Architect

Design processes
This page intentionally left blank

Design processes
What Architects & Industrial Designers can teach
each other about managing the design process
Edited by: Wim Poelman and David Keyson

Edited by: Wim Poelman and David Keyson
Communication and Layout: Matty Cruijsberg
Graphic design: Janita Han
2008 The authors and IOS Press. All rights reserved.
ISBN 978-1-58603-945-5
Published by IOS Press under the imprint Delft University Press
Publisher
IOS Press BV
Nieuwe Hemweg 6b
1013 BG Amsterdam
The Netherlands
tel: +31-20-688 3355
fax: +31-20-687 0019
email: info@iospress.nl
www.iospress.nl
www.dupress.nl
LEGAL NOTICE
The publisher is not responsible for the use which might be made of the following
information.
PRINTED IN THE NETHERLANDS

1
Preface
Contents
Preface
Prof. dr. C.J.P.M. de Bont ________________________________ 3
1 Introduction
Dr. ir. W.A. Poelman ____________________________________ 4
2 Design Processes
Between academic and practice views
Dr. ir. H.H. Achten ______________________________________14
3 Visualization
Sketching is Alive and Well in this Digital Age
Prof. G. Goldschmidt ___________________________________ 28

4 Project Management
Project and risk Management in architecture and
industrial design
Prof. dr. ir. J.W.F. Wamelink and dr. J.L. Heintz _______________ 44
5 Social Complexity
Social complexity in design collaboration
Prof. dr. P.G. Badke-Schaub ______________________________60
6 Decision Making
A decision-based design approach ________________________ 68
Dr. ir. P.P.J. van Loon, ir. R. Binnekamp and ir. J. Burger
7 Technology Diffusion and Design
The metabolism of knowledge
Dr. ir. W.A. Poelman ____________________________________ 90
8 Closing speech
Prof. dr. ir. A.C.J.M. Eekhout _____________________________ 108
Appendixes:
1 Chairmans impressions
Prof. dr. ir. T.M. de Jong _________________________________112
2 Program ____________________________________________120

2

3
Introduction
Preface
This book is a result of cooperation between the Faculties Industrial Design Engineering
and Architecture of Delft University of Technology. It presents the content of a series of
papers presented at the hrst joint conference on Design Processes.
This conference was organized in a special timeframe. On the 13th of may the Faculty of
Architecture burned down. A few weeks later important part of the staff of Architecture
had moved in in the Faculty of Industrial Design Engineering which might have a
greater impact on the cooperation than the conference itself. Directly discussions
between scientists from both faculties started about possibilities for cooperation.
Nevertheless this conference and this book mark an important moment in the 40 year
history after Industrial Design Engineering sprouted from the Faculty of Architecture.
Also on behalf of the dean of the Faculty of Architecture, professor Wytze Patijn, I thank
the reviewers professor Arthur O. Eger and professor Jos Lichtenberg for the effort
they did for improving the scientihc quality of these papers. ! also thank professor Wim
Poelman and professor David Keyson for editing this book.
Professor Cees de Bont
Dean of the Faculty of Industrial Design Engineering

IDE+A
Design Processes - Wim Poelman and David Keyson (Eds.)
IOS Press, 2008 2008 The authors and IOS Press. All rights reserved.
Introduction

5
1 Introduction
Background
This conference has been organized in the context of the cooperation between
the faculties Industrial Design Engineering and Architecture of Delft University of
Technology.
In the second half of the sixties, Professor Joost van den Grinten took the initiative to start
an interfaculty for Technische en Industrile Vormgeving as a spin-off of the faculty
for Architecture, and in cooperation with the faculty for Mechanical Engineering, among
others. Some years later the faculty became independent and the name was changed
into the faculty of Industrieel Ontwerpen or Industrial Design Engineering. As years
went by both faculties developed relatively independently which has had drawbacks
as well as benehts. Of course, more intensive cooperation would probably have led to
more efhciency and a better how of knowledge. On the other hand, some new helds of
knowledge develop themselves more easily in greenhouse-like organizations.
However, after nearly forty years the two organizations still have a lot in common with
the main communality being their focus and vision on society and the role for leading
edge design research. Perhaps more important than what they have in common with
each other, is the design research work which is complementary between the two
faculties. The research subjects within the portfolios of the two faculties differ as does
the approach of the design related research in general. Human factors, methodology
and sustainability are examples of research subjects for which the approach of the
two faculties differs signihcantly. !n these differences lies the greatest opportunity for
cooperation.
A team, consisting of the two deans and several professors of both faculties started
discussing the possibilities of cooperation, a discussion of which the results were
presented at a symposium in December 2005.
This publication is the result of the hrst joint conference held on the 6
th
of June 2008
with the title Design Processes. This title was selected by an organizing committee
consisting of Wim Poelman, David Keyson, Petra Badke Schaub, Teake the Jong and
Hannah Ottens. The committee was of the opinion that the most striking difference
between the disciplines was the attitude against and the practice of methodology in the
design process. It was decided that a preliminary investigation would be organized to
provide specialist with data from practice preparing their papers.
Preliminary Investigation
Four student assistants were invited to carry out the preliminary research, two from
each faculty. Names: Gijs Kappen, Melissa van ter Meij, Maarten Heijmerink and Matty
Cruijsberg.
Next six subjects were dehned for inviting specialist for preparing a paper. These
subjects were: design processes in general (invited specialist professor Henri Achten),
visualization as a design tool (invited specialist professor Petra Badke Schaub), project
management (invited specialist professor Joost Wamelink), social complexity in

6 Introduction - W.A. Poelman
collaboration (invited specialist professor Petra Badke Schaub), decision making (invited
specialist professor Peter Paul van Loon) and technology diffusion (invited specialist
professor Wim Poelman). A questionnaire was set up divided in chapters for each
subject. The specialists were requested to add there own points.
Eight projects were selected, four Industrial Design cases and four Architecture cases.
Interviews were arranged with the involved companies/designers/architects. The
interviews were carried out by two students, one of each faculty.
The interview reports were presented hrst to the interviewees for comments and then
passed to the specialists.
Papers prepared by the specialists were presented to peers, one of the University of
Twente (professor Arthur Eger) and one from the University of Eindhoven (professor
Jos Lichtenberg).
The chairman of the conference professor Teake de Jong of the faculty for Architecture
was asked to comment the overall results of the conference. His comments are recorded
in chapter Chairmens Impression.
The general impression is that specialists were not able to base their paper fully on
the result of the preliminary research. Two facts could be the reason. The hrst is that
many questions could not be answered objectively, so no analysis could be made.
The second is that a lot of interesting information came out of the interviews apart
from the questions that inspired the specialists to elaborate on that specihc issue from
their own point of view. One other aspect might have played a role. For the specialists
the conference was a great opportunity to present their own vision. The cases were
deployed rather for underpinning their own opinion than for analysis in order to come
to new insights.
One of the valuable results of the preliminary research turned out to be the propositions
for which the interviewers explicitly asked. They are presented in this introduction. In the
Chairmens Impressions chapter he will comment these pro-propositions extensively.
The cases
The cases provide several examples of the various characters of design processes.
Not all information, resulting from the preliminary research is free for publication, but
the propositions by the designersfarchitects and the specihc comments in the papers
provide valuable information.
The Westraven building by CePeZed is a project for the government organization
Rijkswaterstaat and based on existing building which is stripped completely until only
a concrete skeleton was left over. This skeleton formed the basis for a modern ofhce
building in which many new technologies were applied. Eye catching in the project are
three squares behind which hoors are broken away to create high open spaces and to
get rid of the boring repetition in the faade. Remarkable are furthermore the textile
screens in the faade which care for sun shading a well as for wind shielding and sound
decrease.
!nteresting are also the large spaces on the ground hoor in which an intermediate
climate is created. Parts of the walls are realized by ETFE inhatable pillows.

Figure 1: Fasade detail
Propositions:
Every advisor has solutions. o
The architect has to take all ideas to a higher level. o
The architect introduces problems, the advisor provides solutions. o
Copies are compliments. o
The A230 chair by Ahrend is a representative example of an
advanced industrial design engineering product. As e result
of much experience with the design of ofhce chairs the
Ahrend team is able to develop a product which is optimized
in every aspect such as ergonomics, form, produce ability,
sustainability, etcetera. Here comes to the fore an important
difference with architecture: development deepness. In
architecture development costs are mostly written of on
one product, while a chair is produced in ten thousands.
Figure 2: A230 chair
Proposition:
Clients have questions. o
Decision making mostly means: how large is the demand. o
The sales agency is our antenna. o
The purchasing agency is an interesting source. o
We write our program of requirements ourselves. o
!n an ofhce, meeting may be more important than work. o
Styling is 10% of our work. o
The image-and-sound (in Dutch, beeld en geluid) building by Neutelings-Riedijk is
a useful example how art and architecture can be integrated. The relation between
architecture and art seems to be quite different than the relation industrial design
engineering and art. As the artist houses more or less in very architect, most industrial
design engineers do not feel like an artist at all. The artistic industrial designer forms even
an apart group within the discipline organized in different professional organizations.
The chair of Ahrend will never be regarded as art, but the knotted chair of Gijs Wanders
is presented in famous art museums. Specihc about the !mage and sound building is

the cooperation with Jaap Drupsteen, a graphical and media designer. In addition to
technology borrowed from the held of graphic arts, quality standards were transferred
to the glass facade to realise this remarkable building.
Figure 3: The image-and-sound building
Propositions:
The scale of a project is not relevant for the way of communicating. o
Steps are similar to those taught at TUDelft + geographical centered o
communication.
Different mock ups to simulate different research questions. o
All knowledge in architecture is common knowledge. o
Also the BeerTender by MMID will never be regarded as art, but it is an excellent example
of industrial design engineering where the link to marketing is crucial. This project is
about a new way of packing, distributing and drinking beer for the home market.
Acceptance by the user of this concept is dependant of marketing communication but
to a large extent of design. The look of the business to business image of the beer
container would not work, not the ergonomics.
Figure 4: The beer container

Propositions:
Beertender is produced in very large series. o
My own style isnt important in this project. o
Style is work method f-d-p (Functionality & technology, Design (look & feel), o
Production & assembly)
I cannot recall decisions that explicitly. o
But there have been moments like that during the project. Time, Money and o
Quality.
The 1-2-3 House by Martini is an extremely interesting project in the context of the
relation between architecture and industrial design engineering. You could say that
an architectural product is developed and produced as an industrial designed product.
From the interview is learned that there are many constraints introducing this kind of
approach in housing industry. Up scaling is necessary to earn back money invested in
the manufacturing process, but the market structure is not suitable to apply marketing
strategies from industry. The housing market is highly bureaucratic.
Figure 5: Turning the tunnel
The Carver of Spark Design & Engineering and carver Europe is based upon the
invention of a hydraulic canting mechanism, which enables stability of narrow vehicles.
The application of the system leads to both a striking driving experience and a striking
visual appearance. In fact, a new archetype of a vehicle is created which resembles a
cross between a motorcycle and a small car. The success of the design is a result of
the collaboration between the engineering company (Carver Europe) and the design
company (Spark Engineering). The design problem is comparable with that of the
Beertender, introducing new product concepts linked to new human behaviour and new
visual appearance. The difference is that Carver does not have a marketing power like
the beer companies. Introduction by immense marketing campaigns is not possible,
so Carver is dependent on a slow introduction via innovators, trendsetters and trend
followers.

Figure 6: The Carver
Propositions:
Robert Barnhorn, Spark:
We see that most women chose the managing side of this profession. o
Investors knew that extra time would be a good investment to there product. o
The one who pays makes the last decision. o
Architecture knows heroes, industrial design the name of the bureau. o
Frank Vermeulens Carver:
A car consists over more than 1200 components. o
Small steps have to restrict high risks. o
Architecture is a specihed direction in product designing. o
Media like to attach a name of an architect to a building. o
A mass product has a lifecycle of one year, but a building has a lifecycle of o
50-100 years.
The Industrial Design Engineering Building, designed by Fons Verheyen The building in
which this conference is taking place is an example of a project in which cooperation
between architects and industrial designers might be expected. Like the CePeZed
building, this building is based on an existing building being the central workshops of
the TU Delft and some ofhce buildings.
This kind of design, which is especially important in architecture, could be specihed
as Supply Driven Design (SDD), which proceeds from existing artefacts. Although we
cannot go into depth about this relatively new, sustainable type of design activity, we
can conclude that more creativity is needed to design something within the limitations of
an existing artefact than is needed to design something completely new. In this regard,
industrial designers could learn from architects, who do this on a regular basis.
Propositions:
Small series, big scale difference. o
The hrst big decision was to decide to do such a big renovation project, then o
deciding upon the hnal amount of square meters and where to place which
function.
The whole idea, to create one big space in which everybody would be able to o
enjoy what others are doing, was one big risk.


Figure 7: Interior sketch
The inhatable care bed of !ndes is the last project to discuss. The bed was not designed
as a synchronous product but as a diachronic script, in which not a special delivery
service, but the homecare nurse herself delivers and installs the bed. The physical
product was simply a way to enable that script. Because traditional care beds did not
ht into that script (too heavy, too big), it was necessary to design a completely new
product.
Script based design represents a growing trend in the discipline of industrial design
engineering and hts in the subtitle of this book life is a theatre". Starting point is: hrst
write the script and then the products necessary to realise the script. In architecture
this might be more common. The use of a building should be described before it is
possible to design a proper building.
Figure 8: !nhatable care bed
Propositions:
Not much attention was given to aesthetics. o
Users played an important role, from the start they were consulted and later o
they were involved when prototypes had to be tested.
The people involved in the engineering phase are already looking over the o
shoulder during the concept development stage.

From the short description of these cases it will be clear that the diversity is so large
and the n" that small that quantitative conclusions are impossible to derive, but also
qualitative conclusions are not easy to formulate because of the diversity of contexts.
Nevertheless, a lot is learned from the cases in combination with the analysis and
rehections of the specialists. Rehections of the chairman professor Taeke de Jong will
follow in Chapter Chainmans impressions.
The subtitle
The subtitle of the conference behind this book reads: life is a theater. Architects care
for the scenery; Industrial designers care for the props; People care for the drama.
This subtitle was chosen in the hrst place to express the link between the faculties of
architecture and industrial design engineering. However, the message goes further than
that. Most people will agree with the proposition that architects and industrial design
engineers should not write the script for human existence. The function of scenery and
props designers is to serve the scriptwriter and the actors with objects supporting the
play. Imagine a situation in which the behavior of a performer has to change because
of the scenery or props. For example, when the actor has to appear on the scene from
the ceiling, or is only able to speak after putting of a mask, without discussing it before
with the scriptwriter and actors, this would lead to an unacceptable situation.
But in real life, this happens all the time. Human behavior is for a large part enshrined
by architects and designers and not anymore by people themselves and spiritual
fathers who acted as scriptwriters for life and still do in religious communities like the
Nuslim community, where hve times praying a day with the face to Nekka is part of
the script.
Nowadays, the script of life is for a large part written by architects and designers. Urban
planning decides how we spread our activities geographical. The design of modern
residential districts determines for a large part how we communicate with each other.
The design of shopping centers determines how we acquire our foodstuffs. Designers
of means for transport decide how we move ourselves and kitchen designers decide
how we cook.
All this has to do with the mechanisms of technology diffusion on which Wim Poelman
will elaborate in his paper later.
The main subject of this conference however is Design Processes and the main issues
of the conference were:
the contemporary interrelationship of Industrial Design and Architecture o
a confrontation of contemporary design practice in both domains with academic o
theory and education

13

IDE+A
Design Processes

15
2 Design processes between academic and practice views
Dr. ir. H.H. Achten
Assistant Professor, Architectural Modeling
Eindhoven University of Technology
Faculty of Architecture, Building and Planning
Design Systems Group
Abstract
In order to speak about the commonalities and differences between industrial and
architectural design, we need a common framework that may capture both design helds.
Practice-based descriptions have a long tradition, and are close to everyday reality of
the helds, but they are too specihc. The scientihc study of design - design research -
is a more recent development, which aims accurately to provide this framework. We
discuss the current understanding of design, its limitations, and some observations
related to the cases of the IDE+A Conference.
Keywords: design theory, design method, design research.
1 Do we understand design processes?
Before we begin the general argument in this paper, we must consider an important
premise that underlies the motivation of the text. At the IDE+A Conference, architects
and industrial designers are invited to address specihc problems in design. The issue
then is this: can an architect or industrial designer discuss aspects of design in his or
her held in such a way that it makes sense to the other? The architect's concerns, after
all, are about bricks, steel, glass, and wood; how to organise the spatial composition
of a building or urban environment, how to make structures and installations work
together, etc. The industrial designers concerns are about plastics, textiles, and various
kinds of metals; how to create effective and ergonomic solutions for people; how to set
up production in the most efhcient and cost-effective way, etc. There is a gap therefore
between the design helds in terms of domain knowledge: the facts and principles that
are specihc to architecture and to industrial design, respectively.
Both architects and product designers (or designers from any other discipline, for that
matter) mainly work on a project basis - meaning that a project is acquired, a single
designer is assigned or a team is put together, and work on the project continues until
its completion (or until its early cancellation). Such projects tend to take a long time,
varying from a few months to several years. Throughout this time projects are subject
to all kinds of change: in the team, in the norms and laws to which the design must
comply, in the user's taste, in materials and production techniques, and so on. Thus,
each project has its own confusing history of contingencies which must be solved for
the project to be completed successfully.
There is a twofold assumption, therefore, when we talk about design processes: that
we can bridge the differences between the design domains, and that we can abstract
enough from everyday practice within each design domain to talk about the general
aspects of design. If either of these assumptions fails (or we choose not to believe in
them) then there is no basis for comparison other than the anecdotal level. Believing

Design processes - H.H. Achten 16
in these assumptions, however, does not mean that all our problems are easily solved.
Design processes have developed over a very long period of time (one could even claim
thousands of years). There is a very close connection between the praxis of design,
its body of knowledge, and design methods. For practitioners it is often very hard to
separate these views. The conception of the design process as something that can be
discussed autonomously is very much a modern methodological-scientihc idea, and a
comparatively young one, having gained currency about hfty years ago. Even though
tremendous progress has been made in the understanding of design, there is still a lot
left to be understood properly.
The perspective that we take in this text, therefore, is academic rather than practice-
based, since the academic view provides a transferable set of theoretical concepts by
which we can discuss design in various domains. !n the hrst section, we will outline
the scientihc concepts dealing with design processes - both in the areas of theory and
methods and sketch the current orthodox view of what design processes are. This
view is certainly not unchallenged, and a number of the most notable problems will be
described. To conclude, we will briehy review the design cases presented for the !DE+A
Conference.
Since the notions established in this paper are the result of research on design in all
kinds of domains, here we refrain from talking about architects or industrial designers,
but use the more generic term designer.
2 Design process, theory and method
In the description of the design process, two perspectives can be utilised: that of design
theory and of design method. Each has a very distinct view of design processes, but it
is fair to claim that there is a very strong interdependency between the two.
Cross (198+) dehnes design methodology as `the study of the principles, practices and
procedures of design in a rather broad and general sense. Its central concern is how
designing both is and might be conducted. This concern therefore includes the study
of how designers work and think; the establishment of appropriate structures for the
design process; the development and application of new design methods, techniques,
and procedures; and rehection on the nature and extent of design knowledge and its
application to design problems.
Cross' distinction between how designing is and might be conducted dehnes the
difference between design processes (how designing is) and design methods (how
designing might be conducted). In order to describe these aspects, it is necessary to
have a theoretical framework for design this is design theory.
It is important to notice that designers and researchers, when talking about design
theory, often mean different things. Professional design theory has been around at least
since Vitruvius (approximately 1st Century BC; see Vitruvius 1960). Professional design
theory is instrumental theory in the sense that very often it instructs or describes how
to get things done. Its main subject is the motivation and starting points for design,
and the systematic description of styles. !t tends to be rehective, based on personal
experience, and is very much object-oriented urban environments, buildings, details,
and so on. Professional design theory, however, is not the view that we take when we
talk about design theory.

17 Design processes - H.H. Achten
2.1 The role of design theory
Theoretical rehection on design is necessary to transfer knowledge about the domain to
others, for example in an educational setting. Theory helps to distinguish between what
is fundamental to the discipline and what is not; which aspects and concepts matter to
design, and which aspects and concepts are incidental. This helps the designer maintain
an overview of the discipline and guards against ad-hoc actions. A strong theoretical
basis can increase efhciency because the designer has a clear view of what the goals
are, and understands the means by which to achieve them. A too-rigid understanding
of design, however, may prove to be stihing; it is important therefore to achieve a good
balance.
In more recent applications, design theory has also been instrumental in the development
of new tools for design in particular in the development and application of computer
tools. !n recent years, the held of architectural design has experimented a great deal
with these technologies. The most notable inhuence comes from the !nternet, which
enables new group processes such as collaborative design and twenty-four-hour design
teams. Also the more direct use of the form and shape generating capacity of computers
in design is inhuencing the design process.
Design theory, to conclude, aims to answer the following questions about designing:
1. What is designing? !s the activity called designing different from other
human activities (for example, cooking, sport, arguing, etc.) If so,
what are the differences, and what do the differences imply?
2. What is a design problem? Are design problems different from other
kinds of problems, and if so, what are their characteristics?
3. What is the design process? What does the design process look like,
what are its general principles, how is it conducted?
+. Who is the designer? How is a designer different from other
professionals? What skills do he or she apply in the job?
S. What is a design solution? What are designs, how are they structured,
what role do they play in the creation of things? How are designs
different from products created in other helds (such as art, industry,
or farming)?
6. What is the role of design theory? What are the various approaches to
the study of design? !s design theory different from other areas of
enquiry, such as history, physics, or law? How does theory inhuence
designers?
2.2 The role of design methods
Design methods concern the actual or desired order of the design decisions that are
taken in a design process. The everyday use of the term `design method' is quite
informal and can mean anything from a habitual working method to highly structured
and controlled processes. Another recurring notion is the personal design method,
which is not communicated with others it is even claimed to be incommunicable. For
a better understanding (and appreciation) of design methods, however, we must clearly
dehne exactly what a design method is. !n our view, something is a design method, if
and only if:
1. !t clearly dehnes a goal in the design process.
2. !t identihes steps to take and the order in which to take them.
3. It is applicable to more than one case.

4. Other people can also apply it.
5. It has criteria to determine when a step has been concluded.
Each aspect of this list has to be present in order for something to be called a design
method.
There are a number of reasons to develop and use design methods. Design methods are
helpful in cases when not using a method means it will take too long to hnd a solution,
or when the cost of not hnding a good or `satishcing' solution is high. This typically
occurs in complex design projects, or when the design(er) (team) takes on a problem
with which it does not have much previous experience. Also, in projects that require
more than an average justihcation of steps or solutions, design methods can be used to
structure the process. Finally, because of the explicitness of design methods, they also
help in coordinating large design teams or multiple experts involved in projects.
There is a sometimes tenuous relationship between design methods and practice. Most
of the designers of the IDE+A Conference cases, when asked whether they followed
a method, replied either that they did not, or that when they did, it closely followed
what they were taught at university. They also noted that practice will most often
lead away from the ideal process, so there is a perceived lack of applicability. When
confronted with new or changed design methods, designers often feel restricted in their
freedom (this is probably a stronger sentiment in architecture than in industrial design).
This should not be very surprising - getting to understand a new method requires
time and effort, which distracts from the job at hand. This is a situation that a skilled
designer wants to avoid. This mechanism can also explain why designers often dislike
talking about their method. Thinking about the design process in terms of method is
a rationalising activity. Design problems, however, as we will see in the next section,
cannot be fully understood in rational terms. Consequently, at some point a designer
has to state where things are explicitly explainable and where they are not. This again
may cause uncertainty or confer a sense of uneasiness. The mark of a skilled designer
is that he or she has internalised design skills so that they do not require explicit mental
effort. Conscious thinking about the act of designing disrupts this because it challenges
the hidden skills to become expressed. Again, this is experienced as an intrusive activity.
Finally, in the domain of architecture in particular there is a heightened status for star
designers. Connected with this status is a tendency to keep the processes or methods
shrouded as some kind of mystery or art.
Most design methods have been developed for single designers. In some cases,
design teams are considered to be one designer consisting of multiple persons. This
may perhaps work for very well-contained design methods that have a limited scope
(brainstorming, creativity techniques, or problem decomposition) but it breaks down
at higher level goals because of group dynamics and mixed expertise. As much of
everyday design takes place in teams or in communication structures with outside
parties, this is a real dehciency in the area of design methods.
To conclude, if we want to describe design processes, we need a theoretical framework
for design. It is basically a descriptive activity with design(ing) as its subject. Based on
theoretical considerations, a design theory may lead to a design method, but this is
not necessarily so. Design methods, on the other hand, may be the subject of design
theory. Design methods are prescriptive and solution-oriented. A design method always
implies theoretical principles because it identihes important steps and issues in the
design process.

