design and Construction
they cannot make these objectives with incremental changes to their traditional
development processes. Repeatedly, companies find that a VDC strategy and
implementation plan is a crucial element of their plan to achieve breakthrough
organizational goals.
With hardware and software vendors and service providers all selling their offerings
aggressively, there is a technology push for VDC models today, and project members
often find them very appealing. In our experience, the biggest driver for VDC methods is
demand-pull: senior management of some AEC providers and some owners see the value
in a competitive market of “faster, better, cheaper” products and services, and they pursue
VDC methods as the best way to achieve such competitive advantage.
10/9/2009 37
Virtual Design and Construction
screen iRoom already makes the display, comparison and management of different
versions feasible.
VDC Limitations
The theoretical framework suggests limits of VDC, each of which we see in practice.
Important limits include:
• Owner management: VDC non-users report that lack of owner request or
initiative limits their interest and willingness to use VDC in practice [Kunz and
Gilligan 07]. Apparently, owners take a limited view of the potential value of
VDC, carefully assessing the costs but not its potential value for projects.
Because cost minimization is comparatively easy and relatively straightforward
managerially, owners frequently establish minimal apparent risk and minimum
first cost as crucial selection criteria for new projects, and they use similar
restrictive criteria as crucial criteria in deciding on incremental changes to a base
project design. Unless asked and encouraged, designers will not even ask the
question of life cycle value of projects or incremental changes to projects. In the
absence of asking about the value proposition seriously, the cost-value tradeoff
defaults to a cost minimization exercise.
• Project-orientation of the AEC industry: owners, designers and contractors all
have a culture and practice of project work. It is difficult, and often not advisable,
to make investments to improve processes when individual projects cannot justify
them. Even if successful, it is difficult to institutionalize the lessons learned about
how to use innovative methods effectively on subsequent projects.
• AEC industry culture: Architects, engineers and contractors all have a culture
and methods that minimize cost. With notable exceptions, many lack a culture
that seeks to maximize value. This culture follows owner preference, but it also
represents a culture that some AEC players accept in order to minimize their
short-term project risks.
• Sharp theoretical basis for VDC methods: The VDC modeling and particularly
the model-based VDC analysis methods are still undergoing theoretical
development. In an industry that appropriately values risk mitigation, the
changing theoretical foundation provides a handy and often an appropriate excuse
to avoid use of the methods.
• Learning: The AEC industry has several practices that limit the ability of
individuals, teams and companies from learning from experience. The ever-
changing project basis of the industry contributes to diffusion of experience,
rather than systematic learning [Taylor 04]. Owners systematically fail to
commission post occupancy evaluation of new projects and recently completed
project development processes, and AEC providers do not do such studies on their
own.
• Tools that are capable and integrated: Users consistently report that VDC
modeling and analysis tools are difficult to use, support limited business
objectives, and do not integrate easily or well with other tools that the project
wants to use. The National Institute of Standards and Technology recently
published a report that attributes nearly a $16B annual cost to the lack of
interoperability in US capital facilities development [NIST 04]. Anecdotally,
10/9/2009 38
Virtual Design and Construction
individual projects incur real costs as developers recreate or reenter the same
information in their models, often developing design details several times though
different stages the design process. However, users report that they use VDC
methods and receive value from its use in spite of the limits that they
acknowledge.
Glossary
4D model: a model that links the 3D description of a product to be constructed with the
plan and time-based schedule to build it. A 4D animation shows the construction of a
project.
Activity: identifiable work to be performed by an actor using a set of resources to
complete an identifiable activity in a process. Actors and a process together define an
organization-process model.
Actor: a project group or individual stakeholder with responsibility for an activity in a
process. Actors and a process together define an Organization-Process model.
Behavior: predicted or observed measurement about an aspect or element of a design,
such as cost, schedule or capacity. Behavior is a major segment of POP models.
CD: See project phases
Conformance: percentage agreement of planned and measured schedule, cost or quality
data. A good project has high measured daily or weekly conformance (> 80%) of planned
2 or 3-week lookahead schedule to actual schedule performance. “Plan Percent complete”
(PPC)” is another term for schedule conformance.
Controllable factor: a condition that a designer or manager can actually control, such as a
design choice about a product, the choice of what teams and people to hire, and the
design of a work process. Controllable factors affect process performance and project
outcomes.
DD: See project phases
Form, or Scope: the choice made by a designer in response to a function requirement,
including physical elements such as a door and abstract elements such as design teams
and activities. Scope is a major segment of POP models.