3 The orthodox view of design processes
The relatively young scientihc study of design processes has gone through a number
of distinct periods (see Cross (1984) and Jones (1980) for good accounts of this
development). Three research approaches have emerged as dominant in the current
view of design processes: rational problem solving, about the structuring of design
problems; information processing, about the thought processes of designers; and
protocol analysis, about the research methods to study designers. Obviously, there are
many other ways to research and investigate design (see for example Oxman et al.
(1995), Achten et al. (2001), and Achten et al. (2005) for an overview), but the three
mentioned above constitute what we might call the orthodox view of design and the
study of design.
3.1 The nature of design problems
In the theoretical research on design, a distinction is commonly made between four
classes of problems with an increasing degree of complexity and unpredictability: tame
problems, well-structured problems, ill-structured problems, and wicked problems
(Lawson (1990), Simon (1973)). The general consensus is that design problems are
wicked problems. The characteristics of wicked problems have been dehned by Rittel
and Webber (1973):
1. There is no dehnitive formulation of a wicked problem.
2. Wicked problems have no stopping rule.
3. Solutions to wicked problems are not true-or-false, but good-or-bad.
4. There is no immediate and no ultimate test of a solution to a wicked
problem.
5. Every solution to a wicked problem is a one-shot operation; because
there is no opportunity to learn by trial-and-error, every attempt
counts signihcantly.
6. Wicked problems do not have an enumerable (or an exhaustively
describable) set of potential solutions, nor is there a well-described
set of permissible operations that may be incorporated in the plan.
7. Every wicked problem is essentially unique.
8. Every wicked problem can be considered to be a symptom of another
problem.
9. The existence of a discrepancy representing a wicked problem can be
explained in numerous ways. The choice of explanation determines
the nature of the problems resolution.
10. The planner has no right to be wrong.
The consequence of the fact that design problems are wicked problems is that a limited
degree of rationality can be applied to solve them. Creating a solution will always
depend to some degree on a creative insight. The phase where solutions are created
is the challenging part where a designer seemingly jumps from a problem setting
to a solution. A match or mapping is made between two distinct things a problem
and a solution. This is not trivial: just why exactly a given solution matches a problem
is still unanswered. Both problems and solutions are complex and they have almost
no common elements in their structure. In most cases, problems and solutions are
even described in different ways: problems as sets of requirements, wishes, demands,

or other verbal statements; and solutions as conglomerations of ordered elements of
urban/city environments, buildings, or objects.
3.2 Structure of the design process
Given the characteristics of design problems, it follows that creating a solution is not
a one-step affair, nor a matter of applying one technique to solve them. Designing,
therefore, is a lengthy process in time, during which the designer iterates and revises
the design many times. Designing is as much about understanding the problem as it
is about creating a solution, in particular in the early phase of design. Therefore, not
only does the designer utilise information and knowledge that is provided at the outset
(brief, site, client, etc.) but he or she also generates a lot of knowledge throughout the
design process.
Roozenburg and Eekels (199S) dehne the Basic Design Cycle (BDC) as a collection of
activities and documents that are created and performed in design. The BDC consists
of the following (terms in italics denote activities):
1. Function statement: a statement about what is needed in the design
problem.
2. Analysis: analysis of the function statement or current state of the
design.
3. Criteria: a set of criteria to which the design has to conform.
4. Synthesis: the creation of a (preliminary) design or solution to a sub-
problem.
5. Provisional design: the external representation, by means of sketch,
drawing, text, or model, of the (preliminary) design.
6. Simulation: the derivation of the expected behaviour or performance
of the (preliminary) design.
7. Expected properties: a prediction of the future behaviour or
performance of the (preliminary) design.
8. Evaluation: a judgement of how well the (preliminary) design
performs, based on the criteria formulated earlier, and the expected
properties.
9. Value of the design: a value setting of the performance, based on the
evaluation and goals set by the designer.
10. Decision: the decision to continue with the design (either through the
creation of a new proposal in Synthesis, or restating the problem
in Analysis), or deciding that the design is hnished, which leads to the
next document:
11. Approved design: the hnalised design.
Roozenburg and Eekels note that the actual order of activities and documents in a
concrete design project is unpredictable, so they do not claim that this order is indicative
for a design project. Rather, they claim that in any given design project, each activity
and each document has to be performed or created at least once, but most likely many
times over.
The BDC may be considered to be the private design cycle for a designer or design
team. Throughout the whole design process, additional structuring is created as well
in architecture this is usually a phased process consisting of sketch design, preliminary
design, hnal design, and execution design. Each phase is concluded with documents
that describe the design solution with increasing precision. The purpose of the phased

structure is to create secure, consistent descriptions of the design which can form
the basis for the next steps in the design process. In that way, the designer avoids
unnecessary backtracking.
3.3 Forms of knowledge in design
Designing is knowledge intensive. Much of design is a matter of applying knowledge of
previous solutions that inform the basic direction in which the current design solution
has to move. Previous solutions can be referred to as precedents (prominent examples),
types (generalised knowledge of classes of buildings or products), and analogies (used
as metaphors rather than literal examples).
The design process itself starts out with many facts, arising from the brief and from
clients desires, from the site where a project is to be realised, from particular technologies
that will be used (for example the 123 House case in the IDE+A Conference), budget,
and so on. Throughout the design process, additional knowledge is generated about
the design itself, and the designer searches also for information based on the needs at
that point in the design.
Constraints are a specihc type of knowledge and information that is used in design.
Constraints put limits or boundaries on the design or the context of design. Client goals,
norms and laws, local regulations, welfare, and so on have to be met in order for a
design to be approved.
3.4 Forms of reasoning in the design process
In order to create (preliminary) design solutions, knowledge and information must be
processed. This involves several forms of reasoning. Reasoning by example is a major
technique used by designers. Whether the previous solution is a precedent, type, or
analogy, the designer takes some element of the example and, based on the perceived
structure of the solution, generates a new solution that is suited to the current design
problem.
A way of reasoning in design that is a bit more explorative or imaginative is through
what-if reasoning or by means of scenarios. In these cases, the designer takes the
current design and tries to imagine how it will perform. In this way, designers can also
use previously experienced episodes with other buildings or urban environments and
aim to duplicate them in the current design.
Given the characteristics of wicked problems, it is not possible to determine objectively
whether a design completely fulhls all the requirements it has to meet. Obviously,
designers try to meet the constraints set out in the brief, and those that are imposed by
the context of the project. However, this does not mean they have to prove that their
design is perfect or the only one possible. Rather, designers try to meet the constraints
as much as possible, and aim to reach at least a minimum threshold of performance or
quality in design. Simon (1973) calls this `satishcing'.
Analytical modes of reasoning are used particularly in the analysis phase of a project, or
when the consequences of design decisions have to be derived. This is also where logic
plays a role in the design process. Finally, the least well understood form of reasoning
is what is generally called visual reasoning. Designers use external representations
such as drawings and sketches a lot, and a considerable amount of generation and

judgement is done visually on the basis of such sketches and drawings. All designers of
the IDE+A cases strongly indicate that they consider sketching to be a vital skill.
3.5 Psychological view of designers
The reasoning and memory abilities of people is limited. Memory is generally conceived
of as consisting of two main functional parts: long-term memory (LTM) and short-term
memory (STM) see Akin (1986) for a good introduction. LTM is where experiences
are stored at length. The recall from LTN is relatively slow, but most signihcantly, it is
not directly accessible for conscious processing. In STM memories are accessed from
LTM and once there can become the subject of thought processes. STM works relatively
fast, but it has a limited capacity to hold information. In general, this is thought of as
roughly seven coherent pieces of information, called chunks. How big the chunks can
be, or how they are organised, remains unclear. It seems evident, however, that more
experienced or skilled designers utilise better or more compressed pieces of information
when they are reasoning.
3.6 External representations in the design process
Limited reasoning and memory capacity is an additional factor that structures design
processes. One role of representations such as drawings and models is to form an
external memory which can store information about the design by similarity. The
designer needs only to glance at the sketch or model to quickly activate the implicitly
stored information.
External representations, in particular those that complete a phase of the design process
(sketch design, preliminary design, hnal design, and production design), also have a
legal status, and they are also used to communicate between parties in the design
process. A large part of the activity in the design process, therefore, is reserved for the
production of accurate and precise drawings and documents.
3.7 Creativity in design processes
Creativity plays an important role in design it is the mechanism with which a designer
is able to come up with a novel solution to a problem. Creativity does not work in
isolation; it needs to be embedded in a work context that provides information and the
right setting to generate an idea.
A common distinction which is made in terms of design solutions are the following three
classes of designs (Brown and Chandrasekaran, 1985):
1. Routine design: the creation of a solution that falls completely within
the range of previous solutions. The solution is adapted to current
needs but does not introduce anything novel. Redesign may also be
considered to be routine design.
2. Innovative design: the creation of a solution which has at least one
additional feature that has not been seen before in this kind of design
solution. Most of the design conforms to existing examples, but one
part is pushing the limits. All the architecture design cases in the
IDE+A Conference demonstrate this kind of design.
3. Creative design: the creation of a solution that has a highly different
structure compared to existing solutions. A creative design does not
have a lot of similarities with existing designs.

The distinction comes from the domain of artihcial intelligence. From that perspective,
the delineation between the classes is fairly straightforward. A routine design simply is
an instance of an already known type or class; an innovative design adds something
new but does not change the structure of the type or class; and in creative design an
altogether new structure for a type or class is created.
The delineation becomes less clear, however, when we try to apply it from a designers
perspective. In particular the distinction between innovative and creative design
becomes hard to make. Especially if we insist on completely new structures, then
most of architectural design simply is not creative a conclusion with which many will
disagree. The difference in innovative and creative, therefore, is more a matter of
the degree to which a design is pushing existing limits by means of innovations.
3.8 Design, designers, the design process
Based on the above, we can now summarise the orthodox view as follows. The
designer can be conceived of as an information processor (STM, LTM, and cognitive
structures) who tries to solve wicked problems. An important design activity is the
subdivision and reformulation of the wicked problem into sub-problems in order to
make them well-structured. The designer has procedural knowledge in the form of
heuristics and design methods. Designing requires declarative knowledge of the domain
(architecture, industrial design, machine engineering, etc.) as well as knowledge of
previous solutions (cases, precedents, and types).
Because of the limitations of STM and LTM, the designer cannot have an overview of
the whole problem (even not when a problem is well-structured, which in design does
not really occur). Consequently, the design process is sequential in time and iterative.
External representations such as drawings and models help to maintain an overview
and understand the consequences of design decisions.
Through the use of phases the designer prevents the possibility that, late in the
process, a small change will necessitate a redesign of the whole project (this does
not always work). Throughout the design process the designer explores both the
solution and the problem. One might claim that only at the end of the design process
is the design problem understood. A design problem does not have one single correct
solution. Furthermore, it is not possible to determine the degree of correctness. The
designer therefore strives for `satishcing' rather than perfect solutions.
4 Challenges to the orthodox view of design processes
The view of design processes sketched above is rather concise, but in broad outlines
provides the contours of our current understanding of design processes. As can be
seen, there is a strong interdependency between theoretical and methodological
notions. Despite the relatively short period of time that design has been an area for
scientihc research, the account seems to be surprisingly coherent. Obviously, this view
is not the ultimate description of what design is about. Many things are still unknown
and there are many challenges to the orthodox view of design processes.
The foundation of the orthodox view of design processes is rational problem solving
(RPS), as dehned by Simon (1996). Simon's work was seminal in setting up a general

framework to talk about design as a scientihc subject at all. !n recent years, a different
view of design has been put forward by Schn (1991), called rehective practice (RP).
Table 1 below sets out the differences between the two approaches.
5DWLRQDO3UREOHP6ROYLQJ 5HIOHFWLYH3UDFWLFH
Design is a form of problem solving . Design is a reflective dialogue between
the designer and the design.
$ design can be decomposed into sub-
problems , each of which can be
solved, and the solutions to each sub-
problem can later be integrated into
an overall solution.
$ designer iteratively names the most
important issues at the start of an
activity, frames these issues in a
context, moves by creating a solution,
and evaluates the outcome with respect
to the frame
The quality of a solution is
measurable.
$ designer reflects in action, about
action, and about work.
Table 1: Rational Problem Solving versus Rehective Practice
Since rational problem solving has the longer research tradition, it is clear that its
theoretical foundations are quite strong. A problem with RPS however is that designers
(in particular architects) do not recognise themselves in the RPS framework. Rehective
practice, as noted by several researchers (see Dorst (1997), Valkenburg (2000), Reymen
(2001)), has a weak theoretical foundation, but strongly appeals to designers. Unless
a designer has a very systematic approach to design, the naming-framing-moving-
evaluating cycle seems much closer to what designers do. In earlier work (see Achten
(2003)), where we investigated the normative stance of three well-known architects
through their published works (Peter Eisenman, UN Studio, and Greg Lynn) in order to
derive their design methods, we have found some evidence for this. This concerns in
particular the decomposition of the problem, which resembles naming-framing more
than decomposition. This is so because there is a strong focus on concept formation.
An additional aspect that RPS ignores is the social aspects of design. Designers do not
operate in isolation, and most of the time they work in teams. The social aspects of
group dynamics such as leadership, dominance, negotiation, and team building are not
dealt with (see for example Foley and Macmillan (2005), Valkenburg (2000), Baird et al.
(2000), Ball and Ormerod (2000)).
Lastly, the idea that the motivation for design, or particular design decisions, is not
purely rational or can be stated completely objectively is a problem. Part of the way
designers in teams persuade each other is by means of storytelling. Another way to
investigate verbal exchanges in design teams is to look at convergence in the use of
words, to see whether a more or less consistent group dynamic is developing (Lloyd
(2000), Turner and Turner (2003), Dong (2005)).
Although RPS pays due attention to the psychological structure of designers, there is
no real differentiation between possible types of designers. In recent work, Lawson and
Dorst (2005) have investigated the notion of the level of expertise at which designers
may be classihed. They distinguish between seven levels: nave, novice, advanced
beginner, competent, expert, master, and visionary. Different cognitive structures, sets
of competences, and ways of organising the design process are associated with each.
Most of the work summarised here (except for Schns work) has begun in the past
decade and is still in development. This is only a brief sketch of additional or alternative

takes on the study of design processes: the held is very rich and diverse and cannot be
done justice in this section alone.
5 The IDE+A design cases
The IDE+A design cases include four from architecture, and four from industrial
design. From the description of each, it is clear that the complexity of the design team
plays an important role in the design process. Given the above outline of the current
understanding of design processes, we can immediately see that this aspect is found
wanting, as team design is not covered much by current research. Nevertheless, we can
make a number of observations about the cases.
1. Most of the designers who were interviewed were able to identify the
authorship of the key ideas in a project without a problem. One might
expect that due to the size of teams and the complexity of the task,
this may be more problematic.
2. In the architectural cases, innovation is much more focused on a
single aspect whereas in the industrial design cases, innovation is
often spread out over a number of key components.
3. !n architectural design, project acquisition through winning a contest
is a common phenomenon. However, this also means that the design
process structure is different from the classical client-meets-architect
model. The competition design leads to a proposal by which the
architect hopes to win the competition, but it is not the same as the
hnished design for the real construction work. Since it is a
competition, the ofhce will also not invest too much effort as there is
a real risk of not getting the job. So in this type of process, there
is a two-stage design process (competition and hnal design), with
different concerns.
4. Because of their length, the structuring of the design process in the
cases is based on the main documents or phases rather than the more
detailed design process for the single designer. The ideal design
process is seen as a point of reference, rather than an attainable
goal.
5. Practice is very demanding and problem-oriented. This means that if
something does not yield immediate results, designers are not eager
to work with it. !n this respect we see that scientihc research often
fails to provide productive frameworks for designers. Findings are
difhcult to translate to a practical setting, which creates a considerable
threshold for their application.
The scientihc study of design has provided us with a framework that allows us to
discuss design in various helds, such as architecture and industrial design, but also
engineering, chemistry, information technology, and so forth. Since design theory in this
sense is quite abstract and distant from practice, in order to gain a good understanding
what design is, it is necessary to reference to practice as much as possible.
6 Conclusion
Do we understand design processes? The modern methodological-scientihc view on
design resulting from some hfty years of research has given us tremendous insights

in the nature of design, designers, and design products. It has also revealed however,
that there is a lot more left to be understood than we currently know. Partly this will
always be the case: the study what design is, will never yield what it is to be a designer.
A methodological rehection on the design process does not give the immediate solution
for the design problem at hand, but it helps in creating the basic skills for the designer.
Finally, from the view of professional and academic responsibility, we need to understand
what we are doing in a systematic, objective, and rigorous way in order to engage in
the creative, unexpected, and joyful way of designing.
References
Achten, H.H. (2003), New Design Methods for Computer Aided Architectural
Design Methodology Teaching; International Journal of Architectural
Computing 1(1), pp. 72-91.
Achten, H.H., Dorst, K., Stappers, P.J. and de Vries, B. (2005), Design
Research in the Netherlands 2005 Proceedings of the Symposium
held on 19-20 May 2005 Eindhoven University of Technology;
Eindhoven: Faculty of Architecture, Building and Planning.
Achten, H.H., Hennessey, J. and de Vries, B. (2001), Design Research in the
Netherlands 2000, Eindhoven, Faculty of Architecture, Building and
Planning.
Akin, O. (1986), Psychology of Architectural Design, London, Pion.
Baird, F., Moore, C.J. and Jagodzinski, A.P. (2000), An Ethnographic Study of
Engineering Design Teams at Rolls-Royce Aerospace; Design Studies
21(4), pp. 333-355.
Ball, L.J. and Ormerod, Th.C. (2000), Applying Ethnography in the Analysis
and Support of Expertise in Engineering Design; Design Studies
21(4), pp.403-421.
Brown, D. C. and Chandrasekaran, B. (1985), Expert Systems for a Class of
Mechanical Design Activity, in Knowledge Engineering in Computer-
Aided Design, ed. by Gero, J.S., Amsterdam, North-Holland, pp.
259-282.
Cross, N. (1984), Developments in Design Methodology; Chichester, Wiley.
Dong, A. (2005), The Latent Semantic Approach to Studying Design Team
Communication; Design Studies 26(5), pp. 445-461.
Dorst, C.H. (1997), Describing Design: A Comparison of Paradigms; PhD
thesis, Delft: Delft University of Technology.
Foley, J. and Macmillan, S. (2005), Patterns of Interaction in Construction
Team Meetings; CoDesign 1(1), pp. 19-37.
Jones, J.C. (1980), Design Methods: Seeds of Human Futures; London: Wiley
Interscience.
Lawson, B. (1990), `How Designers Think: The Design Process Demystihed';
London: Butterworth Architecture.
Lawson, B.R. and Dorst, K. (200S), Àcquiring Design Expertise', !n
Computational and Cognitive Models of Creative Design VI., ed. by Gero, J.S.
and Maher, M.L., Key Centre University of Sydney, Sydney, pp. 211-230.
Lloyd, P. (2000), Storytelling and the Development of Discourse in the
Engineering Design Process; Design Studies 21(4), pp. 357-373.
Oxman, R.M., Bax, M.F.Th. and Achten, H.H. (1995), Design Research in the
Netherlands: A Symposium Convened By the Design Methods Group
Information Technology for Architecture, January 1995; Eindhoven,
Faculty of Architecture, Building and Planning.

Reymen, I. (2001), Improving Design Processes Through Structured
Rehection - A Domain-!ndependent Approach'; PhD thesis,
Eindhoven: Institute for Programming Research and Algorithms.
Rittel, W.J. and Webber, M.M. (1973), Planning Problems are Wicked
Problems, In Cross, N. (1984), Developments in Design
Methodology; Chichester, Wiley, pp. 135-144.
Roozenburg, N. and Eekels, J. (1995), Product Design: Fundamentals and
Methods, Chichester, Wiley.
Schn, D.A. (1991), `The Rehective Practitioner - How Professionals Think in
Action; London, Basic Books.
Simon, H. (1973). The Structure of Ill-Structured Problems, In Cross, N.
(1984), Developments in Design Methodology; Chichester, Wiley, pp.
145-166.
Simon, H. (1996), `The Sciences of the Artihcial'; Cambridge, Nassachusetts,
MIT Press. First Edition 1969.
Turner, S. and Turner, P. (2003), Telling Tales: Understanding the Role of
Narrative in the Design of Taxonomic Software; Design Studies 23(6),
pp. 537-547.
valkenburg, R. (2000), `The Rehective Practice in Product Design Teams', PhD
thesis, Delft, Industrial Design Engineering.
Vitruvius (1960), Vitruvius: The Ten Books on Architecture, translated by
Morris Hickey Morgan, New York, Dover.

IDE+A
Visualization

29
3 Sketching is Alive and Well in this Digital Age
Prof. G. Goldschmidt
Professor, The Mary Hill Swope Chair in Architecture & Town Planning
Faculty of Architecture and Town Planning
Technion Israel Institute of Technology
Abstract
The different modes of visualisation found in the Delft Interviews are explored with
respect to their particular advantages at specihc phases of the design process, and as a
means of communicating with various project stakeholders. Two main conclusions arise
from this exploration. First, that with few exceptions there are no signihcant differences
between architects and industrial designers in the way they produce and use visuals.
Second, despite the proliferation of potent digital visualisation means and their willing
adaptation by design practitioners, freehand sketching continues to be practised by
almost all designers throughout the design process. The extraordinary cognitive advan-
tages of sketching are outlined and it is argued that because of those advantages sket-
ching will continue to reign in design until other means of visualisation will be capable
of emulating its supremacy.
Keywords: DI (Delft Interviews); digital; design; model; sketch; visualisation
Introduction
In a world such as the one we live in it is only natural for young students, who were
born into the digital age, to ask their designer-interviewees: In this digital age is the-
re still a use for sketching?' From the way the question is phrased it appears that the
expected answer is no, but the courteous students allow the designers, practicing
architects and industrial designers, a less-decisive reply by adding a follow-up question:
`. and in what phases would that be?'
Sketching is a mode of visualisation, alongside other modes. All designers in the survey
talk about means of visualisation they used in the particular project on which the inter-
view focuses but they all generalise to other cases as well. Visualisation, in the evidence
provided by the interviews, serves a number of important purposes; hrst and foremost,
as communication in its roles of information and image recording and description, de-
monstration and sharing, explanation and convincing. Apart from freehand sketches
(including annotations), visuals include primarily other manual drawings on paper, di-
gital two- and three-dimensional drawings, and physical models. Digital drawings can
be divided into two distinct types: precise measured drawings, and three-dimensional
images and renderings. Sometimes animation and movies are also added to the arsenal
of visuals. When and for what purpose is each of these modes of visualisation used,
and why? These are the questions we are out to explore in this paper, with an empha-
sis on sketches, which are the most frequently mentioned visualisation modality in the
Delft Interviews (DI), and which, perhaps surprisingly for the interviewers, are still in
frequent use in both architectural and industrial design practice.

30 Visualization - G. Goldschmidt
1 Why visualise?
The question `why visualise' is almost rhetorical, as we all grew up to believe that à
picture is worth a thousand words'. The design of physical artefacts requires that desi-
gners and those for whom the artefacts are designed consider many elements and their
properties, as well as the relationships between them (and in the case of architecture,
also between them and their surroundings). Function and form must be understood,
evaluated and optimised for a successful result. An adequate representation of these
complex parameters is not possible without visualisation, especially the representation
of shapes and forms. It is possible for individuals to entertain internal representation
using mental imagery, and there are reports that designers can go quite some way
using mental imagery only (Athavankar 1997, Athavankar & Mukherjee 2003, Bilda et
al. 2006), but imaging has its limitations and in any event it is applicable only to the
private musings of individual designers; others are unable to share what is locked inside
an individuals mind. Fish (2004) argues that the capacity for mental imagery develo-
ped in humans in prehistoric times for survival purposes as an aid in tasks like hunting;
evolution has not caught up with newer human activities, such as design as we know it
today, and therefore visualisation is used as an extension of imagery, or its amplihcation
(Fish & Scrivener 1990).
Imaging may, though, have to do with preconceived ideas that designers bring with
them at the outset of the design. Nany - though not all - of the D! designers conhrmed
that preconceived ideas and images existed when they started work on their projects:
Pesman (DI.1-Westraven Utrecht) said the image was directly in his head (p. 7); Meer-
tens (DI.5-Beertender) said, the design comes to you (p. 39); and Spark (DI.8-Carver
small car) stated explicitly, the designer always starts with an image of what it has to
look like, this image comes to mind from the beginning (p.69).
But in practically all cases, more than one person was involved in the project right from
the beginning. The team members, whether located in one place or dispersed geogra-
phically, had to communicate during meetings and between meetings. This they did
using visualisations, which were prepared ahead of time and shown in meetings or sent
around, but also produced them in situ, as part and parcel of an ongoing discussion.
Participants in design teams range from a small number of in-house designers to colla-
borations with partners and consultants from elsewhere, in addition to client represen-
tatives. Visualisations help make sure that everyone concerned shares the same mental
models of the products looks and functioning, materials, manufacturing process or a
particular detail thereof that is being discussed. One might say that without visualisati-
ons, it is inconceivable that a shared mental model could be achieved in a design team
(Goldschmidt 2007). This is the foremost reason for visualising in the design process.
We have mentioned that one of the parties taking part in design meetings is the client.
Clients vary greatly in the extent to which they wish, or are able, to get involved in the
design process. But in any event they must approve the design, or select from amongst
alternatives. Designers must therefore make an effort to convince the client of the
virtues of their proposals, sometimes to the point of justifying budget increases. To do
so they must show the client the designed entity in the most complete and attractive
manner possible, and in a mode the client, who is not necessarily technically adept, can
easily understand and appreciate. Digital devices such as graphically potent programs
(3D) are often used for this purpose, and so are models. This is the second reason for
visualisation in the design process.

The third, and the least interesting reason for our purposes here, is visualisation for
the purpose of construction or manufacturing. The visualisations made for this purpose
are technical in nature and today they are almost exclusively produced digitally (2D).
Sometimes the production of these documents is outsourced, frequently to `dot.com'
companies. We shall not discuss these visualisations any further in this paper.
2 The digital age
What do we actually mean, in the design context, when we say that ours is a digital
age'? !n terms of visualisation it means primarily that what has previously been done
manually, can now be done digitally in most cases, and more manipulations than were
previously possible are now achievable quite easily (e.g., photomontages, virtual walks
through buildings that do not exist yet, and so on). There are also new possibilities such
as digital prototyping which hardly existed a decade ago, mainly useful to industrial
designers. Many more new applications are undoubtedly due to make their appearance
in the foreseeable future. There are many advantages to digital drafting and modelling,
such as speed, accuracy, ease of revision, and ease of sharing with others regardless of
where they are stationed. But that is not the whole story, of course: sophisticated algo-
rithms permit the expansion of the world of manufactured and built forms, which are
less restricted than was hitherto the case. For example, the free form of the roof of the
stadium designed by Frei Otto for the 1972 Olympic games in Munich was a painstaking
design effort, realised after countless models were built to approximate the curvatures
of the membranes, which did not conform to mathematically expressible shapes. Nowa-
days digital means can not only easily save the considerable labour invested in building
actual models, but also calculate the structure regardless of its irregular geometry and
thus make it constructible. Frank Gehry is one of the better-known benehciaries of the
ability of digital means to cope with completely free forms in architecture.
!f digital visualisations are so ubiquitously benehcial, why have almost all D! designers
replied in the afhrmative to the question about the relevance of sketching in this day
and age? The answer has to do with the extraordinary advantages of sketching as a
visualisation mode throughout the design process, and especially in its early, prelimi-
nary phase. For experienced sketchers, which include almost every designer (architect
Nichiel Riediek, who testihes that he `does not use sketches that much; he says he can't
draw is an atypical exception (DI.3_Media Museum Hilversum, p. 22)), the production
of a rapid freehand sketch is a fast, hexible and effortless means of representation that
can be executed anywhere and requires no preparation, no equipment, and no skills
that need to be updated periodically. !t is therefore used very frequently in the process
of generating ideas, testing them and discussing them, in a group or even in private de-
liberations with oneself. To date, no digital means are available that come close to emu-
lating freehand sketching in terms of hexibility and ease, as well as speed and cognitive
economy, with the possible exception of academic prototypes that were developed with
unusual insights (e.g., Do 2002; Shapir et al., 2007). Likewise, both industrial designers
and architects continue to produce physical models, with or without the technical assi-
stance of digital means. The physical model is still necessary to allow us to get a better
feel for scale, texture or the mode of operation of an artefact, be it a small hand-held
gadget or a large building; indeed, all DI designers use models at least during the deve-
lopment phase of design projects. Digital devices, then, while helpful and in some cases
indispensable, are not necessarily the answer to every single aspect of the process of
designing. We shall have more to say about sketching in section 5 below.

3 Design education and practice
Industrial designers, and to a lesser degree architects, are taught to work systematical-
ly, according to well-established methods (Roozenburg & Eekels 1995) that specify all
of the design phases and the sequential activities that should be carried out in each. !n
mechanical engineering design the reliance on strict methodologies is even more strin-
gent, with a large body of published research and handbooks to support this claim (e.g.,
Jnsch et al., 2005). In industrial design brainstorming and other group methods are
taught and implemented in practice. However, in real life there are many constraints
and unexpected situations that force designers to divert from the perfect methods
learned at school. Thus the DI car designers state that They [at school] teach you to
follow the perfect process, but in reality it doesnt work that way an innovative project
doesnt keep to planning, it needs freedom. (DI.8_Carver small car, pp. 60-61). One of
the consequences of not being able to work by `solid rules' (D!.7_Ofhce chair, p.S2) is
that there are more iterations, more improvisations, more fresh starts than anticipated,
and this means more exploration and more experimentation. Consequently the best
tools are those best suited for exploration and experimentation, and they usually are
not the digital tools.
Despite the drive to use the latest and greatest methods which inevitably are largely
digital, the tools available are still inadequate for certain tasks, as pointed out above.
In architectural education many studio classes have become paperless, resulting in
projects that are detached from real materiality. Students are less occupied with develo-
ping rich, complex and sensitive spatial solutions and concentrate instead on the gra-
phic qualities of slick renderings. Teachers are unable to draw over students' sketches
and communication has become verbal only, related to PowerPoint presentations. With
that, one important facet of demonstration is lost in the studio critique: without paper
and pencil, the teacher cannot exemplify how something could or should be done, and
is reduced to verbal reactions only to the students work in progress. This is a dramatic
change in the otherwise still largely apprentice-style design education we practise in the
studio, and not a change for the better
1
.
Luckily, in both architecture and industrial design, in practice as well as in the educatio-
nal setting, three-dimensional models are still being made - not only hnal presentation
models but also study-models, often quite rough. The physical object fulhls needs that
no drawing can fulhl: it can be touched and interacted with in ways that are not possible
otherwise. It is therefore not surprising that even long before models are built, both stu-
dents and practitioners hnd ways to use artefacts, including ready-mades that happen
to be in the work environment, to represent or simulate properties of a designed object
even before the object has acquired form (Brereton 200+). The literature addresses the
mediating role of objects in our lives as knowledge translation agents, among other
roles (e.g., Whyte et al. 2007), but in this paper we discuss only visualisations that are
made expressly during the process of designing as a matter of quotidian practice.
4 Design phases interlocutors
The different design phases are distinguishable not only by their contents or the spe-
cihc activities undertaken, but often also according to the participants who take part in
them. It is hardly possible to arrive at a consensual breakdown of the design process
1 The commentary on design education is based on personal knowledge and experience.

into phases; in the Delft Interviews some designers talk about four phases, others
about six, and yet others about a different number of phases. The participants in each
phase may also vary according to the design task and the norms and practices of each
hrm. We shall therefore adopt a pragmatic breakdown into three phases, or rather si-
tuations that require communication between the designer(s) and other parties, whom
we call interlocutors. !nterlocutors are those for whose beneht the designer produces
visualisations, the party with whom he or she (or they) interacts in the normal course
of the design process. The three phases/situations are: a) preliminary design; b) de-
velopment phase; and c) discussions with clients and users. Table 1 maps the modes
of visualisation reported in the DI according to these phases. This mapping cannot be
entirely accurate, since the interviewees were not asked specihcally when they used
particular visualisations, and Table 1 rehects only an interpretation of what was said.
Nevertheless, it does provide a close enough picture to what we assume is the reality
of practice in architecture and industrial design.
The hrst conclusion we can draw is that in terms of visualisation architectural and
industrial design practices are quite similar, throughout the design process. !n both,
sketching is used heavily during the preliminary and development stages, and to some
degree in discussions with clients or users. Clients may be involved throughout the pro-
cess and discussions with them do not constitute a separate phase, of course. Rather, in
this rubric we mean primarily formal and less formal presentations to clients at various
points of decision making.
Preliminary design
At the outset the major means of visualisation is sketching. Sketches are made during
the search for a solution principle, in most cases following an initial, preconceived idea,
by the leading designer(s). Architects make more models than do industrial designers
in this phase, sometimes in compensation for the lack of drawings and sketches (DI.3_
Nedia Nuseum Hilversum), and at other times because some architects may hnd it hard
to imagine complex spatial relations without models. Architectural sketches and dra-
wings, as opposed to product design drawings, tend to be two-dimensional, using the
conventions of orthogonal projections which do not describe spaces directly. Architects
are trained to imagine spaces on the basis of plans and sections, but a model helps
to perceive the space and its proportions, and test the accuracy of the image. Models
are less frequent in the preliminary phase of industrial design. One reason may be that
customary three-dimensional drawings are adequate - and more economical - repre-
sentations at this stage. It may also be the case that rapid prototyping has become the
standard mode of modelling, at least for smaller artefacts; making them is reasonably
cheap and fast, but preparing the necessary CAD hles is time consuming and may also
prematurely hx the design properties. Designers may feel that they prefer the freedom
of sketches before they commit themselves to CAD hles for the purpose of producing
a study model.

Table 1: Visualisation modes in the Delft Interviews
We note that no digital drawings are produced at this phase. This is not surprising as
neither dimensioned drawings nor fancy images are needed in this phase, in which
the designers communicate primarily among themselves, in search of a viable solution
proposal that the designers can defend and which stands a chance of approval by the
client. The sketch, at this phase, is a compact laboratory in which designers can expe-
riment with different ideas freely with no cost or any other negative consequences in

case of failure. This encourages more experimentation with extreme, unusual and po-
tentially innovative concepts which, due to their novelty, require more testing. Suwa et
al. (2001) explain how designers beneht from sketches because they can make disco-
veries in them, including the regrouping of elements, which offers new interpretations.
Fish (2004) and Goldschmidt (e.g., 2002) have advanced similar arguments. Whereas
this facet of sketching is mostly studied in the context of individual designers working
alone, in teams sketching is essential to idea-generation sessions: it does not increase
the number of ideas, but it signihcantly improves the degree to which they build on one
another (van der Lugt 2005), which is normally a precondition for creativity.
Development
The development phase is usually carried out by a larger group of people than the one
involved in preliminary design. It is also more diverse in terms of expertise we include
in the group, or team, all the consultants, internal or external, who are involved in the
project. !n most cases participants have dehned roles, and coordination among them is
a major issue. Therefore the amount and quality of communication is most important
as good coordination results in an efhcient, streamlined process (as much as constraints
permit), whereas poor coordination causes misunderstandings and conhicts that are
costly and demoralising. A key to good coordination is a high level of understanding and
agreement amongst team members regarding the designed entity, which is achieved
through face-to-face meetings and conversations which include sharing of documents,
also when members are not physically co-located. Naturally, visualisation plays a crucial
role in all of these team deliberations. The Delft Interviews show that practically all
modes of drawing and physical models are used in this phase (see Table 1), each for
the purpose it serves best.
Figure 1: Models in dialogue: Denys Lasdun, National Theatre, London, c. 1965
2
.
2 Photo by Behr Photography.

Figure 1 shows a stack of study models made during the long years in which the design
of the National Theatre in London, including three different performance halls, was
being developed in the ofhce of architect Denys Lasdun. Until not too long ago many
large architectural hrms employed full-time model makers, and practically every product
design hrm has at least a small workshop in which models (not rapid prototypes) can be
executed. Most such models are fairly rough and their purpose is study and evaluation.
As evident from Figure 1, the same entity may be modelled again and again, each time
rehecting revisions introduced as a result of assessments of previous versions, until a
satisfactory proposal is achieved. This mode of usage resembles sketching and rough
preliminary models are sometimes referred to as 3D sketches. Students, too, are al-
most always required to produce models in the course of developing design projects in
both industrial design and architecture, and as in practice, these models are different
than `hnal' or presentation models that are often outsourced to specialised experts or
produced as rapid prototypes by 3D printers or similar digital machines. Study models
continue to play an important role in design development, arguably more so in archi-
tecture, especially since all stakeholders, including the client and others who may lack
design expertise, can relate to them easily.
Sketches and other drawings continue to be essential in the development phase. The
state of the design keeps evolving and changes, major or minor, are subject to frequent
discussions and decision sessions. Consultants input needs to be integrated into the
design and this requires considerable coordination efforts, and the resolution of pro-
blems that keep coming up. Communication therefore builds on detailed representati-
ons of the latest versions of design drawings, be they measured plans or still, free-hand
sketches. For communication over distances fax machines and the Internet are used
to transmit information, including drawings. By comparison to the preliminary phase,
in which sketches mainly express ideas and concepts and may be rather abstract and
schematic, in the development phase sketches are more concrete and detailed, and
describe the actual designed entity in its many facets. We begin to see digital drawings
as well: CAD measured drawings are produced so that all designers and consultants
have accurate information as the basis for their interventions. In the case of industrial
design, this includes many more 3D drawings than in architecture. Fancier, so-called
presentation drawings are still rare at this phase, except for interim decision-making
meetings for which they are typically prepared. All modes of visualisation are thus ex-
ploited at school and in practice to help develop a design project, as cogently stated by
Paradiso et al. (2002):
Projects develop through sketches in cardboard and on trace [paper]; they
are pushed further through exacting CNC-milled projects and detailed
renderings. But students are as likely to work through complex details by
hand and to look to the computer as a means to produce quick analytical
sketches. (p 2)
Discussion with clients and users
Discussions with clients and users take place at all stages of the design process, of
course, but are typically built into certain checkpoints in which major decisions are
taken. For those occasions designers prepare visuals that are meant to convince the
client or users of the merits of the overall proposal, or as regards certain aspects of it.
The Delft Interviews show (Table 1) that the means used for that end are mixed: from
sketches, which are probably used in informal meetings in which certain details may be
discussed, through models, to frequent digital drawings (presumable, mostly 3D rende-

rings), and even movies. Often, designers refer to presentations they prepare, which
may indicate the use of tools like PowerPoint in order to show visuals, undoubtedly
accompanied by oral explanations.
This mixed media panorama is most appropriate, and it applies to all branches of de-
sign, architecture and industrial design included. Each mode of visualisation has its own
advantages and is utilised by designers to maximise its benehts. Before the computer
many more manual drawings were made, of course, but even before drawings were
the standard means of visualisation (that is, before paper became readily available and
sufhciently inexpensive, after the moving print revolution in the last third of the 1Sth
century), models were made to be presented to patrons in order to secure their appro-
val. Figure 2 shows a fresco by Vasari from the mid-16th century, depicting the architect
Brunelleschi presenting a model of San Lorenzo to his client, Cosimo de Medici, who
commanded the church. The model is a fairly accurate representation of the famous
Florentine church. Earlier pictures and mosaics bear evidence of the fact that model
presentation to patrons was an established practice (for example, a beautiful mosaic at
the Kariye Museum in Istanbul, dated c. 1320, depicts Theodore Metochites, donor of
the Chora, with a model of the church/monastery).
Figure 2: Fresco by Vasari (1565) showing Brunelleschi presenting the model for
the church of San Lorenzo to Cosimo de Medici
3
.
During the Renaissance, making models for the beneht of clients stood in sharp contrast
with the production of technical drawings which were made for the masons-builders of
edihces (for example, a drawing by Jacopo Berloia from the early 16th century shows
the professional architect, accompanied by his scholarly advisors, presenting plans to
the workmen who were building the Rotunda in Rome). Such drawings became standard
3 From: Ettlinger, L.D. (1977). The emergence of the Italian architect. In Kostof, S.
(ed.), The architect. University of California Press, Berkeley, pp. 96-123; illustration p. 110.

practice following the introduction of orthogonal projections as the mode of delivering
information about the geometry of spaces and objects. Perspective drawings were also
made from that time, of course, and gradually joined models as formal renderings, but
the two co-existed for centuries more as complimentary visualisations rather than rival
or competitive modes of expression.
Drawings that are made for clients or users are, as pointed out earlier, a mixed bag,
depending on their purpose. The more tools we have at our disposal the more there is
to choose from, and the wise designer knows that and specihcally adjusts his choice to
the goals the visualisation is meant to achieve. Whyte et al. (2007) distinguish between
huid and frozen visual materials in design. The former are `mobilized when comment,
input or modihcation is required' (p. 22), which corresponds mostly to the phases of
preliminary design and the subsequent development phases and to a lesser extent to
discussions with clients and users. The latter, frozen visuals are characterized by greater
certainty. The authors remark that such visualisations have several functions, including
use for tactical and political reasons (p. 23). When thus used, the interlocutor is often
the client or the users. Figure 3 captures three instances of the usage of visuals in the
design process of designing a herbarium. In our terms Figure 3a describes a preliminary
design usage of drawings; Figure 3c is taken from the development phase; and Figure
3b exemplihes a `huid' usage of drawings in a discussion with users. As is evident from
this example, in all three cases the interaction among the concerned parties is entirely
dependent on the use of visuals, in this case sketches and drawings.
Figure 3: !nteractions with visual representations, some of which are huid and others frozen, in
design work at Edward Cullinan Architects: a) the founder of the practice and an architect wor-
king on the project talk about the design concept; b) the ideas are presented and discussed with
the library staff; c) working meeting between the project architects and the engineers
4
.
The public
Designers do more than bring into being the best possible buildings and products; they
also take part in the cultural and artistic discourse of their time. For some designers this
becomes a major activity and they are interested in making statements through visuali-
sations they exhibit and publish, in addition to other modes of representation (oral and
written expressions). At times of heated debate designers even publish manifestos and
produce visuals which are loaded with symbolic meaning. This may be more signihcant
in architecture, whereas in industrial design it is the products themselves that are made
with similar intentions. Figure 4 is an example of a drawing made in James Stirlings
ofhce: it is not made for communicating information to fellow designers or to other sta-
keholders in the project, nor is it meant for the builders. Instead, it is a statement about
design thinking and representation, made during the early years of Postmodernism and
4 Figure 3 and its caption reproduced from Whyte et al. (2007), p. 22.

meant for the cultural avant-garde of the time. Stirling chose to present an isolated
selected idea, in an unusual view (worm-view axonometric drawing).

Figure 4: James Stirling and Michael Wilford (1976). Axonometric up-views of major
elements in the Westfalen Museum, Dsseldorf (competition entry, not built)
5
.
Goldschmidt (2004) distinguished between private and public representations. The
former are those visuals that individuals and teams produce for themselves, as thinking
and communication aids; the latter are made in order to advance ideas and concepts vis-
-vis particular interlocutors or the (relevant) public at large, as in Stirlings case. Other
architects and designers produced very different kinds of visuals of the same category;
those visuals served as agents of the unique design identity those designers wished to
publicise. The cultural discourse in which design participates, which is an extension of
practice (or vice versa?), also helps shape practice because every designer works within
a cultural context, even if the level of explicit awareness of its grinding wheels, and the
attention paid to it, may vary considerably from one designer to another.
5 The robustness of sketching
Sketches are the most dominant mode of visualisation in design practise. Today they
are beginning to be produced digitally as well as manually, but sketches on paper are
far from obsolete in the design world. In fact design schools have re-discovered the
necessity of training students in free-hand drawing, after years of somewhat unrealistic
hopes that digital means will happily replace all manual design output. Sketches are not
all of a kind; Ferguson (1992) divides them into the thinking sketch, the talking sketch,
and the prescriptive sketch. In our terminology this means: sketching as a cognitive aid
in the generation of ides; sketching as an agent of communication, and sketching as
instruction for execution (e.g., for the construction of manufacturing). In this paper we
have largely addressed the talking sketch, which is prevalent in design practice where
5 Source: Landesgalerie Nordrhein-Westfalen in Dsseldorf, James Stirling and Partner
with Werner Kreis, Robert Livesey, Russ Bevington, Ueli Schaad, Lotus International, 1977, Vol.
15, 58-67.

40
individuals rarely work by themselves. However, thinking - and design is hrst and fo-
remost a thinking activity occurs above all in the individuals mind, and the thinking
sketch helps in the conversation the designer holds with him or herself. We have already
mentioned briehy the cognitive advantages associated with sketching, which account
for the robustness of sketching in design practice for over half a millennium now, since
paper became the standard medium for visualisations. Figure 5 is a diagram explaining
the status of sketching in visualisation as part of the design problem-solving process.
Problem
Well-structured
ll-structured
Other task
Search space
Non-visual representation Visual representation
nternal (imagery) External: e.g., Drawing,
Notation (+3D objects)
Thinking aid Memory aid Communication/
specification aid
Rapid sketch
Measured drawing
Thinking sketch* Talking sketch* Prescriptive sketch*
*|e|guscn12}
Cognitive advantages include:
- Rapid (minimal cognitive resources)
- Flexible stop-rules
- Only minimally rule-bound
- Refersible/transformable at any stage (overlaping)
- Tolerent of incompletion
- Tolerant of inaccuracy & lack of scale
- Provides stimuli/(unexpected) cues
- Supports feedback loops (internal/external rep.)
Compositional task
Knowledge
base
External
visual
stimuli
Figure 5: Sketching as a mental facilitator in complex, visually mediated tasks.