Function: requirement for a project that must be met because it is intent of an owner or
comes from a municipality or usual design practice. Function drives choice by the design
team of designed form or scope. The designed and ultimately the built scope in turn affect
the behaviors of design, construction processes, such as schedule and cost, and the then
operations, such as energy use. Function is a major segment of POP models.
ICE: Integrated Concurrent Engineering, a way to organize a design team that enables
stakeholders from multiple disciplines to participate concurrently to develop integrated
project designs very rapidly.
LOD: Level of detail, which is a measure of the complexity of a model. The most abstract
(“Level-A”) have about one element in each major section; about ten elements in each
major section in Level-B, and increasing in detail in higher levels.
10/9/2009 39
Virtual Design and Construction
OBS: Organization Breakdown Structure, the definition of the names of generic teams
that design and build a product.
Organization: a team of people that does the work specified in a process to create a
product. POP models represent the Functions, Scopes and Behaviors of project
Organizations.
PBS: Product Breakdown Structure, the definition of the names of generic physical
elements in a physical product design.
POE: Post occupancy evaluation of the quality of a project by affected stakeholders
POP: Product – Organization – Process, the integrated perspective of VDC models,
representing the integrated functions, scope and behaviors of each.
Process performance metric: an aspect of project performance that a team can measure
frequently (hourly, daily or every week or two) and use to judge how well past
management choices (see Controllable factors) are moving toward the final project
outcome objectives.
Process: Activities and procedures followed by an organization team to create a product,
i.e., the work the organization does, or a statement of “what we plan to do.” Plan
activities are statements of the work to do including precedence relationships among
activities, and Schedule activities have a planned start and end. Activities have
responsible actors; they may have coordination and rework dependencies that identify the
other activities with which they must coordinate or that must initiate rework if an
individual activity encounters some sort of failure. POP models represent the Functions,
Scopes and Behaviors of project Processes.
Product: the physical or abstract deliverable of a project as the organization follows the
process, typically a building, facility or design. POP models represent the Functions,
Scopes and Behaviors of project Products.
Project outcome: an aspect of a project that is important and normally can be known only
at the project completion, such as final quality, cost, schedule and safety. Outcome
follows process performance and in turn controllable factors.
Project phases: traditional AEC phases include pre-project planning, which obtains
budget and initial zoning approvals; early Schematic Design (SD), Design Development
(DD), which adds system issues to the design, and Construction Document preparation
(CD), which is the final design phase. Design is followed by Construction,
Commissioning and Occupancy.
SD: See project phases
Stage of VDC implementation: VDC emerges incrementally. Normally the first stage is
visualization to support understanding and decisions of an individual project team, then
systems based integration of multiple models to facilitate description, explanation,
evaluation and prediction of their behaviors, and finally as automation of significant
portions of design and construction activity.
TCE2: Total estimated project Time, Cost, Effort and life cycle Energy use.
10/9/2009 40
Virtual Design and Construction
TEI: Total Economic Impact®, a simple pro forma financial model that shows both the
value added to a business and the costs incurred of a technology investment.
VDC: Virtual Design and Construction, the use of integrated multi-disciplinary
performance models of design-construction projects to support explicit and public
business objectives.
Virtual model: a model in the computer of some aspect of a project. Virtual models can
complement and often replace physical models, and they can be built long before the
actual product, organization or process emerges in real life. Our experiences are that if a
team cannot build a project in the computer, it cannot build it in real life, and that
building virtual models can significantly decrease project risks.
Visualization: presenting a model in a way that is meaningful to diverse stakeholders,
which is normally visual, such as a 3D model of a product, a network of actors and
activities for organization and process models, and time-based (4D) animations of
product construction as well as time graphs of building performance.
WBS: Work (or Process) Breakdown Structure, the definition of the names of generic
activities to design and build a project.
Acknowledgements
We appreciate the encouragement, technical help and financial support of the CIFE
member companies over the years that we have taken to develop this work. CIFE has
provided the majority of the financial support to do this work. We also appreciate the
passion, skill and hard work of dozens of our wonderful students, and specifically Calvin
Kam who developed the models in our test case example with the help of his colleagues
and sponsors at the US General Services Administration.
References
[aecXML 05] North American International Alliance for Interoperability (IAI) chapter
web site, last accessed 5 October 2005, http://www.iai-na.org/about/resources.php
[ASTM 05]: American Society of Testing Materials (ASTM): Standard Classification of
Building Elements and Related Sitework - UNIFORMAT II, last accessed 26 September
2005, http://www.uniformat.com/
[Bedrick 05] “BIM and Process Improvement,” AECbytes Viewpoint #20
December 13, 2005, http://www.aecbytes.com/viewpoint/issue_20.htm, last accessed 23
December 2005.