The thinking sketch, on which we wish to focus here, does not need to be complete or
precise. In fact it may be partial, vague, incomplete, inaccurate, not necessarily true to
scale, and its level of concreteness of abstraction may vary sharply (within and between
sketches). Furthermore, it can be stopped at any time without losing what was done
to that point.
We shall conclude the discussion with a brief enumeration of what we hold to be the
major cognitive advantages of the sketch, which designers recognise and capitalise on,
and which secures its utility in the design process for the foreseeable future.
The rough sketch:
is rapid (minimal cognitive resources). Sometimes a few pencil strokes
on paper are enough to capture an idea, a shape, a mechanism or a
relationship among parts.
has hexible stop rules (no algorithm governs the steps to be taken).
The sketcher may stop any time and ultimately when the outcome
looks satisfactory (and not when rules are satished).
is only minimally rule-bound (example-based versus rule-based
reasoning). More than anything else, the thinking sketch is a tool of
reasoning. Reasoning is said to be either rule-based or example-based
(Sloman 1996); sketching facilitates example-based reasoning which
enjoys considerable freedom from rules (other than the rules of
orthogonal projections, which are normally adhered to). This in turn
has the potential of expanding the design space in which a solution is
sought and may therefore enhance innovation and creativity.
is reversibleftransformable at any stage (overlaying facilitates rapid
changes). The sketcher may change his or her mind at any time and
retract any number of steps.
is tolerant of incompleteness and vagueness. Especially when the
sketch is made as part of the dialogue the designer holds with him or
herself, shorthand is enough; the designer will recognise intentions
and will be able to mentally complete any missing or vague
information.
is tolerant of inaccuracy and lack of scale. When concepts or broad
intentions are important, accuracy and correct scale are not always
necessary and there is no need to labour over them.
supports feedback loops. Sketching and mental imagery work in
tandem: one informs the other. The ensuing cycle is in fact a feedback
loop which helps push the process forwards.
6 In conclusion
Whatever the differences in the design process between architecture and industrial de-
sign, practitioners in both helds `think visually' and constantly visualise their thoughts.
Often visualising is in fact thinking and not merely the recording of thoughts that had al-
ready been entertained in the mind. Designers in both helds use all representation and
visualisation means available to them, from freehand sketching and manual drawing to
digital drawings, through physical models and various simulations and movies. Natu-
rally, more sketches are made in the front edge and more two- and three-dimensional
digital drawings are produced later in the design process. Models are built throughout
the process: they tend to be manual in architecture and digitally based prototypes in

industrial design. In essence, the kinds of visuals that are made in practice are not
very different from the ones that have been made for hundreds of years, although we
can now produce many of them digitally. A notable exception are the visuals made for
display and publication not in the context of regular practise but rather as participants
in a cultural discourse, where the norm is to break conventions and present innovative
breakthrough concepts. The means utilised are correspondingly often novel.
!n practise, where efhciency is an over-riding value and goal, means are adjusted to
ends, and the most effective visuals are used for each purpose, i.e. the most conve-
nient, most economical and most potent modes of visualisation are selected at any
given time. The fact that sketching is still in wide and frequent use is evidence of the
fact that, at least for the purposes of study and exploration, we have no tool that rates
higher. We must therefore conclude that sketching has advantages that to date cannot
be emulated by any other mode of visualisation. Sketching will continue to be in good
currency as long as it is the state of the art.
Acknowledgment
The writing of this paper was partially supported by a grant to the author from the fund
for the promotion of research at the Technion, hereby gratefully acknowledged.
References
Athavankar, U. A. (1997). Mental Imagery as a Design Tool, Cybernetics and
Systems, Vol. 28, 25-47.
Athavankar, U. A. & Mukherjie, A. (2003). Blindfolded Classroom: Getting De
sign Students to Use Mental Images. In Human Behaviour in Design,
edited by Lindemann, U., Springer Verlag, Berlin, pp. 111-120.
Bilda, Z., Gero J. S. and Pyrcell, T. (2006). `To Sketch or not to Sketch? That is
the Question, Design Studies, 27 (5), 587-613.
Brereton, M. (2004). Distributed Cognition in Engineering Design: Negotiating
between Abstract and Material Representations. In Design
Representation, edited by Goldschmidt, G. and Porter, W. L., Springer
Verlag, London, pp. 83-103.
Do, E. Y-L. (2002). Drawing Marks, Acts and Reacts: Toward a Computational
Sketching Interface for Architectural Design. AIEDAM, Vol. 16,
149-171.
Ferguson, E. S. (1992). Engineering and the Minds Eye. The MIT Press,
Cambridge, MA.
Fish, J. (2004). Cognitive Catalysis for a Time-lagged Brain. In Design
Representation, edited by Goldschmidt, G. and Porter, W. L., Springer
Verlag, London, pp. 151-184.
Fish, J. and Scrivener, S. (1990). Amplifying the Minds Eye: Sketching and
Visual Cognition, Leonardo 23, 117-126.
Goldschmidt, G. (2002). Read-Write Acts of Drawing. In TRACEY (Internet
journal dedicated to contemporary drawing issues); issue on Syntax of
Mark and Gesture. http://www.lboro.ac.uk/departments/ac/tracey/
somag/gabi.html. Loughborough University, UK (accessed April 28,
2008).
Goldschmidt, G. (2004). Design Representation: Private Process, Public
Image. In Design Representation, edited by Goldschmidt, G. and
Porter, W. L., Springer Verlag, London, pp. 203-217.

Goldschmidt, G. (2007). To See Eye to Eye: The Role of Visual
Representations in Building Shared Mental Models in Design Teams,
CoDesign 3 (1), 43-50.
Jnsch, J., Nissl, A., Strasser, C. and Bruch, C. (2005). The Importance of the
Integration of Design Methods in Robust Engineering Design. In
Proceedings of 15th International Conference on Engineering Design -
ICED 2005, edited by Samuel, A. and Lewis, W. Institute of Engineers
Australia, Melbourne (CD-ROM).
Paradiso, A., Baxter, E. and Baumberger, M. (eds.) (2002). Foreword.
Retrospecta 01-02. New Haven: Yale School of Architecture.
Roozenburg, N.F.M. and Eekels, J. (1995). Product Design: Fundamentals and
Methods. Wiley, Chichester.
Shapir, O., Goldschmidt, G. and Yezioro, A. (2007). Conceptual Design: An
Operational Prescription for a Computer Support System. In Computer
Graphics, Imaging and Visualisation: New Advances, IEEE & Computer
Society, 4th CGIV07 International Conference, Bangkok, edited by
Banissi, E., Sarfraz, M. and Dejdumrong, N., pp. 513-521.
Sloman, S. A. (1996). The Empirical Case for Two Systems of Reasoning.
Psychological Bulletin, 119 (1), 3-22.
Suwa, M., Tversky, B., Gero, J. S. and Purcell, T. (2001). Seeing into Sketches:
Regrouping Parts Encourages New Interpretations. In Proceedings of
2nd Conference on Visual and Spatial Reasoning in Design:
Computational and Cognitive Approaches, Bellagio, edited by Gero, J.
S., Tversky, B. and Purcell, T., pp. 207-219.
van der Lugt, R. (2005). How Sketching Can Affect the Idea Generation
Process in Design Group Meetings. Design Studies, 26 (2), 101-122.
Whyte, J. K., Ewenstein, B., Hales, M. and Tidd, J. (2007). Visual Practices
and the Objects Used in Design. Building Research & Information, 35
(1), 18-27.

IDE+A
Project Management

45
4 Project and risk Management in architecture and
industrial design
Prof. dr. ir. J.W.F. (Hans) Wamelink
1
and dr. John L. Heintz
2
1
Professor, Department of Real Estate and Housing,
Faculty of Architecture,
Delft University of Technology,
2
Assistant Professor, Department of Real Estate and Housing,
Faculty of Architecture,
Delft University of Technology
Abstract
This paper describes the ways in which factors of project environments determine
the application of management concepts, particularly risk management, in industrial
design engineering (IDE) and architectural projects. The paper is based on a set of
eight design cases prepared for the IDE+A conference. Given the limited number of
cases and the constraints imposed by the overall case study design, it was necessary
to supplement the insight derived from the cases with a review of generally accepted
accounts of the design process in IDE and architecture. By sorting the cases according
to the emergent dimensions of internal vs. external project and market- vs. client-driven
and comparing the applications of project management concepts in each case, we will
hnd that the environmental factors provide a clearer picture of how and why different
project management concepts are applied than do the disciplinary factors. Indeed, by
focusing on the project environment factors we may be in a better position to predict
project management approaches will be required in design-build or other unusual or
innovative project organisations.
Keywords: Architectural Design, Industrial Design Engineering, Project Management,
Risk Management
Introduction
Although designers of all sorts are accustomed to operating under uncertainty, and
engage in many activities intended to reduce that uncertainty, they tend not to think of
their work in terms of risk or project management. As design projects ht the management
dehnition of `project' extremely well this omission seems odd. Further, while designers
are sometimes reluctant to speak of risks, and the word risk is seldom used in the
design literature, the literature does cover most of the issues that are covered by the
notion of risk, and has done so for some time. However, which risks are considered
to be signihcant, and which project management techniques are used seems to vary
greatly between different design projects. By examining a range of cases across from
the helds of !DE and architecture we hope to shed more light on the question of when
and where to best apply different notions of risk and project management in design
projects.
We will begin by comparing how project management is applied to IDE and architectural
projects. We will then use the cases to derive two signihcant dimensions of the project
environment. By sorting the cases along these dimensions and again comparing the
applications of project management concepts we will hnd that environmental factors

46 Project Management - J.W.F. Wamelink and dr. J.L. Heintz
provide a clearer picture of how and why various project management concepts are
applied than do disciplinary factors. Because our focus is on the management of design
projects, we begin with a description of projects as seen from the perspective of
management.
Project managment
!n management literature, a project is usually dehned as a one-time set of activities
with a dehnite beginning and ending point. Nanaging projects is dehned as the task
of completing these activities on time, within budget and according to specihcations
(Robbins & Decenzo, 2004). Robbins and Decenzo attribute the growing popularity
of project management to the increasing rates of change in the contemporary world.
Project management hts well within a dynamic environment, and it responds to the
need for hexibility and rapid response. Organisations are increasingly undertaking
projects that are somewhat unusual or unique, have specihc deadlines, contain complex
interrelated tasks requiring specialised skills or are temporary in nature. These types of
projects are not well suited to the standardised operating procedures that guide routine
and continuous organisational activities. Other means are required to manage this type
of projects successfully.
Any project is about causing a change in an uncertain situation. in other words, a
project involves developing or making something new. Thus one common characteristic
of all projects is discovering the unknown. Inherent to such endeavours are risk
and uncertainty. However, the characteristics of risk and uncertainty differ in across
projects.
Lock (2007) classihes projects in four different general types:
- Civil or chemical engineering and construction projects (buildings,
tunnels and bridges)
- Manufacturing projects (automotive, pharmaceuticals, aircraft)
1
- Management projects (implementing new IT systems, reorganisation
projects)
- Scientihc research projects
The hrst two types are relevant in the helds of !DE and architecture. Lock states that
construction projects incur special risks and problems deriving from their organisations.
They may require massive capital investment, in addition to rigorous management
of progress, hnance and quality. !n many cases, a large number of independent
organisations are involved in the design and construction of a new building. Furthermore,
the design process continues even after construction has been contracted out, as the
primary contractor, sub-contractors and suppliers all redesign the various components
and details of the building. This may also be the case for some large-scale and complex
manufacturing projects (e.g. aircraft), in which a number of organisations collaborate
to develop highly complex products. Such internationally oriented projects are prone
to higher risk and difhculties in control and coordination that arise from organisational
complexity, national rivalries, contracts and other factors. Most industrial products,
however, are simpler, with the design concentrated within a small group of actors, and
1 Specihcally Lock considers manufacturing projects to be either one of a kind, in which
case design engineering and production, are all included within the project; or series produc-
tion, in which only design is included in the project.

47 Project Management - J.W.F. Wamelink and J.L. Heintz
with the possibility of a global distribution of part sourcing. In most product-design
processes, control of both design and production is much held more closely within the
design team. Even the simplest product, however, involves unexpected interactions
between design intent, user preferences, available technology and production
systems.
In the building industry, project management is generally carried out by project
management consultancy hrms. !n industrial design projects, it is often carried out
internally by persons bearing titles such as manager of product development.
These characteristics imply that projects are surrounded by risk and uncertainty. An
important aspect of managing these projects is therefore dealing with these risks and
uncertainties. Winch (2002) described the project process as the dynamic reduction of
uncertainty through time (see Figure 1.). At the inception stages of a project, uncertainty
is very high: the asset of the future is little more than an idea and possibly a few
sketches. How high depends upon a number of factors, such as the extent to which
standardised components and solutions can be used. It is clear that reducing uncertainty
is an important part of managing projects: As the project moves through the life cycle,
uncertainty is reduced as more information becomes available ambiguities in design
are resolved. (Winch, 2002).
The aim of the project manager , or more in generally the function of project
management, is to achieve success in all aspects of the project. Conditions for the
successful application of business strategies are also referred to as success factors.
`Success factors have been dehned as the critical key areas where things" must go
right for the business to hourish' (Koutsikouri, Dainty et al., 2006; Rockart, 1979).. !t is
necessary to distinguish between the success factors, which lead to successful projects,
and the success criteria, which are used to measure project success (Cooke-Davis,
2002). Thus, although success factors and success criteria commonly address similar
issues, we must clearly delineate the differences between cause and effect.
0 Time
Amount of
information
processed
Uncertainty
amount of
information
required
none
none
all
all
inception completion
0 Time
Amount of
information
processed
Uncertainty
amount of
information
required
none
none
all
all
inception completion
Figure 1: The project process as the dynamic reduction of uncertainty
(Winch et al., 1998)

Traditionally, these success criteria have been understood to refer to the three basic
aspects of project management - cost, time and quality - as described in the mid-1980s
by researchers such as Barnes (Barnes, 1988). Barnes later replaced the concept of
quality with performance: the notion that, upon completion, a project should do what it
was intended to do (Lock, 2007). Barnes drew these three project objectives as a triangle,
to illustrate that the three primary objectives are interrelated. A management decision
to place greater emphasis on achieving one or two of these objectives must sometimes
be made at the expense of the remaining objectives. Other scientists expanded the
model to include additional aspects, such as people (to stress the importance of the
management, organization and motivation of the people involved in the project) (Kliem
& Ludin, 1992).
Performance
People
Cost
Time
Figure 2: triangle of project objectives
A second distinction that must be made is between 1) the internal characteristics of
project organisation such as time cost and performance goals, and 2) the external
characteristics, such as customer satisfaction (Shenhar, Dvir et al., 2001; Koutsikouri,
Dainty et al., 2006; Meredith and Mantel, 2006). It is conventionally assumed that
success, as measured by internal project characteristics, will necessarily lead to
customer satisfaction. but the Sydney Opera House, however, is a famous example
of the potential for a disconnect between the two. More importantly, building projects
are lengthy and client organisations are in a constant state of change. as the frequent
practice of altering recently completed buildings attests.
Key themes in the description of project management in the IDE+A cases
In the section above, several basic aspects of projects and project management were
introduced. Although theories of project management are much more mature than
indicated here, for the objective of this paper this theoretical basis is sufhcient
2
. To
describe the differences concerning project management between the eight selected
cases within the IDE+A project, we use the most important concepts from the foregoing
section:
- Environmental properties of the project, in terms of risks and uncertainty
- Important management activities (risk management, estimating, scheduling,
organisation)
- Project results, in terms of budget, time and performance
2 Readers interested in more in-depth reading on project management may refer to
(Lock, 2007; Winch, 2002; Morris, 1994; Morris, 2001).

Performance
People
Cost Time
Project results
Managment activities
Project environment
Risk and
uncertainty
Figure 3: Aspects investigated in the cases
As show in Figure 3, risk and uncertainty are determining factors in the description of
the project environment. We therefore discuss risk and risk management as it appears
in design projects and in the design literature.
In the last decade, risk management has become an important consideration in project
management (Lock, 2007). The term has emerged from management studies, and
has slowly become accepted in the building industry. However, the term seems still to
be novel in IDE, as indicated by its absence from recent books such as Von Stamm
(2003), which contains (only a single mention of risk). Older texts .such as Roozenburg
8 Eekels, (1991) sometimes do not mention `risk' at all. !n the scientihc literature, we
were only able to hnd one team (at the Eindhoven University of Technology) working
on risk management in product design. This does not mean that the concerns of
risk management have been ignored. For many of the issues associated with risk are
considered to be standard issues in the product design process. Keizer, Vos and Halman
have studied perceptions of risk in product product-design processes (Halman, 2002;
Keizer et al., 2005). They have found that, when prompted, product design teams
identify a large number of risks in their projects. In one study, Keizer et al listed 142
different risks identihed by their interviewees (Keizer et al., 200S). Clearly, product
designers are well aware of the risks associated with their projects. It is simply that they
consider them to be normal to design practice.

RunkIng PerceIved rIsks requency
1 CommunIcuLIng Lhe new roducL Lo LurgeL consumers z6
z OrgunIsuLIon und munugemenL oI Lhe rojecL z
SLubIIILy oI Lhe roducL, whIIe In sLoruge In roducLIon
IunL, In shoJwurehouse, durIng LrunsorLuLIon or uL
zz
q QuuIILy und suIeLy requIremenLs oI roducLIon sysLem
(IucIIILIes und ersonneI)
18
ConsLunL und redIcLubIe quuIILy oI suIy by suIIers 16
6 PossIbIe neguLIve exLernuI reucLIons by key oInIon
Iormers or InLeresL grous
1
; AdequuLe roducLIon meuns (equImenL und LooIs)
uvuIIubIe when needed
1
8 New roducL IuIIIIs InLended IuncLIons 1
q New roducL meeLs consumer sLundurds und demunds 1
1o New roducL`s ueuI Lo generuIIy ucceLed vuIues
(heuILh, suIeLy, nuLure, envIronmenLuI Issues)
1z
Table 1: Nost frequently perceived risk issues within categories,
adapted from Halman (2002).
From their list of perceived risks, Keizer et al derived a shorter list of the 10 most
frequent.
It is interesting to note that in the case study material, the industrial design engineers
provided very little information on risk management in their responses to either the
original questionnaire, or to a set of follow-up questions. On the other hand, the
respondents frequently mentioned the individual perceived risks noted by Keizer et al.
Thus, it seems as if, while the terminology is not widely accepted in IDE, the issue is
fundamental to how industrial designers go about their work. Indeed MMID devotes an
extensive section of their website to risk management (MMID 2007). It is possible that
within a few years the terminology will be standard in the held of industrial design.
In construction management, the term risk is more widely accepted, and researchers in
this domain have also indexed perceived risks. Contractors have long been understood to
run risks, the weather being only the most obvious, but contractors identify a signihcant
number perceived risks (El-Sayegh, 2007; Mbachu & Vinasithamby, 2005). Consultants
too perceive risks in their work. In their study of Australian building consultants and
contractors, Nbachu 8 vinasithamby (200S) identihed ++ distinct sources of risk internal
to the project. The eight external sources of risk they identihed were all related to the
ability to complete the project and to inhuences on project costs. End users and the
market for buildings were not perceived as risks by either consultants or contractors.
One thing emerges clearly in comparing lists of perceived risks in Architecture and IDE.
In the construction industry, perceived risks are narrowly focused on project organisation
and management issues. In contrast, the risks perceived in IDE span a wide range of
issues, including consumer acceptance and marketing, public acceptance risks, and
commercial viability risks. These perceived risks could be compared as follows:

PerceIved rIsks DE ArchILecLure ConLrucLor
CommunIcuLIng Lhe new roducL Lo LurgeL
consumers
yes no no
OrgunIsuLIon und munugemenL oI Lhe rojecL yes yes yes
SLubIIILy oI Lhe roducL, whIIe In sLoruge In
roducLIon IunL, In shoJwurehouse, durIng
LrunsorLuLIon or uL home
yes no no
QuuIILy und suIeLy requIremenLs oI roducLIon
sysLem (IucIIILIes und ersonneI)
yes no yes
ConsLunL und redIcLubIe quuIILy oI suIy by
suIIers
yes no yes
PossIbIe neguLIve exLernuI reucLIons by key
oInIon Iormers or InLeresL grous
yes yes no
AdequuLe roducLIon meuns (equImenL und
LooIs) uvuIIubIe when needed
yes no yes
New roducL IuIIIIs InLended IuncLIons yes yes no
New roducL meeLs consumer sLundurds und
demunds
yes no no
no New roducL`s ueuI Lo generuIIy ucceLed vuIues
(heuILh, suIeLy, nuLure, envIronmenLuI Issues)
yes yes
Table 2: Comparison of perceived risks in IDE and Architecture; IDE risks
after Halman (2002), others supplied by the authors
3
The key difference between the two disciplines is the degree to which risks associated
with the market or with production are perceived, carried and dealt with by the designers.
In architectural projects, the designers carried little or no risk. In IDE projects, the
designers were often situated within an organisation carrying the project risk, and be
therefore more attentive to these risks and more able to address them.
The cases
Turning now to the cases, we began our analysis of the cases by creating a table in
which we could compare a number of salient characteristics of each project. We were
looking for patterns, for predictors of project management behaviours.
3 The lists of perceived risks in architecture and construction were compiled by the
authors based on traditional project organisations, in which design and construction are carried
out by different parties. The distribution of risk perceptions may be different in newer integrat-
ed project organisations.

DIscIIIne DesIgner OrgunIsuLIon ProjecL PurLIes PeoIe
ArchILecLure Ceezed DesIgn IIrm WesLruven,
ULrechL
; zoo+
VVKH DesIgn IIrm DE BuIIdIng
TUDeIIL
;+ qo?
NeuLeIIngs
RIedIjk
DesIgn IIrm MedIumuseum
HIIversum
1;
MurLInI
ArchILecLs
DesIgn IIrm 1-z- HuIs
DE MMD DesIgn IIrm BeerLender q 1q +
ndes HeuILh Cure
ProducLs
CureIIIL 1 1o+J qo+
Ahrend urnILure
munuIucLurer
Azo Desk
ChuIr
(consuILunL
desIgners?)
+
Surk DesIgn &
Curver
EngIneerIng
DesIgn IIrm &
deveIomenL
IIrm
Curver z meI-18
Table 3: The cases
Organisations
There are two types of organisations represented among the cases: 1) design hrms,
which consisting of a staff of sometimes multidisciplinary designers (some of whom
are multidisciplinary) with, normally, no investment in the product and no productive
capacity, and 2) large companies whose business is the design, production and
delivery of consumer products. For simplicity, we can call projects in the hrst type of
organization external projects, as the design team is external to the producer. The
second type of organisation has an in-house design staff, and is responsible for the
organization of, if not the actual, production of their products. We refer to the projects
in these organisations as internal projects, as the design team is (largely) internal to the
producing organisation. In all of the architectural design cases, the design was carried
out by an external design hrm, this was also the case in two of the !DE cases, although
in one of these cases there is a very close relationship between the designer (Spark
Design) and the producer, Eurotool/Carver Engineering. In the other two IDE cases, the
design was carried out by large product hrms using both internal designers and external
design consultants.
Thus, it is already clear that we cannot say that one organisational form is inherent
to either !DE or architecture. And even though we see only design hrms, among the
architecture cases, this is not a matter of principle. but only of custom. In the case of
the 1-2-3 Huis, the architectural hrm had a much closer relationship with the production
company similar to that of Spark Design and Carver Engineer. The advent of design-
build and other integrated contract forms is leading to new organisations where, at
least for the term of the project, design and production are more integrated. In Japan,
the building industry is dominated by large hrms with in-house designers. We can
venture to conclude that the discipline does not determine the organisational form of
either the design team or the design project.

Drivers
Further, we notice two types of projects in the cases. One project type is a one-off
product, which must conform to a pre-established list of requirements. The other project
type is the development of a product to be marketed to a mass audience. We can call
these client-driven and market-driven projects, respectively. Figure 3 shows the cases
arrayed in a matrix according to these two project environment dimensions. We will
contend that these dimensions give us a much more reliable indication of how project
management considerations are typically applied in design projects.
Murket-
driven
II6: CureliIt
II,: Ao Chuir
IIS:Curver
Aq: 1-- Huis IIg:
Beertender
Client-
driven
A1: Westruven A:
II Building A:
Mediumuseum
Figure 4: Matrix showing cases grouped according to key environmental factors
Market-driven projects
Figure 5: Cases
The projects in this category are characterised by the fact that the design activities are
carried out by and for businesses that will market, produce and distribute the products
themselves. The project environment is therefore market oriented. In this category, we
have placed, not only all of the IDE examples, but also one of the architectural cases:
the 1-2-3 Huis., (although in this case the concept was not developed completely in-
house).
Initially, there was a great deal of uncertainty about the production costs for these
products. Uncertainties that played a role in this regard include the demand for the
product, the price that the market would bear and the manufacturing technologies that
would be required to obtain the desired relationship between price and performance.
The reaction of anonymous end users was of great importance throughout the entire

design process. Output factors of signihcance include time to market and performance
in relation to the desires and needs of the end users.
In market-oriented projects management activity tends to focus on the concept
development phase. !n all hve cases, the project began with a market analysis. !n
general, these market-driven projects are also driven by technology driven. This is well
illustrated by the case of the 1-2-3 Huis, were the product is not the design of a single
house, but a production system that allows customers to order custom-made houses
tailored to their unique needs. This project is also an example of how such projects are
often responses to social needs. The 1-2-3 Huis was intended to respond to the need
for the production of houses to replace the existing post-war stock, which no longer
meets the requirements of present day consumers for inexpensive, adaptable, and
commodious houses supplied without long waiting periods.
Dealing with these uncertainties and risks is an important part of project management.
Characteristic of this is a phased approach to the design process with clear decision
moments. Most design processes can be seen as proceeding according to following the
following sequence: analysis, idea generation, concept development, detailed design,
model building and testing, engineering, production start-up, and series production.
In some cases you can observe a structured risk analysis sometimes using standard
techniques (e.g. Failure Node 8 Effect Analysis, Design for Fabrication Analysis). Often
risks are allayed through extensive testing of prototypes. In this manner, the designers
have attempted to match product performance to user expectations, in accordance with
the business model driving the product development process.
Remarkably, the designers in this group undertook the management of the entire
product development process. Planning, estimating and monitoring seem to be seen
as core activities. by the design team. No only the costs of the design projects, but
also the costs of production and delivery were carefully analyzed and optimised by
the designers. for cases in which the budget for the design of the product proved
insufhcient., additional funds were made available - for both internal and external
projects. For the designers, the primary management goal was to optimise the return
on investment for the project as a whole.
Client-driven projects
Characteristic of client-driven projects is the fact that they are based on a brief supplied
by the client - one specihc client. The designers often also has have opinions about the
brief as well, but in general this leads only to slight changes in the brief. The clients
requirements are not limited to those indicated in the brief. Beyond these, there may
be additional design constraints such as typical project management goals as budget
and time. The client often contracts the management of the project out to a project
management consultancy hrm. The designer is, therefore, often not involved in the
management of the project, and is therefore in general less able to steer the design
project.
Client-driven projects usually involve a large number of independent parties. Different
aspects of the design are normally carried out by different hrms. Thus both the overall
complexity and the complexity of the construction phase are increased. The designer is
reduced to the status of one of the links in the supply chain to be managed by the project
management hrm. Communication between and organisation of the different parties is

more difhcult and required more management attention in client-driven projects than
in market market-driven projects, and it is more likely to leads to disagreements. To
meet with this increasing complexity., additional management capacity is usually added
to the team.
The budget for design is usually determined in advance, and is normally set as a
percentage of the total investment. This hxed budget encourages designers to ration
the time they invest in the project. The designer is, therefore, not always encouraged
to invest in a very thorough study to hnd the best resolution possible for the client's
brief. Issues such as Design for Fabrication normally fall outside the architects scope of
interest. The architects scope is negotiated anew for each new project with the client
and the other design consultants.
Conclusions
While noting that the exact form taken by project management activities is determined
by the specihcs of the individual project environments, we can venture to draw a number
of tentative conclusions regarding how the general character of the project environment
determines project management. We may begin drawing conclusions by examining the
tradeoffs most frequently made between the different management factors (time, cost,
and performance) in market- and client-driven projects.
In market-driven projects budget overruns are not always considered negative project
results. On the contrary, additional expenditures seem to be readily accepted in Rather,
in cases where other factors are more highly valued additional expenditures seem to be
readily accepted, if they lead to higher performance and therefore a higher expected
return. Often it is time, specihcally time to market., that seems to dominate here.
Performance and budget (i.e. development budget) may therefore be traded off against
each other relatively freely. The budget for the Beertender, for example, was expanded
several times during the project, and no budget was specihed at all in the case of the
Carver.
Time, however, seems to be the crucial constraint in market-driven projects. The internal
project manager makes a global plan for the project. This schedule seems rarely to be
extended.
Performance
People
Cost Time
Performance
People
Cost Time
Market driven Client driven
Figure 6: Comparison of tradeoffs between management factors in market- and client-driven
projects

In client-oriented projects, project success is more likely to be was measured on the bases
of compliance with previously established indicators for time, cost and performance.
At the beginning of an architectural project, the client usually provides a relatively
detailed brief. Yet it is actually other factors that seem to dominate. Time is generally
assumed to be beyond the control of the project team. Delays are accepted as a natural
part of the process. The time factor is determined primarily by external factors (i.e.
factors external to the design process or the design team), such as building permits,
and regulations), and there is little that can be done about them. Thus, although the
delivery date may slip because of these factors, such shifts are often not seen as a
particularly negative result; rather they are seen as a fact of life. Time only becomes
an important management tissue once construction has begun. However, when costs
begin to escalate, management intervenes and performance must be reduced to bring
costs back into line, as for example, when the sustainability aspects of the IDE building
were reduced. This is true not only of production costs, but of the design costs as
well. In practice, client-driven projects are budget driven, and the level of performance
achievable within the stated budget is accepted, even when this is less that then stated
in the original brief.
We can also observe differences in the organisational relationships between design
and construction. In client-driven projects we see a separation of design and project
management. In architectural projects project management is often performed by an
external project management consultancy hrm. !n market-driven organisations, there
is often an internal project manager. Thus, in IDE projects the designer is responsible,
not only for the design, but also for the management of the project.
!n client-driven environments, an additional difhculty is posed by the difference between
the client and the end user. !n the case of the !DE building, there was frequently
tension between the owner of the building (TU Vastgoed) and the users (Faculty of
Industrial Design staff). Further, while building design projects often span a period of
years, and the organisations to be accommodated continue to evolve throughout the
duration of the project. This often leads to changes in the users needs and negotiations
that lead to deviations from the originally stated brief. On the other hand, the market
determines what the expected performance should be, through market research and
product testing.
The architectural design process is more complex. There are more parties involved,
and many aspects of the design are contracted out to other parties. In some cases, the
architect will provide only the concept design, and the working out of that design in
detail and specihcations will be done by another party.
Risk management seems to be important in both market- and client-driven projects.
However, the risks receiving the most attention are different. In market-driven projects,
the most important risks to be managed are those associated with the market itself
price, and consumer demand. In client-driven projects, the most important risks are
internal project risks the client is concerned with managing the designer, and does not
share their concerns for the market (in those cases where the product will eventually be
brought to market) with the designer. The designer plays no part in market research,
but receives the hndings of this research in the form of a design brief. The emphasis
is on arriving at a previously conceived result rather than maximising performance,
production cost, or delivery time.

Risk management varies between internal and external projects as well. In internal
projects it is possible for designers to deal with risks associated with production, where
as, we can see in the external projects, particularly in the architectural cases, that no
account is taken by the designer of production risks.
Thus we see that risk management, while not being named as such, is carried out in a
more structured fashion in IDE projects than in architectural projects. In architectural
projects, risk management is conhned to project managers, whether they are either
external consultants or internal to the client organisation. This leads to a sort of
conservatism, in which meeting predictable ends is more important than maximising
performance. Innovation in architecture is, therefore, exceedingly gradual, and it tends
to be focused on aesthetic issues. In this respect it should be noted that the only
architectural project where patents were sought was the market-driven 1-2-3 Huis
project. The only market-driven project not to seek patents was the A230 Chair, were
Arhrend sought alternative means to protect their intellectual property.
From this short study it can be seen that there are many similarities as differences
between the ways in which project management as it is applied in IDE and architecture.
In summary, we can say that IDE projects are managed to meet product and investment
performance expectations, while architectural projects are managed to achieve
compliance with briefs.
More interestingly, the distinction between IDE and architecture is not always evident. The
categories of market-driven and client-driven projects are more illuminating, as are the
categories of internal and external projects. !n many ways, therefore, an !DE hrm and an
architectural hrm that are both working on client-driven projects are likely to have more
in common than would an !DE hrm and an !DE department in a large company working
exclusively on internal projects. Indeed, focusing on project-environmental factors may
enhance our ability to predict the types of project management approaches that are
required in design-build and other unusual or innovative project organisations.
References
Barnes, M. (1988). Construction Project Management. International Journal
of Project Management 6(2): 69-79.
Cooke-Davis, T. (2002) The real success factors on projects. International
Journal of Project Management 20(3): 185-190.
Halman, J. I. M. (2002) Ontwikkeling van een risicoreferentielijst voor product
innovatieprojecten. Bedrijfskunde 74(5): 35-45.
Keizer, J. A., J.-P. Vos, et al. (2005) Risks in new product development:
devising a reference tool. R&D Management 35(3): 297-309.
Kliem and Ludin (1992) The People Side of Project Management. Gower,
Aldershot.
Koutsikouri, D., A. Dainty, et al. (2006). Critical success factors for
multidisciplinary engineering projects. 22nd Annual ARCOM
Conference, Birmingham, UK, Association of Research in Construction
Management.
Lock, D. (2007) Project Management (9th edition). Aldershot, U.K., Gower.

Mbachu, J. I. C., & Vinasithamby, K. (2005). Sources of risks in construction
project development: An exploratory study. Paper presented at the
Queensland University of Technology Research Week, Brisbane,
Australia.
Meredith, J. R. and S. J. Mantel (2006). Project Management; A managerial
approach. New York, Wiley.
MMID. (2007, 20/08/2007). MMID full service design team (corporate
website). Retrieved 12/05/2008, from www.mmid.nl.
Morris, P.W.G. (1994) The Management of Projects. London, Thomas Telford.
Morris, P.W.G. (2001) Updating the Project Management Bodies of
Knowledge. Project Management Journal 32 21 30
Robbins, S. P. and D. A. DeCenzo (2004). Fundamentals of management:
essential concepts and applications. Upper Saddle River, N.J., Prentice
Hall.
Rockart, J. F. (1979). Chief executives dehne their own data needs." Harvard
Business Review 57(2): 81-93.
Roozenburg, N.F.M. & Eekels, J. (1991) Produktontwerpen, Structuur en
Methoden. Utrecht, Lemma.
Shenhar, A. J., D. Dvir, et al. (2001). Project success: A multidimensional
strategic concept. Long Range Planning 34(6): 699-725.
Stamm, B. von (2003) Managing Innovation, Design & Creativity. Wiley,
Chichester, UK.
Winch, G., A. Usmani, and Edkins, A. (1998) Towards total project quality: a
gap analysis approach. Construction Management and Economics 16:
193-207.
Winch, G. (2002) Managing construction projects : an information processing
approach. Oxford, Black

59

IDE+A
Social Complexity

61
5 Social complexity in design collaboration
Prof. dr. P.G. Badke-Schaub
Professor Design Theory and Methodology
Faculty of Industrial Design Engineering
Abstract
This chapter focuses on the causes and consequences of enhanced complexity of de-
sign activities by the social context. The eight design projects, which were used as
stimulating material, were analysed towards the variables which contributed to the so-
cial context. All interviewees discussed collaboration between different stakeholders as
one of the main ambiguous issues in the design process. In the paper the challenges of
the three problems prevalent in most projects are analysed in further detail: unshared
or contradictory goals between different stakeholders involved in the process, the need
for cross-disciplinary communication and the uniqueness of the projects. Finally, two
concepts are presented and further detailed in how they may provide opportunities of
inhuencing these complex social processes in a desired direction.
Keywords: coordination, communication, contradictory goals, team mental models
Introduction
In the past the designer was a creative genius, a creator and the artist behind the prod-
uct. Today, its common to state that design is a social process (e.g. Bucciarelli, 1994)
since many design projects are far too complex for individual designers. Technological
advances have led to increasing specialisation, with the consequence of an increasing
need for teamwork in the context of multi-disciplinarity. The former sequential prod-
uct development processes within one organisation have changed to concurrent engi-
neering processes, often involving several organisations. Thus, the designer is often a
member of a multi-disciplinary product development team including disciplines such as
marketing and mechanics, software, product control, and more.
The same is true for architects, whose work includes collaboration with disciplines such
as statics, installation, construction, etc., each of them contributing their particular
domain-specihc expertise.
At hrst glance this situation seems to be highly effective because the only option for
coping with these complexities is to integrate the expertise and knowledge of different
disciplines. This synergistic effect is especially emphasised in the theoretical framework
of social cognition:
Knowledge is commonly socially constructed, through collaborative efforts towards
shared objectives or by dialogues and challenges brought about by differences in per-
sons perspectives. (Pea, 1993, p.48)
However, multidisciplinary teams also run the risk of a variety of problems, to name
only two aspects of any collaboration in a project team across disciplines and organisa-
tions:
- different language: different disciplines use different vocabularies; the
same word can indicate different phenomena (for example, the word
function) or different words can refer to the same feature;

62 Social complexity - P.G. Badke-Schaub
- different background: members of multidisciplinary teams often run into
conhicts due to their lack of a shared mental model (Badke-Schaub et al.,
2007).
Considering only these two aspects it becomes obvious that the social context adds ad-
ditional complexity and thus requires from the designer to encompass a broader set of
capabilities apart from hisfher domain-specihc role in design projects.
In the following chapter the main challenges which constitute social complexity in de-
sign collaboration will be outlined and in the third part (chapter 3) some theoretical
analyses explain the aspects which are most important when considering how to cope
with these challenges successfully.
1 Challenges
Working with other professionals in a project team requires focusing on the accomplish-
ment of the task at hand while embedded in a complex social process. The challenges
resulting from this situation are of various kinds; the three challenges discussed here
are taken from the interviews of four architects and four industrial designers (from 8
different projects) who were involved in the design of well-known Dutch buildings and
products (8 different projects). These projects were chosen because they represent
design success stories in architecture and product development.
1.1 Unshared and contradictory goals
According to the dehnition by Katzenbach and Smith (1993), à team is a small number
of people with complementary skills who are committed to a common purpose, per-
formance goals, and approach for which they hold themselves mutually accountable.
Furthermore, team members strive for a common goal. However, these characteristics
do not hold for teams we see more and more working in the globalised world, such as
virtual teams, geographically dispersed teams, etc.. These teams may have a common
goal in the broader sense such as in project teams where each discipline brings in its
own, often hidden agenda. For example, one of the architects interviewed describes
the conhict between the architect and the construction company as a continuous goal
conhict: whereas the architect is interested in the highest quality, the construction
company gains the most with a project that is as cheap as possible; in order to save
money the construction company does not always stick to the plans. The same problem
was also mentioned by another architect: The architect wants to create beautiful things
where the building contractor wants the building to be cheap.
Obviously, the need to cope with different and often contradictory goals is not only a
part of the task process but also of the social process. The designer has to balance
between individual and domain-related goals and project goals, or as Bucciarelli (1994,
p.1S9) states: `the (design) process is necessarily social and requires participants to
negotiate their differences and construct meaning through direct, and preferably face-
to-face exchange. Thus, collaboration can be successful if the interaction focuses on
joint objectives.
1.2 Cross-disciplinary communication
The main contribution to the overall success of a complex design project is communica-
tion between the various parties involved, including the user and the client. However,
the different individual backgrounds of the parties, visible as an amalgamation of differ-
ent expertise, perspectives, values and goals, frequently poses difhculties for adequate
communication (see Figure 1). As a direct consequence of these difhculties designers
often refer to, and budget for, the time needed to come to a decision.

Coordination
Knowledge
nformation
transfer
Goal/task
accomplishment
Communication
Shared
mental model
ndividual
background
Collaboration
Figure 1: Nodel of inhuences on collaboration.
The main aim of communication is the exchange of information, which in complex
projects usually leads to an information overload for the individual professional. Hence,
the integration of information is necessary in order to transfer information into knowl-
edge. Some projects try to enable this process by using a sophisticated documentation
system. The structure of such a system has to be transparent for all parties involved
and the vocabularies used need to be understood by all disciplines in the same way.
Furthermore, a documentation system should clearly describe the decisions that have
been taken and why. The integrated knowledge should be more or less shared by all of
the team members.
1.3 Structured procedures
Structured procedures are, in the eyes of many designers, too structured. One designer
related that in his company there is a standard procedure for handling a project. This
procedure is similar to what he learned at university. They teach you to follow the per-
fect process, but in reality it does not work that way. Its neither preferable nor workable
because each project needs its own approach. Every project is one of a kind; we start
by asking ourselves what this project needs. And from there we start the project.
Although there are several methods which support project work there is a rather low
rate of acceptance by professionals. Contrary to these structured approaches, brain-
storming is widely accepted and used in daily work; however this method is not always
used as prescribed by the inventor (Osborne, 1953) but more as a tool for unstructured
discussions.
2 Essentials of social complexity in design
!n this chapter two basic concepts are suggested which require further research if we
want to understand and support multi-disciplinary design collaboration: coordination
and team mental models.
2.1 Coordination
One of the most important requirements of design projects - which has also been
conhrmed by the designers interviewed - is the how of information which has to be
organised and distributed effectively in terms of team, time and space. As projects

always have to cope with the intersections between disciplines it is necessary to make
sure that the individual contributions are in line with the various interconnections. The
more interconnections the more coordination is necessary, Thus, adequate coordination
is a precondition for precisely aligning individual contributions to the team as well as
contributions between teams. The way to coordinate may be different, depending on
the use of tools, channels and media (see Figure 2).
Coordination
Tools nformation Channels Media
Location
Time Team
Allocation of
- tasks
- roles
- responsibility
Coordination of
- communication
- cultural aspects
Timing of
- actions
- deliverables
- group dynamic
Figure 2: Different helds and elements of coordination.
As indicated in Figure 2, an important element of coordination determines the team
structure by allocating tasks, roles and responsibilities, in which the coordination of
roles is a main factor inhuencing team performance. A clear allocation of roles pro-
vides the group with a transparent group structure and a clear allocation of tasks and
responsibilities according to the preferences and competencies of the team members
is a prerequisite of successful team performance (Stemphe, Hbner 8 Badke-Schaub
(2001). In this way the team members develop a shared team mental model and may
be at the same time a transactive memory.
A further major coordination issue is the allocation of time and careful scheduling so
that enough time is available to accomplish tasks and for group development. A clear
schedule clarihes the timing of actions and milestones for deliverables.
The coordination of the interaction between different locations becomes more important
with increasing globalisation and hence the increasing virtualisation of collaboration.
Another important topic with regard to geographically distant collaboration concerns
cultural differences and their impact on different aspects of design work.
Empirical studies reveal that the more team members coordinate their contributions,
in relation to task and process, the better they perform (Gurtner, 2003). Certainly, the
need for coordination depends on the complexity of the task, the number of different
parties involved and the degree of interconnectivity between the parties; the more in-
terdependencies the more coordination is needed.
Groups working together for a longer period of time develop a common history and as
a consequence implicit coordination, especially for well-known and structured parts of

their work. However, when collaboration in teams is begun, coordination is communi-
cated explicitly and this creates a common ground (Clark & Brennan, 1991), which
rehects shared information and shared beliefs.
There is also empirical evidence that teams tend to avoid coordination or postpone the
start of coordination activities until it becomes obvious that the current muddling-through
strategy is not successful (Hackman, Brousseau & Weiss, 1976). Gersick (1988, 1989)
derived from an analysis of project teams that halfway through the project timeline a
transition phase occurs, characterised by a sudden change of strategies. Resources can
be wasted as the consequence of omitted coordination, but in addition poor decisions
can be made, which affects the result of the whole project in a negative way.
2.1.1 Regular face-to-face meetings
Although all interviewees reported that during the projects information was also shared
by computer-mediated communication, all participants underlined the need for regular
face-to-face meetings. One designer explained that he does not use the phone or email
that much, because he wants to see people and read their body language: The scale
of a project is not relevant to the way of communicating. Designing concerns tangible
items. It is therefore also important that you see each other face to face and with a
drawing or a mock-up.
The importance of face-to-face meetings is conhrmed by hndings that face-to-face
meetings support the trust-building process (Tang & Isaacs 1993). Furthermore, the
possibility of a shared view of sketches, models and mock-ups is a basic part of most
face-to-face meetings in design teams. This tangible aspect is mainly stressed by ar-
chitects: face-to-face meetings are needed because especially at the beginning of a
project - a lot of things are being specihed.
Some of the designers interviewed distinguished between two kinds of meetings; on
the one hand, the more structured meetings usually chaired by the project leader, and
on the other, unstructured brainstorm sessions. Brainstorming seems to be the method
used in all projects and for different purposes, such as to identify various aims.
2.1.2 Reection on task and social context
You always have to stay critical about why you are doing something. This goes for the
building but also for the management.
This statement by a designer emphasises `rehection' as an important part of the design
process.
There aren't many empirical studies analysing rehection as part of human cognition.
Gurtner et al. (2007) showed that rehection enhances performance, mediated by struc-
tured communication and the similarity of mental models. Another interesting result
refers to the differences between individuals and groups in terms of rehexivity: indi-
vidual rehexivity was superior to group rehexivity because group rehexivity increased
the discussion of very general strategies. Obviously rehection enhances performance as
long as it focuses on task-specihc strategies.
Currently there are no empirical studies that investigate the role of rehection on the
social context and its inhuence on performance.
2.2 Team Mental Models
Mental models are internal representations that humans build in order to understand,
predict and act in the world (Craik, 1943). There are different assumptions about the
patterns of representations; however, researchers agree on two basic types of models
(Cooke, Salas, Cannon-Bowers, & Stout, 2000; Klimoski & Mohammed, 1994; Rentsch
& Hall, 1994): those concerned with the task and those concerned with the team. The

task mental models encompass all aspects related to the execution of the task, while
the team mental model covers all representations related to the team and the team
members who are essential to working together. Team mental models are generally
dehned as the organised understanding of relevant knowledge that is shared by team
members (Cannon-Bowers, Salas, & Converse, 1993; Klimoski & Mohammed, 1994).
It has been shown that teams sharing a common understanding of the task, the team,
and the situation perform better (e.g. Lim & Klein, 2006; Mathieu, Heffner, Goodwin,
Cannon-Bowers, & Salas, 2005).
More precisely, one designer interviewed describes his view of the kind of team-related
knowledge that is required: `You have to be a group; you have to come together as a
team. You have to know who does what. You have to know what you can expect from
each other. And how you will work together. What are the ins and outs of the project?
What is the project's focus? !t is very important to short circuit it with everybody in-
volved.
Here two main aspects contributing to successful projects are specihed:
- team cohesion: you have to come together as a team
- shared team mental model: you have to know what you can expect
from each other.
Knowledge about the team includes knowledge about competencies, roles, tasks and
responsibilities of the team members: who is responsible for which (sub-)tasks, what
kind of competencies, capabilities, strengths and weaknesses characterise a specihc
person? This kind of knowledge is relevant to successful team performance.
!n an empirical study Bierhals et al. (2007) found a signihcant correlation between a
high level of shared team mental models and problem-solving performance but not
between shared task mental models and performance, which underlines the relevance
of the social context for successful task accomplishment.
3 Conclusions
This paper aimed to focus on the signihcance of social complexity to collaboration in
design. Social complexity relates to the social context that a project team works in and
thus inhuences the decision-making process of the project work.
Interviews with designers responsible for successful Dutch projects in architecture and
product development formed the background for the analysis presented here of the
impact of social complexity on collaboration in design teams. Although all projects were
successful the interviewees also reported some restrictive factors, such as contradicting
goals between the different parties involved. Finally, two major theoretical concepts
(coordination, team mental models) are discussed which provide further ideas to suc-
cessfully dealing with complex design projects in social context.
References
Badke-Schaub, P., Neumann, A., Lauche, K., & Mohammed, S. (2007). Mental
models in design teams: A valid approach to performance in design
collaboration? Co-Design, 3, +-19.
Bierhals, R., Schuster, I., Kohler, P., Badke-Schaub, P. (2007). Shared mental
models - linking team cognition and performance. Co-Design, 3,
75-94.
Bucciarelli, L.L. (1994). Designing Engineers. Boston: MIT Press.
Cannon-Bowers, J. A., Salas, E., & Converse, S. (1993). Shared mental models
in expert team decision making. In N. J. Castellan, Jr. (ed.), Individual
and group decision making: Current issues (pp. 221-246). Hillsdale,

NJ: Lawrence Erlbaum Associates.
Clark, H. H., & Brennan, S. E. (1991). Grounding in communication. In L. B. Resnick, J.
M. Levine, & S. D. Teasley (eds.), Perspectives on socially shared
cognition (pp. 127-149). Washington, DC, USA: APA Books.
Cooke, N. J., Salas, E., Cannon-Bowers, J. A., & Stout, R. J. (2000). Measuring
team knowledge. Human Factors, 42, 151-173.
Craik, K. J. W. (1943). The nature of explanation. Cambridge, UK: Cambridge
University Press.
Gersick, C.J.G. (1988). Time and transition in work teams: Toward a new
model of group development. Academy of Management Journal, 31,
9-41.
Gersick, C.J.G. (1989). Marking time: predictable transitions in task groups.
Academy of Management Journal, 32, 274-309.
Gurtner, A., Tschan, F., Semmer, N.K. & Ngele, C. (2007). Getting groups to
develop good strategies: Effects of rehexivity interventions on team
process, team performance, and shared mental models. Organizational
Behavior and Human Decision Processes, 102, 127-142.
Hackman, R.J., Brousseau, K.R. & Weiss, J.A. (1976). The interaction of task
design and group performance strategies in determining group
effectiveness. Organizational Behavior and Human Performance, 16,
350-365.
Katzenbach, J.R. & Smith, D.K. (1993). The Wisdom of Teams: Creating the
High-performance Organization. Boston: Harvard Business School.
Klimoski, R., & Mohammed, S. (1994). Team Mental Model - Construct or
Metaphor. Journal of Management, 20, 403-437.
Lim, B.-C., & Klein, K. J. (2006). Team mental models and team performance:
a held study of the effects of team mental model similarity and
accuracy. Journal of Organizational Behavior, 27, 403-418.
Mathieu, J. E., Heffner, T. S., Goodwin, G. F., Cannon-Bowers, J. A., & Salas, E.
(200S). Scaling the quality of teammates' mental models: equihnality
and normative comparisons. Journal of Organizational Behavior, 26,
37-56.
Mohammed, S., & Dumville, B. C. (2001). Team mental models in a team
knowledge framework: Expanding theory and measurement across
disciplinary boundaries. Journal of Organizational Behavior, 22,
89-106.
Osborn, A.F. (1953) (rev. 1957, 1963). Applied Imagination: Principles and
Procedures of Creative Problem-Solving. New York: Charles Scribners
Sons.
Pea, R. (1993) Practices of distributed intelligence and designs for education.
In G. Salomon (Ed.), Distributed cognitions: Psychological and
educational considerations. Cambridge University Press: Cambridge.
Stemphe, J., Hbner, O. 8 Badke-Schaub, P. (2001). A functional theory of
task role distribution in work groups. Group Processes and Intergroup
Relations, 4, 138-159
Tang, J. C. 8 !saacs, E. (1993). Why do users like video?: studies of
multimedia-supported collaboration. Computer Supported Cooperative
Work, 1, 163-196.