[BLIS 00] BLIS Software demonstrations: IAI International Council Summit, October
2000, website last accessed 6 October 2005, http://www.blis-
project.org/demos/index.htm, 2000.
[BLIS 02] BLIS project website, last accessed 6 October 2005, http://www.blis-
project.org/index2.html, August 2002.
[Chachere 04] Chachere, John, Kunz, J., and Levitt, R., “Observation, Theory, and
Simulation of Integrated Concurrent Engineering: Grounded Theoretical Factors that
10/9/2009 41
Virtual Design and Construction
[Clayton 96] Clayton, M. J., J. C. Kunz and M. A. Fischer, “Rapid Conceptual Design
Evaluation Using a Virtual Product Model,” Engineering Applications of Artificial
Intelligence, Vol. 9, No. 4, Elsevier Science Ltd., pp. 439-451, 1996.
[Cormier, 02] Cormier, Bob, "The Total Economic ImpactTM (TEI) of Deploying
Network Appliance’s NearStore Product for Backup and Recovery," Giga Consulting
Group, last accessed 6 October 2005,
http://www.netapp.com/tech_library/ftp/analyst/ar1010b.pdf (September 2002)
[Dym et al. 88] Dym, C. L, R. P. Henchley, E. A. Delis and S. Gonick, “A
knowledge-based system for automated architectural code checking,” Computer-Aided
Design 20(3): 137-145, 1988.
[Flemming and Woodbury 95] Flemming, U., and Woodbury, R., “Software
Environment to Support Early Phases in Building Design (SEED): Overview,” Journal of
Architectural Engineering, Vol. 1, 147-152, 1995.
[Froese 02] Froese, Thomas, Current Status and Future trends of Model Based
Interoperability, eSM@rt Conference Proceedings Part A, pp. 199-208, University of
Salford, 2002, ISBN 0902896415.
[Garcia 04] Garcia, A.C.B., Kunz, J., Ekstrom, M., and Kiviniemi, A. 2003], Building a
Project Ontology with Extreme Collaboration and Virtual Design and Construction,
Advanced Engineering Informatics, Volume 18 # 2, April 2004, Pages 71-83, also CIFE
TR152, http://cife.stanford.edu/online.publications/TR152.pdf
[Gero 90] Gero, J.S. (1990). “Design Prototypes: A Knowledge Representation Schema
for Design,” AI Magazine, 11(4), 26-36.
[Gliedman 03] Gliedman, C., “the total Economic ImpactTM of Deploying the Sun-
Sybase Enterprise Data Warehouse Reference Architecture,” Giga Consulting Group,
http://www.sun.com/products/architectures-
platforms/wp/GigaInformationGroup_TEI_Sun-Sybase.pdf (August 2004)
[Haymaker 05] Haymaker, John; Ayaz, Engin; Fischer, Martin; Kunz, John; Kam,
Calvin, Ramsey, Marc; Suter, Ben and Toledo, Mauricio, “Methodologies to Manage and
communicate multidisciplinary design processes and information for the Stanford
University Green Dorm Feasibility Study, ITCON, Fall 2005
[Hunt 96] Hunt, Daniel V., Process Mapping: How to Reengineer Your Business
Processes, John Wiley and Sons, 1996.
[IAI 05] IAI international web site, last accessed 26 September 2005, http://www.iai-
international.org/Model/IFC (ifcXML) Specs.html
[Jin et al. 95] Jin, Yan, Raymond E. Levitt, Tore Christiansen, and John C. Kunz, “The
Virtual Design Team: A Computer Simulation Framework for Studying Organizational
Aspects of Concurrent Design,” SIMULATION Journal, Vol.64, No.3, pp. 160-175,
March 1995.
10/9/2009 42
Virtual Design and Construction
[Johansson 02] Johansson, B., Fox, A., and Winograd, T. “The Interactive Workspaces
Project: Experiences with Ubiquitous Computing Rooms,” IEEE Pervasive Computing
Magazine 1(2), April-June 2002.
[Kam 02] Calvin Kam and Martin Fischer, PM4D Final Report. CIFE technical Report
143, Stanford University 2002, also available at
http://www.stanford.edu/group/4D/download/c1.html
[Koo and Fischer 00] “Feasibility Study of 4D CAD in Commercial Construction”,
Journal of Construction Engineering and Management, ASCE, 126(4), 251-260, 2000.