IDE+A
Decision making

69
6 A decision-based design approach
Dr. ir. P.P. van Loon
1
, ir. R. Binnekamp, ir. J. Burger
1
Associate professor Design and Decision Systems
Faculty of Architecture
Introduction
!n the held of engineering design (industrial design, architecture, as well as urban
design) a specihc design perspective, known as decision-based design, has been
explored for several years now.
Over the past decades, design theory research has taken several twists and turns,
as computational tools became the standard for how engineers of all disciplines did
design. In an early National Science Foundation Workshop report (Newsome et al.,
1989), research was categorised into topical areas focused on the design process
that included the computational modelling; the cognitive and social aspects; the
representations and environments; the analysis tools including optimisation and the
design for, such as for manufacturing. At that time, the NSF programme was called
Design Theory and Methodology and consisted of three components that essentially
captured these hve topical areas: the hrst, `Scientihcally Sound Theories of Design',
established a home for proposals that were directed at creating the scientihc basis for
the design process. The second, Foundation for Design Environments, was aimed at
advancing the understanding of fundamental generic principles that could be used and
understood across engineering domains. The third, Design Processes, was focused on
the how and why of the design process, including early work on life-cycle concepts and
concurrent design (Durham, 2006).
At this point, you might ask, `So what is new?' The tools certainly have advanced over
the years, from early computer-aided design (CAD) through solid modelling capability.
The introduction of virtual reality, computer integration engineering, and collaborative
and distributed design processes created demands on the community to focus on how
decisions were made, under what conditions and to what purpose. Decision-based
design became a major thrust for the research community, with the issues of uncertainty
and predictive modelling capability becoming the foci. As with any science, the theories
must be put forward, tested for consistency and completeness, and then incorporated
(or not) into the framework of the science. This is true, too, for engineering design, if it
is to become more than just an ad hoc, intuitive process that is domain-specihc.
During the late 1990s, members of the engineering design research community articulated
a growing recognition that decisions are a fundamental construct in engineering design.
This position, and its premise that the study of how engineering designers should make
choices during the design, represented the foundation of an emerging perspective on
design theory called decision-based design (DBD). DBD provides a framework within
which the design research community could conceive, articulate, verify and promote
theories of design beyond the traditional problem-solving view. As dehned by Chen et
al. (2006):
Decision-based design (DBD) is an approach to engineering design that recognizes the
substantial role that decisions play in design and in other engineering activities, largely
characterized by the ambiguity, uncertainty, risk, and trade-offs. Through the rigorous
application of mathematical principles, DBD seeks to improve the degree to which these
activities are performed and taught as rational, that is, self-consistent processes.

70
At TU Delft, the Open Design Research Group has been working in the held of decision-
based design for some years. The groups focus is on a collaborative approach to
architecture, urban planning, and project management. It offers concepts and methods
to combine technical and social optimisation into one integrated design process
(Binnekamp et al., 2006).
IDE+A Case Study Analysis, IDE+A Workgroup TU Delft (2008), pp. 4-5:
Introduction - Case 1: Westraven Utrecht - Company: Cepezed
Could you give a short introduction of your project?
The Westraven project consisted of an ofhce building that had to be
renovated. The building was designed by Jan Lucas (of Lucas & Niemeijer)
in 1975. The building stands at a particular spot that is known as the
bellybutton of the Netherlands. Jan Pesman has looked at the exact middle
point of the Netherlands but this was not the correct location. However, the
location is characterised by the crossing of waterways and roads such as
the Amsterdam-Rhine canal and the A2 highway. This location is therefore
very precious to an organisation such as Rijkswaterstaat (the Ministry for
Transport, Public Works and Water Management). There was a contest, in
the form of a European tender. The requirements were an enlargement
of the working space and that surrounding buildings should have glass
windows which could be opened. In order to open the windows they
developed a double faade system with a semitransparent fabric in order
to hlter light and reduce wind speed.
Social Complexity in Collaboration - Case 1: Westraven Utrecht -
Company: Cepezed
Describe the structure, mutual communication and the relationships during
the collaboration by those involved in the project.
What was your role in this project?
He was the architect. According to Jan, he is at the top of the food chain. You
generate ideas, which you discuss with the design team. These proposals
are then taken to the customer (in this case, the Rijksgebouwendienst
(Government Buildings Agency)). As the architect, you are president of the
design team. This means that you also have a vote about which external
advisors take part in the project.
Did this change during the project?
In general, the role of the architect is reduced little by little. Jan Pesman
says that architects revolt against this. He sees more and more responsibility
being placed in the hands of the construction company. The architect has
to make a nice drawing and the construction company executes the plans
the way they like it. The danger here is that the construction company aims
to make as much hnancial gain as possible by making the project as cheap
as possible. Jan advises making a design hnal, down to the smallest detail,
and then going to a construction company.
How many parties and people were involved in the project?
A total of seven parties were involved.
Who were they?
1. The client / user: Rijksmonumentenzorg (Netherlands Department for
Conservation) and Rijkswaterstaat.
2. The design team: architect (Chair of the design team), Construction
company, structural and installation technology
3. External advisors or specialists
4. Cost management
Could you sketch the organisation's structure?
Only the architect was from Cepezed. All of the other parties involved were
from outside the Cepezed organisation.
A decision-based design approach - P.P. van Loon, R. Binnekamp and J. Burger

71
What was the balance between the genders?
Only a few women were involved in this project.
How many people worked on this project?
A maximum of 200 people worked on this project. There is the project
manager, who represents a group of 30 people who in turn represent the
users during the project; there are 15 engineers, 4 people in executive
management, and the architect.
Were there changes in the number of parties and people that were involved
during the project?
The number of people did not change, due to the European tender.
Did these people came from inside Cepezed or were they external?
Only the architect came from Cepezed. The rest are from outside the
organisation.
Have you worked before in this type of collaboration?
Yes, this is common.
Was there a more intensive collaboration between some parties?
The architect works in close collaboration with the advisors. This is better
for developing ideas.
Which forms of communication did you use during the project?
It is important to have regular meetings in order to develop solutions. In
addition, there was a level of technical executive control.
How did you choose which method to use, and when?
In big projects you have specialised agencies manage the executive
functions.
How was communication organised?
All decisions were recorded in project reports. Because of the complexity
of this project they set up a hle sharing service. This is a database in which
all of the information regarding the project is collected. There are written
reports, drawings, and the latest status reports. Jan Pesman also stresses
the importance of meetings, saying you need interaction. In a meeting
quick sketches can be shown and discussed. This is much quicker than via
e-mail.
Was there an initiation meeting?
Yes, there was.
Did the absence of specihc persons inhuence the advancement of the
project?
Because Cepezed had only technical executive control they were not always
present at the construction site. But sometimes the construction company
will not stick to the plan, in order to save money. So they have to carry out
thorough checks of the work.
Did the innovative nature of the project (on a international scale) have an
inhuence on the means of communication?
The innovative nature of the project led to more communication between
the advisors and the architect. More meetings and more sketches lead to
more solutions.
1 Decision making in an architectural design process
Designers working in architecture are confronted more and more frequently with huge
solution spaces, hlled with hundreds of alternative combinations of possible sub-solutions
supplied by a host of specialists. As a result of this, their quest for the optimum design
tends increasingly to run aground in a combinational explosion. At such moments,
there are too many options, too many opinions and too many alternatives.
If designers bring in the expertise of specialists to reduce the size of the solution space,

72
their position becomes even more difhcult. Specialists only select those combinations of
options that lie within their own discipline. When the designers go on to combine these
selections, they hnd that the specialists were not sufhciently capable of independently
assessing which combination was important for the whole, and which was the mostly
likely to meet the goals of all involved.
Many designers attempt to deal with this dilemma of too many combinations at the
start and too few options after selection by setting up a broadly based design team.
On such a team, the designers work with specialists, jointly exploring the solution space
and determining the best combination of sub-solutions. Unfortunately, this approach
often also tends to run aground, when the client and users fail to approve the result
produced by the professional design team. They might hnd it `too ambitious', ùnfeasible'
or feel that it `fails to cater adequately to the social context'.
To prevent this kind of rejection, designers often enlist the aid of process experts, asking
them to devise a decision-making process for the team. This process sets out what has
to be produced, when, and who should decide what: hrst produce and approve this
sub-design, then the next, etc. This enables the team to work towards a result with
some degree of certainty, but also entails the risk that a series of sub-optimum design
decisions will lead to a sub-optimum design result.
IDE+A Case Study Analysis, IDE+A Workgroup TU Delft (2008), p. 6:
Decision making - Case 1: Westraven Utrecht - Company:
Cepezed
How are decisions made during the process? Who is involved in decision
making?
The architect makes decisions regarding the design. He is advised by his
team of advisors. He takes their advice into account when drawing up the
design. Then he takes the design to the client. The client normally follows
the advice of the architect.
Who makes the hnal decision?
The client.
What factors play a role?
The requirements of the user.
Why do these factors play a role?
The users are the ones who have to use the building. Their wishes and
demands have to be represented in the design.
Could you give your point of view (related to the project) about these non-
scientihc decision-making methods?
Non-scientihc decision-making criteria:
1. `First come, hrst served' (other ideas are unlikely to be offered).
2. The loudest voice is the one the gets heard (others are too modest)
3. Power (the boss decides)
4. Authority (the dominant person decides)
5. Everyone thinks his own owl is a falcon (for example, the boss son)
6. Anxiety (risky decisions are taken)
7. Haste (the proposal that looks like it will take the shortest time to
complete is chosen)
8. Tenderfoot (trying to get in the good graces of those making the
proposal)
9. Fast talking (the best presentation wins)
10. Last-best (the last presentation is the one that is best remembered)
Types of leadership:
1. Laissez-faire (from the French, meaning do what you want to do, I will
judge the result)
2. Catalytic (stimulating)

73
3. Participatory (collaborative)
4. Directive (little room for those who are carrying out the work)
Laissez-faire is not applicable. If they do not do something, nothing will
be done. Stimulating by all means. The engineers who are involved are
challenged in the hrst phase of the project. `Participatory' also applies to
the organisation. The production team is a connected whole. However, the
execution is not up for discussion. When the plans are ready the execution
is carried out in a directive fashion.
To what extent is the process of decision making put on paper? What is
recorded?
All decisions are recorded in reports.
Did you agree with all of the decisions that were made?
No, not with all of the decisions, such as the colours that were chosen by
the interior designer.
How were you involved when decisions were made?
The architect leads the building team. If a decision had to be made the
architect, as the leader, had the ability to really make decisions. But as
always it is a team effort. So advisors give advice and that is presented to
the client.
Was there an alternative when making decisions (Plan A andfor Plan B)?
There is always an alternative when making decisions. The architect always
has a backup plan, but he really tries to go for Plan A.
2 The Sjoelbak game: a decision-based design situation explained
A designer who had been given a very complex design commission wanted to know
before he started whether the client realised what he was getting into. He therefore
invited him to his design studio so that he could become acquainted with the fascinating
but often difhcult work of design. The client was entirely unfamiliar with the design
profession, and had never commissioned a designer before.
The designer had come up with an unusual way of immediately and tangibly illustrating
the design profession. He had set up a Sjoelbak (shufheboard) game right in the middle
of his studio. Sjoelbak originated in Friesland, a province in the north of the Netherlands,
where it is a popular family game. It consists of a long rectangular wooden box (the
shufheboard itself), and wooden disks, similar to ice hockey pucks, known in the old
Frisian language as sjoelen. Players slide the disks from one end of the box to the
other with their hand, through a number of holes (originally four), each of which has
a different score. The person who scores the highest number of points with a given
number of pucks is the winner (Figure 1(a)).
The designer began playing and pushed a number of pucks in the direction of the holes.
Then he stopped. The state of the game at that point was just right for explaining the
complexity of the design commission (Figure 1(b)).
1
3
5
4
2
Des|gne|
1
3
5
4
2
(a) (b)
Figure 1: The hrst game

74
`Let's say,' said the designer, `that each puck is one of your design requirements. !f a puck
is in the hole, then a certain percentage of your requirement has been met, depending
on the hole's score. However, if the puck is still on the board, your requirement has not
been met. The pucks that still have to be played are all the requirements we have yet
to look at. Because you cant get all the pucks into the holes at once and because not
all the pucks will ht in the hole with the highest score, !'ll have to play the game several
times to get the best score - or the best combination of requirements. How often will
! have to do this? When will ! have achieved the best combination? Who can say? !s it
indeed possible? And how can ! improve my chances?'
The client studied the situation carefully, remembering the game from his youth. Why
have you made things so difhcult by trying to use so many holes at once? !f you leave
a few open and close the rest it will be a lot easier, you'll be hnished a lot sooner, and
!'ll get my design more quickly'.
Of course the designer had expected him to say this. He called two experts from his
ofhce into the studio. They played the game together, on a shufheboard the designer
had prepared beforehand so that three people could play simultaneously. The experts
were told to send the pucks in a certain direction by pushing the sides of the board in
(Figure 2(a)).
The designer now pushed a number of pucks towards the holes, while the experts
pushed the sides so that the pucks landed in their preferred holes. After a few more
pucks had been played, the game nicely illustrated how reducing the number of holes
with the aid of experts could inhuence the outcome (Figure 2(b)).
1
3
5
4
2
Des|gne|
1
3
5
4
2
Expe| e }
Expe| e }
1
2
(a) (b)
Figure 2: The second game
If we dont use all the holes, explained the designer, there are fewer possible
combinations so there's more chance we'll get a quick result. As you can see, the
experts did a great job helping to reduce the number of holes that were in play. But
there's a big risk here. Did the experts send the pucks in the right direction? Was it
acceptable to ditch a number of requirements? !s the score really bad in the area the
experts weren't directing the pucks to?'
The client looked desperate. Listen, youre well known as a good and, above all, clever
designer. That's why ! chose you. So how exactly are you going to deal with this?'
Of course the designer had hoped the client would say this. To illustrate his planned
approach, he adjusted the shufheboard further so that more players could play at the
same time. The designer had a few more people from his ofhce come over to the studio.
They all, including the client, stood around the board. It had been adjusted so that the
direction of the pucks would be affected in a variety of ways. The players at the sides of
the board could steer the pucks by pushing or pulling at the sides, which were hinged
(Figure 3(a)).
After the designer had pushed a number of pucks towards the holes, between the now
hexible sides, the situation was such that he could explain precisely how he intended to
approach the design commission (Figure 3(b)).

75
1
3
5
4
2
Des|gne|
1
3
5
4
2
Ee e e e e
1 2 3 4 5
Ee e e e e
7 8 10

(a) (b)
Figure 3: The third game
Look, he said, now everyone has affected the outcome, but we dont know who had
what effect. We do know however that everyone had to take everyone else into account.
We were all connected by the box, the hexible sides, the pucks and the holes. The
outcome was not known at the outset. No one was able to steer the pucks entirely as
he wanted. If we play the game like this a few times, well get better at it, and be able
to arrange the sides properly, so that we eventually arrive at something that rehects our
group preferences. And we can also arrange to leave open the possibility of pushing the
pucks into certain holes at a later stage.
The client now fully realised what he had embarked upon. He was particularly pleased
with the idea that he could involve people other than professional designers directly in
the process.
3 Decision-based design by means of the combination of sub-solutions
The design situation illustrated above by the Sjoelbak game is known as design by means
of the combination of sub-solutions. This method was developed in the late 1960s and
early 1970s in the framework of what was known at the time as systematic design or
`methodical design'. At that time the basic ideas about and techniques for the rational,
step-by-step combination of sub-solutions into one design were developed. Initially
these ideas and techniques were intended for the individual designer. After they had
been successfully applied in practice by many designers they came to be used by design
teams too, and subsequently in the design processes of entire organisations. At the
time, problems were sometimes encountered by the last two applications. There were
designers who were beginning to see that individually oriented combination techniques
could not simply be applied to design teams and design organisations, especially not
in complex and large-scale projects, such as large buildings, residential areas, cities,
regions and trafhc systems. These projects often ran aground in the combination phase,
due to the impossibility of bringing together the large number of parties involved, all
with their own design goals and design ideas, and of incorporating the ideas into one
whole, one general design. They therefore began to seek combination techniques which
were aimed specihcally at large groups of people designing together and producing
large numbers of sub-solutions.
Systematic design emanated largely from classic conceptions of rational and modern
design. While the systems approach, which originated in the thirties, had a strong
inhuence in this respect, disciplines such as mathematics and logic and helds like
decision theory and management science also helped shape the combination method,
albeit to a lesser degree.

76
ne c|oss|c
ccncep|cn cf oes|gn
ne sysems
opp|cocn
Monemo|cs ono
|cg|c
Dec|s|cn nec|y ono
monogemen sc|ence
Menco|co| oes|gn: ccmc|no|cn cf suc-sc|u|cns
Figure 4: Approaches to methodical design
The classic conception developed in the 1950s when the search for new design
methods was underway. Designers and design theorists had seen that commissions
were becoming more complex and that existing design methods were proving to be
inadequate and unable to cope with this complexity. Personal intuition and experience
were important features of these methods (Foque, 197S p. 9). However, the users of
products, whether these were houses, residential areas or trafhc infrastructures, were
beginning to ask why these products had been designed in a particular way. People
wanted to discuss the effectiveness and the effects of new products before they were
made. The designers personal vision was no longer enough. People wanted a rational
and even a scientihc foundation for design solutions.
At the time the systems approach and cybernetics had a strong inhuence on the
development of the new design methods which were needed. Jones (1970), Foque
(1975) and Tzonis (1982) wrote a great deal about this. A whole school emerged
around what is now known as systematic design. Even today there is interest in these
views and they are widely propagated in the framework of developments in computer-
aided design and the role of information systems in design processes, including those
used for complex design commissions. Designers are slowly beginning to realise that
the computer has created new techniques for which there is no foundation in the
classic conceptions of design. As Mitchell (1990 p. 13) put it: We must embrace the
possibilities of design that have ambiguous and unstable structural descriptions. He
goes on to say that we can no longer use only the stable, universal design rules of the
1950s and 1960s, on which computer-aided design is still often based.
Foque (197S) mentions two aspects in connection with classic systematic design: the
form-function dichotomy and goal orientation. In the 1950s, when new design methods
were being developed, there was a shift from ontological thinking to functional thinking.
It was felt that meanings are not immutable and exclusive entities that reside within
things and can be discovered by the creative force of an exploring subject, but are,
on the contrary, functions within a given context' (Foque, 197S pp. 16-17). Function
must determine form. The design process must begin with an analysis of functions and
then move to a synthesis of appropriate forms (sub-solutions). The systems approach
provided a conceptual framework on the basis of which all manner of systematic and
objective analysis and synthesis techniques could be developed. The integration of
cultural norms and values into design, which had dominated thinking on design until
then, faded into the background. The systems approach placed the goal orientation of
design activities in the foreground. From then on all design considerations and goals
had to be clearly and logically dehned. However, this requirement was never properly
put into effect.

77
4 A general phase model for the combination of sub-solutions
The literature contains many general phase models for the division of the design
process based on the method of combining sub-solutions. Roozenburg and Eekels
(1991 pp. 92-96) provide an overview of these models in the held of product design.
Jones (1970) and Hamel (1990) do the same for architectural design. McLoughlin
(1969), Chadwick (1971) and Faludi (1973) provide overviews for the held of urban and
regional planning.
Roozenburg's and Eekels' (1991 p. 96) conclusion, with respect to their held of expertise,
would concur that these models all have the same structure, which is simply presented
in a different manner in each case. Figure 5 shows this structure in a form which is
suitable for our purposes.
This general phase model shows that the cycle of formulating and combining sub-
solutions (the divergence-convergence cycle) may take place many times during a
design process. In Figure 5 it takes place twice: from the outset, up to and including
the possible solution, and from then until the hnal solution.
Figure 5: The general phase model of the combination process
The general model does not show who determines the division of the search space and
the direction of the search pathway. This would require the expansion of the model to
include its decision-making environment.
IDE+A Case Study Analysis, IDE+A Workgroup TU Delft (2008), p. 11:
Interview with Fons Verheijen and Krijn Tabbers
Fons did not have time to do the whole interview. Therefore he contacted
Krijn, who worked on the project daily, to do the interview. Before the
interview Fons did have time to give us his view of the differences between
architecture and industrial design in the approaches to design. He is a
teacher at TU Delfts Faculty of Architecture but has also tutored to a few
students in Industrial Design.
- There is a big scale difference. The scale is much bigger in
architecture.
- !ndustrial Design requires much more training in protocols: this is
how you do it (hrst analysis, then you make a matrix with pros
and cons, and then you have your answer)

78
- Architecture puts less focus on analysis but it is still needed in
order to hll the mind with data
- The design process in architecture is much more intuitive
- Along the way you slowly discover the rules for the design,
rules you have to obey, everything can change
- !t is a trial and error process
- Slowly you combine what you, as an architect, want from the
project, what the materials can permit and what the projects
preconditions are. This is what you learn to cope with at the
Faculty of Architecture.
- Architecture uses sketching much earlier in the process.
- !n architecture how the building is made is very important. There
are companies that only do concepts. But according to Fons it
gets really exciting when you need the hgure out how you can
actually make what is in your mind.
Two big differences are:
- Architecture is like a equation with a hundred unknowns, you
have to learn to cope with this.
- How to cope with doubt:
- Be a daredevil
- Be skilful
- !t is important to know what is most important at any given
moment in time
- For this, intuition and experience is needed.
(In architecture this is much more necessary than in
industrial design)
- Therefore, it is very important to work with intuition instead of
with protocols.
- Architects have secret clients, in addition to the client who hired
them. These secret clients are society and the general well-being.
Unasked, architects take all of society into account, and this can
inhuence the design. Luckily, architecture is an old profession and
this is accepted. !n industrial design you have to hght a lot more
if you want to do something different.
5 Decision making in the combination process
Decision making in a design process that is based on the combination of sub-solutions
will be geared mainly to determining and restricting the number of sub-solutions and
combinations of sub-solutions. After all, this largely determines the progress and duration
of a combination process. The more sub-solutions there are, the more combinations
will be possible. Many combinations means that many evaluations and choices have to
be made. The number can be limited on the basis of the methods that the designers
themselves use to ensure that their combination process has a workable structure,
both for themselves and for each other. These methods provide a number of bases for
managing the process: the order of the combination process; the allocation of tasks
and decisions; the structure of the search space; and the laying out of the search path
and the combination strategy.
As described above, the combination process begins with a number of parallel individual
combination processes, in which the designers draw up their own plans and make
their own syntheses. This will present few decision-making problems, irrespective of
how these individual processes are carried out. At the outset, each designer has the
opportunity to work independently of the others, in terms of both content and method.
But thereafter, the process can become difhcult. Once he has made his own alternative

79
plans, each designer will want to know what the others alternatives are, in order to
select those of his own plans that give him the best chance of achieving his own goals,
in combination with the plans of the others. A plan that might in the hrst instance
function well in terms of the designers own goals could, in combination with the others
plans, create an overall situation in which these goals are achieved only partially, if at
all.
In practice, people will often hold a brainstorming session at this point, at which everyone
puts ideas forward freely and, in so doing, gains an idea of all the proposals. It is
assumed that the participants will have this freedom. If not, a decision-making problem
that is typical of the combination process will arise: each designer will wait for the
others to reveal their ideas before he is prepared to reveal his. In such a situation, one
designer or group of designers will probably take the initiative and propose an overall
plan which will include their own sub-solutions, and those that the other designers
were supposed to have produced, in an attempt to gain a lead on the others. The
party taking the initiative will formulate an overall plan that is favourable for them,
but which includes elements that actually belong to the decision area of others. If, at
such a moment, there is still confusion as to the allocation of the decision areas, the
process can run aground. Everyone talks about and decides on everything. If certain
parts of the plan drawn up by the break-away party taking the initiative seem to be
unfavourable or even incorrect to a designer who has knowledge of that particular held,
there will be little he can do about it. In such a situation, the designer has been known
to call on his own organisation or department to block the implementation of the plan
(on a hierarchic basis).
The rules that must be applied to ensure that the combination process runs smoothly
are similar to the rules of a game. In a game, the rules (the combination rules) are
hxed and each player has his own pieces (the resources). He can decide how to use
them, within the rules of the game. These are the two prerequisites for the game. With
resources of their own, but no rules, it is impossible for the players to devise a strategy.
With rules but no resources of their own, each player can use any resources.
For design in general, rules and individual resources are even more important than in
a game because there is feedback during the design process. A series of moves might
be stopped and reversed, if the outcome is wrong. The `design game' is not hrst played
to the end, but is partially repeated along the way. But which moves may be reversed,
which sub-solutions must be withdrawn, and who may decide?
We may conclude from the above that it is possible to control the combination process
only if we know in advance what each designers decision area is, and what the
combination rules are.
This does not preclude everyone proposing sub-solutions on any aspect, but certain
individuals are authorised to decide whether sub-solutions for certain aspects may be
included in the solution. !t is quite possible that changes might occur during the process,
shifts in decision areas and changes to the combination rules. These will be subjected
to negotiation and decision making.
6 Management of the combination process
Management can be described on the basis of its two main components, coordination
and control. Coordination is the linking of the activities and decisions of different
individuals. This allows a particular piece of work to be carried out as a complete entity.
Coordination is normally based on the allocation of responsibilities within the work
process; control is steering the process in the desired direction. This mainly entails
correcting any mistakes.