[Kam 04] Kam, Calvin and Fischer, Martin, “Capitalizing on early project decision-
making opportunities to improve facility design, construction and life-cycle performance
POP, PM4D and decision dashboard approaches”, Journal of Automation in
Construction, 13, 1, pp. 53-65, Elsevier, 2004.
[Kunz and Gilligan 07] Kunz, John and Gilligan, Brian “Value from VDC / BIM Use,”
last accessed 13 April 2007, http://cife.stanford.edu/VDCSurvey.pdf
[Kunz 05] Stanford Civil and Environmental Engineering class CEE243, last accessed
26 September 2005, Virtual Design and Construction,
http://www.stanford.edu/class/cee243/.
[Kunz, A and Ballard 04] Kunz, Alex, and Ballard, Glenn, "Opportunities Offered by
Computer Modeling for Restructuring Construction," Proceedings of the 2004 CIB World
Congress, May 2004.
[Kunz 98] Kunz, John, Christiansen, Tore R., Cohen, Geoff P., Jin, Yan, Levitt,
Raymond E., “The Virtual Design Team: A Computational Simulation Model of Project
Organizations,” Communications of the Association for Computing Machinery, pp. 84-
92, November, 1998.
[Levitt 02] Levitt, Raymond and Kunz, John, “DESIGN YOUR PROJECT
ORGANIZATION AS ENGINEERS DESIGN BRIDGES,” CIFE Working Paper #73,
2002.
[Londoño 89] Londoño, F., Cleetus, K. J. Reddy, Y. V. “A Blackboard Scheme for
Cooperative Problem-Solving by Human Experts”, CERC-TR-TM-89-001, Concurrent
Engineering Research Center, West Virginia University, last accessed 26 September
2005, http://www.cerc.wvu.edu/cercdocs/techReports/1989/cerc-tr-tm-89-001.pdf.
[Macomber 03] Macomber, John, “Follow the Money: What really Drives Technology
Innovation in Construction”, ASCE Construction Research Congress, Proceedings of
Construction Research Congress, March 19-21, 2003, Honolulu, Hawaii; Sponsored by
Construction Institute - Construction Research Council, American Society of Civil;
Engineers Construction Engineering and Management Program, University of Colorado
at Boulder, edited by Keith R. Molenaar, (E) and Paul S. Chinowsky, Reston, VA, 2003
[Mark 02] Mark, G., "Extreme Collaboration" Communications of the ACM, Volume 45,
Number 6, pp. 89-93. 2002.
10/9/2009 43
Virtual Design and Construction
[NIST 04] "Cost Analysis of Inadequate Interoperability in the U.S. Capital Facilities
Industry," NIST GCR 04-867, last accessed 26 September 2005,
http://www.bfrl.nist.gov/oae/publications/gcrs/04867.pdf (August 2004)
[Prasad 96a] Prasad, B., Concurrent Function Deployment – an Emerging Alternative to
QFD: Conceptual Framework, Advanced in Concurrent Engineering: Proceedings of
CE96 Conference, USA, pp. 105 – 112.
[Prasad 96b] Prasad, B., Concurrent Engineering Wheels, in The RASSP Digest - Vol. 3,
1st. Qtr. 1996, last accessed 26 September 2005,
http://www.eda.org/rassp/documents/newsletter/html/96q1/news_6.html
[Shea and Cagan 99] Shea, K. and Cagan, J., “The Design of Novel Roof Trusses with
Shape Annealing: Assessing the Ability of a Computational Method in Aiding Structural
Designers with Varying Design Intent,” Design Studies, Vol. 20, 3-23. 1999.
[Stiny 80] Stiny, G., “Introduction to Shape and Shape Grammars,” Environment and
Planning B: Planning and Design 7, 343-351, 1980.
[Tarandi 03] “Editorial: IFC - product models for the AEC arena,” ITcon Vol. 8, Special
Issue IFC - Product models for the AEC arena, pg. 135-137,
http://www.itcon.org/2003/11, last accessed 3 January 2006.
[Taylor 04] Taylor, J. and Levitt, R., “Understanding and Managing Systemic Innovation
in Project-based Industries.” In: Slevin, D., Cleland, D. and Pinto, J. (Eds.), Innovations:
Project Management Research 2004, pp. 83-99. Project Management Institute, Newton
Square Pennsylvania, 2004.
[Teicholz 04] Teicholz, Paul, “Labor Productivity Declines in the Construction Industry:
Causes and Remedies” AECbytes Viewpoint #4, April 14, 2004,
http://www.aecbytes.com/viewpoint/issue_4.htm, last accessed 3 January 2006.
10/9/2009 44