80
Generally speaking, a process will have been managed properly only if the results are
consistent with the values and characteristics determined beforehand. Management
ensures that the process is steered towards those results. Representing this as a simple
control model, we can say that the management body determines the interventions that
are necessary to the process and its support to obtain an output with those particular
values and characteristics. This is represented in Figure 6.
Figure 6: Management of a process (after In t Veld, 1989 p. 47)
Since, at the outset, the outcome of the design process is at best vague, management
of this process will focus mainly on clarifying the outcome (= design, hnal decision) step
by step. Moreover, since it is not entirely known at the outset how the design process
will be structured, management will also have to focus on setting it up and altering it
during the process: changes in the phasing, reallocation of the tasks that have to be
performed, links between the phases, etc.
The design literature, and particularly the literature on decision theory, mentions a
number of ways of achieving an effective structure for the design-decision process
and a good design-decision result. I shall simply set out the general framework for the
structuring of the design-decision process, using the model Herbert Simon has devised
for a decision-making process (in: Davis and Olsen, 1985 p. 199). His model is simple
and, partly as a result of its simplicity, has become very well known. According to this
model, a decision-making process can be structured around three process phases (Figure
7): intelligence, the phase during which problems and possibilities are investigated;
design, the phase during which problems and possibilities are analysed, and feasible
solutions are generated; and choice, the phase during which options are selected from
the various possibilities, and the chosen option is put forward for implementation.
Figure 7: Process phases in a decision-making process (after H.Simon)

81
Each phase can be further divided, using the principle of phased decision-making
(Wijnen et al., 1993 p. 13). In other words, the process in each phase can be divided
into a number of logical parts, and there will be a moment of decision between activities
(see Figure 8).
Figure 8: Phased decision-making (Wijnen et al., 1988 p. 13)
If phased decision-making is incorporated into Simons decision-making model, the
result is as depicted in Figure 9. The diagram now includes intelligence activities and
decisions, design activities and decisions, and choice activities and decisions.
Figure 9: Decision-making process with phased decision making
7 A case study: Stedelijk Museum Amsterdam
Two important elements of architectural design are the quantities of various architectural
resources that are applied (rooms, corridors, ofhces, entrance hall, etc.), and the location
of those resources in architectural space. Traditionally this was the architects problem
to solve; in modern practice the owner/principal as well as a whole range of technical
experts and organisations (e.g. the hre department) inhuence these allocations.
Increasingly, the prospective users themselves (as distinct from management or
developers) also demand and receive a voice in these negotiations. This has led to
a dramatic increase in the complexity of design processes, in which the design object
can sometimes be forgotten.
We will describe a design tool drawing from the helds of architecture, urban planning,
building design management, operations research, and measurement theory. It enables
a number of stakeholders from different disciplines to optimise and steer the design
together, each from their own perspective, by indicating preferences and restrictions on
function-location combinations, in an iterative search for a better design.
This new tool, the Architectural Design/Decision Room, builds on an earlier tool which
was successfully used in the design negotiations around the renovation and expansion of
one of Amsterdams major museums, the Stedelijk Museum. The Architectural Design/
Decision Room also shares many ideas and technologies with the Urban Decision Room,
a related tool which has recently shown great efhcacy in complex urban restructuring
questions.
The hnal section goes into ways of measuring stakeholders' preferences, and covers
a number of limitations faced by any preference-based system due to the nature of
preference and the current state of knowledge of measurement theory on this issue.

82
7.1 Project history
The Stedelijk Museum Amsterdam, or SMA, houses the citys contemporary art
collection. Its main building is located on the Museumplein, a large public area in and
around which other museums such as the Rijksmuseum and the Concertgebouw are
also located (Figure 10(a)). The original SMA building was designed in 1895 by A.W.
Weissman. In the 1950s and 60s, its capacity was expanded with a number of annexes
and intermediate hoors. Half a century later, the amount of space required for a modern
museum has again outgrown that which was available at the SMA. There was need for
a large-scale renovation of the existing building, and the construction of a signihcant
extension behind the current building, which would also house a new main entrance.
The Portuguese architect Alvaro Siza Vieira made initial plans for both the renovation
and the expansion. These were used to write a preliminary bill of requirements, which
was approved by the municipality along with a budget. At this point a number of
architects in succession were asked to develop more detailed plans. Each ran aground
on conhicting ideas between the municipality and the museum staff. Additionally,
the latter group felt left out and ignored in the decision-making process. The project
management consultancy hrm PKB was then asked to rehne the bill of requirements. !n
order to address this question to the satisfaction of both the staff and the municipality,
PKB used a set of computer models developed in conjunction with TU Delft. These
models will be described in the following sections.
!n July 200+, hve architects were invited to develop sketch designs based on the hnalised
bill of requirements. The design by Benthem Crouwel Architecten was selected, and
construction began in 2006 (Figure 10(b)).
Figure 10: (a) Exterior view of the Stedelijk Museum in Amsterdam;
(b) Artists impression of the new expansion, currently under construction.
7.2 Description of original single-input tool
The tool used by PKB in the development of the SNA bill of requirements consisted
of two connected computer models: one numerical, one geometrical. The numerical
model contained bandwidths for the amount of area required by the various functions,
and the budgetary restrictions imposed by the municipality. The geometrical model
contained the areas of spaces available in the main SMA building and a depot at another
location (geometrical restrictions), and dehned a hexibly sized space representing the
yet-to-be-designed extension.
Into the geometrical model, the PKB consultant entered the museum staffs preferences
on which functions were allowed to be allocated to which spaces: they could express
the htness of each particular room for use for every function, using Boolean values. For
instance, not all rooms were ht to be used for exhibitions (Barendse et al., 2006).
Combining the functions, spaces, and permitted allocations between the two, the
numerical model could generate an optimal allocation of functions to spaces. This was

83
output back to the geometrical model, by colour-coding the different spaces according
to their allocated function, so that the SMA staff could view the results. In other words,
the numerical results became graphical (Figure 11). Rehecting on these outcomes gave
rise to changed preferences for permitted allocations, which could be applied to the
next run of the models. !n this iterative way the staff could rehne their preferences until
they were satished with the proposed layout.
A detailed mathematical description of this numerical model was given in an earlier
paper (Van Loon et al., 2006). In brief, for allocable functions f, available spaces s,
allocation preferences p, and resulting allocations a, with the indices i, j identifying the
individual functions and spaces respectively, the function of the numerical model can
be given as:
given all f
i
, s
j
, p
ij
, maximise (1)
Figure 11: Visualisation of functions allocated to spaces in the original SMA tool.
(a) All hoors; (b) Close-up showing available area and allocation preferences per room.
7.3 Description of multi-stakeholder tool
A disadvantage of the tool used by PKB as described above is that it has only a single
set of preference inputs. In other words, all the various stakeholders involved in a
decision must agree on the allowed allocations to be entered into a model for a given
allocation run (Figure 12(a)). In the discussion round following the presentation of
the allocation, the participants rehect on the outcomes, and adjust their preferences.
As a consequence of the single-input nature of the tool, in the discussion round the
participants must also negotiate on the next set of preferences to be used (Figure
12(b)). Besides taking up valuable time, the distinction between the rehection and the
negotiation becomes clouded.
Figure 12: (a) A single set of preferences are used for all stakeholders; (b) The discussion
round consists of negotiations on the next set of preferences to run as well as rehection on the
results.

84
There is a second reason why a single-input tool is less desirable than it could be. An
important factor in structuring a decision-making process following the open approach
described in earlier sections is the issue of the ownership of constraints throughout the
whole process. A decision-support model must accept individual constraints from all the
parties involved, and maintain the independence of those constraints. In the context of
an allocation model such as the one we are dealing with here, the various stakeholders
must be able to enter their own preferences for each of the possible function-space
combinations independently of the others (Figure 13).
Figure 13: (a) An independent set of preferences is used for each stakeholder;
(b) The discussion round consists only of rehection on the results.
The ability to accept multiple sets of preference inputs can be added to the model
described by equation (1) quite easily, subject to one assumption being true. The model
can be extended by adding a set of users to the input variables, which then allows the
function-to-space preference to receive an additional index k representing the user
whose preference it is. Ultimately, of course, the numerical model must optimise to
some aggregate or combined resulting form of these multiple preference sets; the
single preference set which describes this form can simply be substituted into equation
(1), which then functions as before.
For functions f, spaces s, users u, preferences p, and allocations a, the function of the
numerical model can be given as:
given all f
i
, s
j
, u
k
, p
ijk
, preference gives p
ij
(2)
then given all f
i
, s
j
, p
ij
, maximise a
ij
(1)
The intriguing question raised by equation (2) is of course the nature of the preference
operation. The hnding of a valid mathematical form for this operation is the necessary
assumption mentioned above. Unfortunately this is not a trivial question, and the issue
of preference values and operations involving preference will be looked at further in
section 9.

i j

85
Figure 14: Visualisation of functions allocated to spaces in the new tool.
8 The measurement and addition of stakeholder preferences
In models where there is only a single set of preference inputs, or where the various
stakeholders have unique constraint areas (which is often the case in urban planning
models), there is no need for any form of preference addition, as the constraints can
be applied independently. As soon as more than one stakeholder is able to express a
preference regarding the same thing here the allocation of a particular function to a
particular space these distinct preferences clearly have to be combined, somehow, at
some point, for the model to hnd a group solution. Depending on the type of preference,
this can be necessary in the constraints of the model or in the construction of the
objective function.
8.1 Boolean and tri-valued veto preferences
In the models we have constructed to date, the stakeholders preferences have been
implemented in a very limited way, namely as Boolean values allowing or disallowing
a particular function to be allocated to a particular space. The number 1 or TRUE
means that a function may be allocated, the number 0 or FALSE means that it may not.
These preferences effectively act as veto criteria, and as such are part of the model
constraints. They can be implemented as follows:
ij: a
ij
< s
J
x p
ij
(3)
Expressed in words: the area of function i allocated to space j must be less than the
area of space j multiplied by either 0 or 1, i.e. either less than 0 or less than the area
of space j.
If Boolean preferences are to be combined, there are two simple implementations:
a) each constraint is treated independently, so that every stakeholder must assign
TRUE for a function to be allocable; b) all constraints are evaluated, and if any one
stakeholder assigns TRUE, the resulting value is also TRUE. There are more complex
forms imaginable, for instance requiring a majority (0+1+0 0; 0 1 1 1).
For completeness, it is of course also possible to implement a Boolean veto preference
as may/must instead of may not/may. This is not often used, however. Alternatively,
one can easily extend the Boolean veto system to a tri-valued system, may not/may/
must. This has been implemented in a limited way in our urban planning models, with
the session leader being able to hx functions in the model with group approval. Where
preference combination is required, however, more questions are raised: do may not
and must cancel each other out? Or does one have priority over the other? Does a
particular stakeholder overrule the others?
We will now introduce an alternative extensible notation for preference values, which is
j
ij
a
A

86
also usable for tri-valued or higher systems. Its nomenclature is based on the meaning of
the preference rather than the accident of the corresponding method of implementation,
as with the Boolean notation above. After the terms used in the previous paragraph, the
negative veto preference will be labelled N for may not; the positive veto preference
will be labelled M for must; the neutral may or allowed preference will be labelled 0
(zero).
Using the new preference notation we can construct a combination table for two
stakeholders using a tri-valued veto system {N, 0, M} (Table 1). May not and must
both overrule may, being veto criteria; may not must yields a conhict (C). This
sounds like a problem, but is in fact quite acceptable within the broader approach:
an apparent conhict between the views of two stakeholders has been revealed, which
needs to be discussed prior to the next run. Either the problem is truly intractable,
in which case infeasible is the correct outcome; or one or both stakeholders can be
provided with an alternative solution elsewhere, relaxing the conhicting constraint; or
the organisational or contractual arrangements are such that one party can overrule the
others, transparently within the process. These rules hold when the system is expanded
for three or more stakeholders.

N O M
M
O
N
C M M
N O
N

Table 1: Combination table for {N, 0, M} (tri-valued veto system) and two stakeholders.
These rules can be expanded for an arbitrary number of stakeholders.
8.2 Multi-valued relative preferences
Participants in workshops using these models often hnd that the Boolean or tri-valued
veto approach described above is too restrictive to express their preferences as they
would wish. They want to be able to indicate relative values: I dont particularly mind
allocation x, but I would much prefer allocation y. I dont really like allocation z, but can
live with it if its absolutely necessary. An important change needs to be made to the
numerical model to support this. The model must no longer simply maximise the total
allocated area, but the objective function must maximise the stakeholders preferences
for the allocated areas.
An individual stakeholder can express relative preferences on an ordinal scale. An
approach such as the hve-valued relative preference set {--, -, 0, +, ++) often seen
in marketing surveys comes to mind, or grading on a scale of one to ten. However,
due to the ordinal nature of these scales, there is no information on how much better
++ is than +, or 8 than 6; only that it is `better'. Consequently it is not possible to
perform further comparative mathematical operations on these preferences, and they
are unable to be used as a measure for optimisation. (Strictly speaking, mathematical
operations can be performed; it is just that the outcomes are formally undehned and
hence meaningless.)
Preferences expressed on an interval scale can be worked with. In urban planning
models where a rent-based bidding system is used as part of the allocation (maximising

87
return), we have observed participants use different rent values as a mechanism to
introduce relative preferences implicitly. This can be implemented in the objective
function as follows:
given all f
i
, s
j
, p
ij
, maximise a
ij
x p
ij
(4)
It should be noted that preference itself cannot be measured on an interval scale
(Barzilai, 2005). There is no unit of preference, nor an absolute zero of preference
with respect to which a unit could be dehned. !n the urban planning example above,
the participants used a separate scale measured in rent-euros to approximate their
preferences. Though they are using the same scale, each participants mapping of
preferences to euros is different, and it is not possible to determine how far the resulting
group optimum deviates from the true group optimum.
Preferences for three or more alternatives can be expressed on a relative, proportional
scale. In recent years there has been considerable debate on preference in the
held of measurement theory, and for these models it is not immediately apparent
how to implement either the measurement of the preferences or their subsequent
implementation in the objective function. Research is ongoing in both these areas.
8.3 Implementing a limited veto and relative preference system
The {N, 0, M} veto system described in section 9.1 can be extended to include a
two-valued relative system, allowing the implementation of a limited mixed system
{N, 0, 1, M}. While 0 still represents may or allowed, 1 (one) represents preferred.
Clearly this is a purely ordinal scale, with all the limitations that entails. Nonetheless,
due to the nature of zero and one, the problem of measuring the difference between
the two categories disappears: only the area allocated in the category preferred, 1, is
counted. The problem of different participants different interpretations of the verbal
label preferred still remains; the participants need to be very aware of this if they opt
to try this system.
Expanding the combination table in Table 1 for {N, 0, 1, N) reveals another consequence
of these ordinal categories: preferred preferred what? !t may be tempting to
say that the relevant allocation is twice as preferred and so should have a value of
two, but in ordinal categories this is meaningless, and any non-veto operation involving
preferred maps back to preferred (Table 2). These rules are again extensible to three
or more stakeholders.
N 0 1
M
1
0
N
C M M
N 1 ?
N 0
N
M
M
N 0 1
M
1
0
N
C M M
N 1 1
N 0
N
M
M
Table 2: Combination table for {N, 0, 1, M} and two stakeholders.
(a) The problem of 1 1 ? (b) 1 anything (except veto values) always maps back to 1.

i j

88
9 Conclusion
The use of both numerical and geometrical models greatly reduced the time it normally
takes to develop a bill of requirements for such a complex project. The open process
made the staff of the Stedelijk Museum feel their wishes were taken seriously and
not swept under the carpet. In contrast to traditional approaches, PKB could provide
conhdence that the bill of requirements would satisfy both the budgetary restrictions
imposed by the municipality and the geometrical restrictions imposed by the existing
buildings. In the traditional approach some rules of thumb would be used to establish
whether the bill of requirements would meet both budgetary and geometrical restrictions,
which often give rise to unpleasant surprises later on in terms of overruns in time and
money.
The design process for construction projects has become increasingly complex in recent
decades, as more and more parties inhuence the decision-making process in diverse
ways. Ideas from management theory and operations research, and mathematical
models which make these ideas operational, can aid in bringing the design process to
a successful conclusion. This paper has shown how a preference-based single-input
tool which has already proven itself in this held can be extended to support multi-
stakeholder use directly. This new tool is currently being developed, using the SMA case
as experimental subject.
It has been observed that stakeholders wish to extend the range within which they can
express their preferences. However, a number of strict and severe limitations on the
ability to measure preferences for this purpose have been shown, stemming from the
current state of knowledge in measurement theory.
References
Barendse, P., Binnekamp, R., Graaf, R.P. de, (2006), Integrating linear
programming optimisation and geometric modelling, in: Aouad, G., et
al. (eds.), 3rd International SCRI Symposium, proceedings, University
of Salford, Manchester, pp. 295-304.
Barzilai, J., (2005), Measurement and Preference Function Modelling, Int.
Trans. in Operational Res., Vol. 12, pp. 173-183.
Binnekamp, R., Gunsteren, L.A. van, Loon, P.P. van, (2006), Open Design, a
Stakeholder-oriented Approach in Architecture, Urban Planning,
and Project Management, Research in Design Series, Vol. 1, IOS
Press, Amsterdam.
Chadwick, G., (1971), A Systems View of Planning, towards a Theory of the
Urban and Regional Planning Process, Pergamon Press, Oxford.
Chen, W., Lewis, K.E., Schmidt, L.C., (2006), The Open Workshop on Decision
Based Design, in: Lewis, K.E., Chen, W., Schmidt, L.C., Decision
Making in Engineering Design, ASME Press, New York.
Davis, G. B., and Olson, M. H., (1985), Management Information Systems,
McGraw-Hill Books, New York.
Durham, D.R., (2006), The Need for Design Theory Research, in: Lewis, K.E.,
Chen, W., Schmidt, L.C., Decision Making in Engineering Design, ASME
Press, New York.
Faludi, A., (1973), Planning Theory, Pergamon, Oxford.
Foque, R., (197S), Ontwerpsystemen, een !nleiding tot de Ontwerptheorie,
Spectrum, Utrecht.

89
Hamel, R., (1990), Over het Denken van de Architect, een Cognitief
Psychologische Beschrijving van het Ontwerpproces bij Architecten,
AHA Books, Amsterdam.
Jones, J. C., (1970), Design Methods, J. Wiley, London.
Loon, P.P. van, Burger, J., Graaf, R.P. de, (2006), Optimum architectural group
design as rehection in action', in: F. Scheublin et al. (eds.), Adaptables
06; proceedings of the joint CIB, Tensinet, IASS international
conference on adaptability in design and construction, Eindhoven
University of Technology, Eindhoven, pp. 12-103-12-107.
McLoughlin, J. B., (1969), Urban and Regional Planning, a Systems Approach,
Faber, London.
Mitchell, W. J., (1990), The Logic of Architecture, Design, Computation and
Cognition, MIT Press, Cambridge Mass.
Newsome, S.L., Spillers, W.R., Finger, S., (eds.), (1989), Design Theory 88,
Springer-Verlag, New York.
Roozenburg, N. F. M., and Eekels, J., (1991), Produktontwerpen, Structuur en
Methoden, Uitgever Lemma, Utrecht.
Tzonis, A., (1982), Het Architectonisch Denken, Socialistische Uitgeverij,
Nijmegen.
Veld, J. i. t., (1988), Analyse van Organisatie Problemen, een Toepassing van
Denken in Systemen en Processen, Stenfert Kroese, Leiden.
Wijnen, G., Renes, W., and Storm, P., (1993), Projectmatig Werken, Spectrum,
Utrecht.

IDE+A
Technology Diffusion and Design

91
7 The metabolism of knowledge
Dr. ir. W.A. Poelman
Associate Professor Product Development
Faculty of Architecture
Introduction
Discussing the subject of Technology Diffusion and Design can take place on different
levels. The hrst level is the role that design plays in technology diffusion on the macro-
level and the responsibility connected to society. The second level is the role design
plays on the micro-level in the context of design processes.
Of course the emphasis will lie in this paper on the second level, but in the context of
the subtitle, some attention will be paid to the macro-level.
Referring to this subtitle of the conference Life is a theater, a conclusion in the
introduction reads: Nowadays, the script of life is for a large part written by architects
and designers. Urban planning prescribes how we spread our activities geographical. The
design of modern residential districts determine for a large part how we communicate
with each other. The design of shopping centers determine how we acquire our
foodstuffs. Designers of means for transport decide how we move ourselves and kitchen
designers decide how we cook.
Of course these conclusions are too easy. We cannot just claim that the designers
of television sets and programs decide for us that we spend our evenings before the
television set and not around the table playing family games. However, we cannot deny
an immense inhuence of this developments on the behavior of people.
Research on this phenomenon, carried out as Constructive Technology Assessment is
in the hrst place a task for sociologists and psychologists, but it is welcome if architects
and designers participate actively in this discussion, and of course they do already.
Important in the introduction of new technology applications is the phenomenon: what
is may become ought. Let us discuss some examples.
Twenty years ago we would not even think about listening to music in trains with a
headphone and a portable audio device. However, it was already possible. In museums
you could hire such devices years ago for guiding purposes. The headphone culture
started however when Sony introduced the existing functionality in a new coat for the
purpose of listening to music in public spaces. After that it became part of the script of
life. If Sony says so you can do it.
The same thing happened with MacDonalds Drive Inns. It was not ought to eat in a
car and for many people it still isnt. However, when a brand as McDonald suggests that
it is acceptable, many people will accept a visit to a Drive Inn as an alternative for the
script of having lunch, which used to be a social event.
Coming back to television and the design of residential districts. Could you blame
designers of television sets and programs that the behavior of people has changed?
Could you blame the designers of the Bijlmer for social problems originating from
architecture? Of course some designers feel embarrassed by the inhuence they have,
but they have to keep in mind that they play a minor role in a process on a higher level

92 Technology Diffusion and Design - W.A. Poelman
which acts like a train which is not easy to control.
New technology becomes available in a spectacular tempo as a result of research
activities and designers are always eager to hnd applications of that technologies in
products, either driven by their own ambition or by the ambition of clients who want to
earn money or score in another sense with innovative products such as buildings.
Probably this mechanism determines our future more than a mechanism in which
values of life are the starting point for concrete wishes which are translated in products
fulhlling these wishes.
Many writers have thought about the destination of this train and it could be heaven or
it could be hell. Aldous Huxley wrote his book Brave New World in 1932 and every part
of it is subject of discussion nowadays: mood drugs, biotechnology, consumer-society,
birth control, etcetera. (see www.huxley.net)
Figure 1: Cover of Brave New World Figure 2: Cover of Nineteen Eighty Four
The story is the same with James Orwells Nineteen eighty four: technology enables
new scripts of life and how will these scripts evolve?
In brave new world the necessary technological means to keep people happy are
all applied, effectively leading to a society which we would not want. The same we
see in the Natrix, the hlm released in 1999 by Wachowski Brothers, Joel Silver and
Warner Brothers, and claimed by Sophia Steward to be based on her book The Third
Eye, copywrited in 1981. Also the famous Ninority Report, a 2002 science hction hlm
directed by Steven Spielberg, loosely based on the Philip K. Dick short story of the
same name, hts in this range of attempts to describe a scenario which might result
from technological developments. Of course there are many other examples like 2001
Space Odyssey of Stanley Kubrick (1968), Alphaville of Jean Luc Godard and ExistenZ
of David Kronenburg.

Figure 3 Poster of The matrix
How far are we with the evolvement of these scripts? Are we the frogs in the kettle
which is slowly warming up, keeping us inside? What inhuences have designers on
these scripts? !s there a concept of life on which architects and industrial designers
base their work? Do designers have a controlling task in the application process of new
technology or are designers just prostituting themselves for industry as professor Jan
Jacobs, former director education of the School for Industrial Design Engineering of
Delft University of Technology claimed once supposed during a conference.
Within the disciplines of industrial design engineering and architecture there is a lot of
organized discussion about their societal role. However, there is not enough discussion
about the way they are embedded in the overall process of technology development and
application. This is necessary in order to position them selves. With respect to rehection
on their position in society, architects can build on a long history, industrial designers
can not. Architects are consulted regarding social issues, industrial designers hardly.
However they could fulhll an important role in the process of scenario development in
general from their ability to imagine a non existing future. Design is nothing else than
creating a non existing future. A world in which a certain product does not exist is per
dehnition different from a world in which it exists. This fact represents an important
constraint in market investigation.
visualizing future worlds is now mainly done by science hction artists, but professional
industrial designers are not often employed for this purpose. Of course there are
examples like an industrial design agency which received an assignment to visualize
possible means for military defense in the future. However, many design engineers see
it as a risky affair. When Leonardo da Vinci would have worked at Delft University now,
probably he would be scientihcally shot down. !t is not ought nowadays to design things
that are not possible to produce yet.
1 Technology diffusion and design processes
What is the relationship between knowledge diffusion and design? All of the designers
who were interviewed in the eight cases acknowledged the matter of knowledge
diffusion, but they did not discuss what knowledge diffusion involves. The designers
apparently regarded technology diffusion itself as innovation, through the application
of technology from third parties. We should nevertheless take a more fundamental look
at the phenomenon of knowledge diffusion and design, and doing so requires sound
dehnitions of the concepts that are used. !n the context of this paper, knowledge is

dehned as the ability to apply information and resources to achieve a dehned goal".
Diffusion is dehned as the implementation of new or existing knowledge within new
helds of application". Design is limited to industrial and architectural design.
1.1 The cases
The cases provide several examples of the various characters of knowledge diffusion.
We consider each one separately.
The Westraven building by CePeZed provides an example of the diffusion of technology
from technical-textiles applications to facades. In addition, it involves the diffusion of
inhatable pillows from dome constructions to hat facades.
Figure 4: Westraven
The A230 chair by Ahrend apparently involves no direct
example of technology diffusion, although the development of the hinges seems to be
quite new. Bas Pruijser works as a designer for both a manufacturer of ofhce furniture
(e.g. Ahrend) and a manufacturer of garbage-management devices (e.g. Bammens).
He inevitably transfers technological knowledge from one application to the other.
Figure 5: A230 Chair
The image-and-sound (in Dutch, beeld en geluid) building by Neutelings-Riedijk offers a
clear example of technology diffusion between branches. They collaborated closely with

Jaap Drupsteen, a graphical and media designer. In addition to technology borrowed
from the held of graphic arts, quality standards were transferred to the glass facade to
realise this remarkable building.
Figure 6: Beeld en Geluid gebouw
The BeerTender by NN!D is a hne example of a special kind of knowledge transfer, namely
from the professional market to the consumer market. This type of knowledge transfer
occurs on a regular basis. Examples include do-it-yourself tools, kitchen devices and
audio devices. Direct translations are seldom possible, however, because of the fact that
business structures differ. In the example of the BeerTender, professional maintenance
services cannot be utilised and ergonomics must be adapted to inexperienced beer
drafters. One of the most important differences is that the time span until a container is
empty is longer in the consumer market then in the professional market. This required
the development of a new system.

Figure 7: Beertender container
The 1-2-3 House by Martini is an interesting example of a different situation. This
case illustrates a match between knowledge about prefabricated buildings (from the
concrete industry) and industrial manufacturing systems.


Figure 8: 123 huis 1
In this case, concrete, integrated building elements are handled in the same way as
objects in the car industry are handled. Building elements are manipulated by robots
for logistic purposes and ergonomic comfort.
One interesting aspect is that magnets are applied to keep cables and other inserts in
position during the moulding process. The integration of functions allows the use of
high-quality materials. This case reveals an important difference between the building
industry and other industries. Transport costs limit the geographical expansion, and
thus the room for investment, of products like the 1-2-3 House.

Figure 9: 123 huis 2
The next case involves Carver, which can be regarded as a clear example of knowledge
diffusion between pure, advanced mechanical engineering and industrial design.
The product is based upon the invention of a hydraulic canting mechanism, which
enables stability of narrow vehicles. The application of the system, however, inevitably
leads to both a striking driving experience and a striking visual appearance. In fact, a
new archetype of a vehicle is created which resembles a cross between a motorcycle
and a small car. The success of the design is a result of the collaboration between the
engineering company (Carver Europe) and the design company (Spark Engineering).

Figure 10: Carver
The building in which this conference is taking place the Industrial Design Engineering
Building, designed by Fons Verheyen is an example of a project in which cooperation
between architects and industrial designers might be expected. In reference to a new
technology, Verheyen mentions that the fencing is done without balusters. Although
this is an obvious example of technology diffusion, it can also demonstrate the diffusion
of modern building technology to a project aimed at giving an existing building a
complete new function.
This kind of design, which is especially important in architecture, could be specihed as
Supply Driven Design (SDD), which proceeds from existing artefacts. Although we
cannot go into depth about this relatively new, sustainable type of design activity, we
can conclude that more creativity is needed to design something within the limitations of
an existing artefact than is needed to design something completely new. In this regard,
industrial designers could learn from architects, who do this on a regular basis.
Figure 11: IDE building
The last project to discuss is the inhatable care bed, which combines several technologies
from different applications in one product. The reason for this combination is open to
speculation. One option is that it was due to a different approach to the design process,
which proceeded from the design of goals instead of from the design of means. The

bed was not designed as a synchronous product but as a diachronic script. The physical
product was simply a way to enable that script. Because traditional care beds did not
ht into that script, it was necessary to design a completely new product. !n this project,
the product was not based upon available technology; the technology was selected to
meet the design goals. It was therefore necessary to look outside the technologies that
have traditionally been used in care beds. Because this diffused technology was not
developed for this goal, it was necessary to invest considerable effort in making this
technology appropriate.
Figure 12: Care bed
With these projects in mind, the following section discusses a new paradigm on design
and technology diffusion.
2 Design as information processing
The new paradigm can be dehned as follows: `!n principle, design is information
processing. The design process can be regarded as a black box, in which information
goes in and information comes out.
Black box
design
process
information about
user needs,
technological
possibilities,
etcetera.
information for
marketing and
manufacturing
Figure 13: Design process as a black box
How should we consider physical presentation models and technical prototypes?
Should they also be considered information outputs? Absolutely! Nodels, prototypes
and drawings are simply information carriers or media. It has been said that products
are not the result of the design process, but of the production process. Within the black
box, all manner of explicable and inexplicable event take place. In this paper, we do not
focus on the design processes that take place in the black box. We concentrate instead
on the aspect of information processing.

Our concentration on information processing requires a dehnition of a product (which
could also be a building). As claimed by Kotler (2002), A physical product is just a
means to create functions
1
. People do not need the physical product; they need the
functions that products enable. The physical aspects are generally a necessary evil. The
physical aspects occupy space, need maintenance and pollute our world.
Enabling user functions requires functionalities, most of which continue to require
physical means. This may change over time. Consider the `roll of hlm', which has
disappeared as a physical means for storing and transporting pictures. Consider also
communication cables, which have largely been replaced by wireless technology.
2.1 Supply of and demand for information
In the context of this paper, we organise design information into two categories: supply
and demand. Starting with the last one, demand information is linked to the design of
goals, which precedes the design of means. In current times, more products (including
buildings) are failing because of defects in the design of goals than because of defects
in the design of mean. Although possibilities continue to expand in a technical sense, it
is not always easy to hnd room in the market for new productfmarket combinations.
functionalities
functionalities
functionalities
functionalities
functionalities
functionalities
functionalities
functionalities
potentialities
potentialities
potentialities
potentialities
potentialities
potentialities
potentialities
potentialities
properties
properties
technology
objective
functions
objective
functions
product to
design
Strategic
product plan
Operational
product plan
a
s
s
o
c
i
a
t
i
o
n
p
r
o
c
e
s
s
Figure 14: Product development and the diffusion of technology
In general, goals can be described in terms of objective and subjective functions,
2
each
of which is realised through functionalities. In this context, functionalities should be
interpreted as indivisible functions, such as keeps warm or changes colour. Product
functions can generally be described by arranging a large number of functionalities in a
tree structure. The design of goals can be described using descriptions of functionalities
in predehned contexts.
On the supply side, technology can enable potentialities. Although potentialities can
be dehned as indivisible functions, the fact that the functions themselves have yet to
be dehned requires a new noun, `potentiality'", which refers to the capacity to enable
functionalities. The number of potentialities increases at the speed of technological
1 We assume that Kotler included the realization of emotional values in his concept of
functions
2 We will not elaborate on these concepts in this context. See Poelman 2005.

development. New and interesting potentialities are discovered every day. Although
examples can be found in nanotechnology or other disciplines, they occur in nature as
well.
We could regard the design process as a process of association between the demand
side (as expressed in functionalities) and the supply side (as expressed in potentialities).
Because it is impossible to make associations with unknown information, we can
conclude that designers should have as many potentialities in their minds as possible.
It is important to note that technical background information is not necessary. In the
design stage, it is important to know only what might be possible. It is not necessary
to know how it is possible. More commonly stated, It is enough to have heard the bell
ring; we can use the internet to hnd the clapper'.
The lack of emphasis on technical explanations is not a matter of which designers and
architects should be ashamed, and most of the good ones are not. Nonetheless, even as
non-professionals in helds of manufacturing, they manage to obtain maximum results
from their suppliers (see the glass facade of Neutelings/Riedijk/Drupsteen). Suppliers
start by saying that something is impossible, as the costs of doing it differently are
sometimes high. Some degree of pressure by the designer is often benehcial to both
parties. This can result in a better product, ensuring that the supplier then has more
to offer.
2.3 Towards a new paradigm for the knowledge-diffusion process
Assuming that designers are able to develop sound designs of goals, and assuming
that they have enough knowledge about potentialities at their disposal, the process
of matching potentialities and functionalities is the key activity for designers. Although
industrial designers should be trained for this task, traditional design methodology
unfortunately does not provide solutions for such training.
The process of associating functionalities and potentialities involves more than simply
a designer sitting and thinking. It is a complex process involving many media, people
and organisations. It cannot be explained by traditional organisation models. Many sub-
processes can be distinguished that usually have nothing to do with the design process
itself.
Analysis of these processes has led to an attempt to use the metaphor of an organism
rather than an organisation to describe the general process. The difference can be
explained as follows. An organisation is created to enable a process and often more than
one process. In contrast, an organism is not created but evolved, and it is dedicated to
dehned tasks.
In nature, even in one-celled creatures, we can observe processes taking place in an
organism that can do nothing other than carry out that specihc process. A process and
an organism can be seen as two sides of the same coin. The process is the diachronic
organisation of activities. The organism is the synchronic, functional organisation that
facilitates the activities (see hgure 1S).

Acting system
(material organisation)
Concerning system
(to be transformed)
Functional organisation
Method
Acting process
Synchronic
Diachronic
Relation between method
and organisation
Action subject
Action object
Figure 15: Relation between method and organization
!t is not sufhcient to state that the design company represents the organism that carries
out the design process. Every interviewee in this preliminary research expressed in their
own words that the situation is much more complex. Many efforts have been made to
describe the external design organisation in traditional schemes (Poelman 2005). In
general, but cover only part of the situation.
Let us analyse the process of knowledge diffusion. Assuming that some kind of organism
is carrying out this process, we considered the possibility of using metaphors from other
domains to describe the organism. This exercise resulted in the knowledge metabolism
model.
Figure 16: Model of knowledge metabolism for development projects

Design
Processes
In this model, a set of mechanisms can be distinguished, which can be projected onto
the eight cases. Before presenting this projection, we must explain the one-celled
organism that represents a design project (not a design company).
Each project has a strategic level (1), a tactical level (2) and an operational level (3).
1. The strategic level represents know-what. It is the planning level, which can be
compared to DNA. When something is wrong with the DNA of the project, it will fail or
lead to an unexpected outcome. Such outcomes are sometimes better than expected.
After all, evolution is partly based upon imperfections in copying genes. Naturalis in
Leiden has probably become more successful because the planning changed from city
centre to the outskirts.
2. The success also emerged from the hexibility of the project team at the tactical or
know-how level (second level). At this level, skills are developed that can be compared
to the proteins in a biological cell. As before, the biological organism represents the
project as a whole and not the design company. Skills represent both the skills within
the design company and those of every involved party. One important skill of the
design company, however, is to involve the right parties. This proved a crucial aspect in
nearly every case. According to Neutelings-Riedijk, We do not have preferred supplier.
Companies involved in creating a building can be compared to a travelling circus. One
moment, they are all there with their knees in the Dutch clay; the next moment, they
are all gone, back to where they came from.
3. The real work of design takes place at the third level: the operational or know-where
level. This level involves hnding the right information and the right people. At the
operational level, output information is produced and packaged in such media as texts,
drawings, models or computer simulations.
The next question concerns the process itself, which is executed within the project
organism. This process is freely derived from research done by Hargadon (1997) in
IDEO, an international design agency. Hargadon discovered that, as soon as knowledge
enters the project, it is processed in several steps: acquisition, generalisation, association,
application and recording. The interpretation in this model is as follows:
Acquisition: the actual transfer of knowledge
Generalisation: the analysis of knowledge and dehnition of
potentialities
Association: the creative process of hnding potentialities to realise
dehned functionalities (through product-function
analysis)
Application: the integration of knowledge in industrial product design
Recording: preparation of information for later use, in which the
identihcation of potentialities plays an important role
Knowledge acquisition is a complex process. This process is easier to understand by
breaking down the process into leads, follow-up and transfer. This breakdown is
borrowed from the discipline of direct marketing. Because it is impossible provide the
whole world with information about a product, it is necessary to hnd leads: indications
that particular prospects might be interested (e.g. they returned a reply card. After
prospects have shown interest in the product, it is necessary to follow up in order to
learn whether they are truly interested. The third phase, the transfer of the order, is of
course the most important.

The same applies to the process of knowledge acquisition. !t is impossible to involve
all of the information in the world in a given project. Leads (or potentialities) are
necessary. Leads are the capital of a direct marketing hrm, and they are likely to serve
the same function in a design hrm. Designers who are inept in follow-up, however, will
not be successful. Designers should be skilled in motivating suppliers to provide more
information and invest in the project.
In the third step, knowledge transfer, learning and engineering capabilities become
important. While any of the cases could be used to illustrate these steps, let us consider
the media building of Neutelings-Riedijk. One (external) party in the project team was
Jaap Drupsteen, an expert in exploiting the potentialities of new technology. He knew
that glass could be moulded and printed with modern computer-supported techniques.
This lead was followed up with visits to companies who could accomplish this. The
transfer was realised through a difhcult iterative process, in which quality goals of
Drupsteen competed with economic goals of the glass producer. Drupsteen is skilled
in translating potentialities (the tricks that we know) into visually spectacular effects.
With respect to recording, according to Neutelings-Riedijk, all knowledge is common
knowledge in architecture.
As mentioned above, having leads is an important selection criterion for knowledge
acquisition. Other mechanisms play a role as well, however, and we refer to them as
constraints. Constraints operate in both positive and negative ways. One negative
function is that they can prevent useful knowledge from coming through. A positive
effect of constraints is that they can serve a pre-selection function. Knowledge that
triggers no interest will not be processed and will thus be prevented from entering.
For the sake of clarity, the various kinds of constraints are not shown in the hgure. As
a former CEO of a chemical company, Prof. Johannes Eekels advised using piping as a
metaphor for dehning the following series of constraints:
Valve: A valve is a mechanism that prohibits knowledge from howing into or out of
the project, as with an embargo on speaking with certain companies. Such valves can
also be identihed in the eight projects, in some cases blocking the how of knowledge
between governmental parties and companies. Intellectual property (IP) considerations
form another common reason for blocking knowledge transfer.
Narrowing: Narrowing is a mechanism that limits the how of knowledge into or out
of the project, as illustrated by a lack of capacity. There is no time for reading. The
stacks of information that we wish to consume increase throughout the course of our
careers.
Semi-permeable membrane: A semi-permeable membrane is a mechanism that
selectively prevents knowledge from coming through, as in the case of marketing
information that is withheld by the management. Everyone has specihc hobbies, and
people tend to defend their own areas of specialisation.
One-way valve: A one-way valve is a mechanism lets knowledge through in only one
direction (e.g. from source to recipient); no information is provided to information
sources about information needs. Students should be told about the importance of a
win-win situation. One receives information only when one provides information. In
order to understand the kind of information that is of interest to a particular contact, it
is essential to be interested in the activities and opinions of that contact.
Filter: A hlter is a mechanism that prohibits complex knowledge items from passing

(e.g. dehciencies in education concerning the analysis of scientihc information). This
mechanism is widely familiar. Faced with the choice between a two-page article and a
+00-page report, we tend to choose the article hrst. There are so many articles to read
that we often do not manage to read the more profound materials.
Labyrinth: A labyrinth is a mechanism that slows down the information how (e.g. a
manager who insists upon seeing the information hrst and then passes it on to the wrong
person). A familiar example involves magazine issues that arrive on an employees desk
half a year after publication.
Leakage: Leakage is a mechanism that leads to the loss of knowledge (e.g. hling systems
that do not function properly). Because knowledge is a crucial asset of a company, the
leakage of knowledge is a severe crime, and guilty parties should be punished. Jan
Pesman (CePeZed) states, All knowledge that is gained is stored in the project and for
use in future projects. This knowledge increases the toolbox. Every project is a learning
process, and the key moments from this process can be reused at any moment even
after hfteen years'. !t would be interesting to learn more about he accomplishes this.
Compatibility: Compatibility is a mechanism that prevents information from diffusing
(e.g. the extent to which the source and the recipient of the knowledge are able to
communicate). This is probably the most interesting constraint in this world of polarisation.
Every discipline has its own language, culture and set of ontologies. Although this is
often perceived in a negative light, the successful development of disciplines depends
upon these aspects.
The successful diffusion of multi-disciplinary knowledge can be stimulated by paying
more attention to the interface between disciplines the designer. Designers should
have an interdisciplinary attitude, be conversant (but not necessarily huent) in more
than one jargon and feel empathy with professionals from other disciplines.
With regard to the organism in Figure 16, two concepts have yet to be discussed:
the sensor function and the ejector function. The sensor function is quite simple.
The ability to notice `weak signals' (Payens 1996) is quite important to the process of
knowledge diffusion within a project. A project is not a closed entity; it develops itself
in a cosmonomy, as represented in Figure 3.

Figure 17: Cosmonomy of projects and knowledge diffusion
A project can be regarded as an entity hoating in an ìnformation soup', together with
other projects. Sensors detect what is going on in their own neighbourhoods.
Ejectors send out information in order to allure interesting partners (pheromones). In
some cases, this can lead to the mating of projects. This is an essential function in the
project. Good project teams communicate intensively about the activities with which
they are occupied. This increases the chance that other parties will take an interest in
collaboration.
Discussion
The metabolism model for knowledge diffusion should be regarded as a result of an
attempt to make the complex issue better understandable. Fishbone diagrams do not
describe sufhciently the real way people work together in development processes.
Sociograms do better because they pay attention to the informal organization which is
often of greater importance than the informal organization. However, also sociograms
present only a part of the organization as such, just the synchronic part. The diachronic
part is mostly described apart in the context of methodology.
Models in methodology can be divided in three basic categories (Roozenburg/Eekels
1995): activity models (fundamental design cycle), phase models (VDI 2221, Pahl &
Beitz) and aspect models (eekels, Andreasen, Archer).
The metabolism model could be added in a fourth category: function models.
Activity models refer to hours to be spent. Phase models refer to results in-between
and aspect models refer to points of attention. A function model refers to skills needed
in the development process in different stages. In that sense a function model forms
a bridge between the process and the organization. Furthermore this function model
forms a bridge between the design process and technology diffusion issues. Finally, a
function model hts in the way of thinking of designers. After all, a designer is a creator
of functions as Kotler claims.

With respect to the relation between architecture and industrial design engineering we
can conclude that hgure 17 provides an opportunity to think about cooperation between
the two disciplines. Industrial designers and architects meet each other more and more
in the cosmonomy of development projects such as the creation of new buildings,
transport facilities and the composition of public space.
Knowledge of the two disciplines has, since the 40 years of existence of the faculty of
industrial design engineering become more and more complementary. It is time for
intensifying knowledge diffusion between the two disciplines.
References
Poelman, W.A. (2005,) Technology Diffusion in Product Design, thesis, Chair
Design for Sustainability, Delft University of Technology, Delft
Kotler, Ph. (2002), Marketing Management.Analysis Planning & Control, 11
th
edition, Prentice Hall International, London
Hargadon, Andrew & Sutton, Robert I. (1997). Technology Brokering and
Innovation in a Product Development Firm. Administrative Science
Quarterly, vol. 42, December, p 716-749.
Payens, Ruud. (1996). Het Zesde Zintuig, Stichting Innovatiecentrum Noord &
Oost Gelderland, Apeldoorn.
Roozenburg, N.F.M. & Eekels, J. (1995), Industrial Product Design:
Fundamentals and Methods, Wiley , New York

107

IDE+A
Closing Speech

109
8 Closing Speech
6 June 2008 IDE+A symposium Design Processes
Prof. dr. ir. A.C.J.M. Eekhout
on behalf of the dean of the faculty of architecture.
One of the objectives of the conference `Design Processes', concerning the scientihc
held of Design Nethodology, is to make the invisible visible" in the immaterial held.
The literal meaning of this is to make the invisible preparation process, which precedes
the production of new building and industrial products and components, visible and
understandable by a textual and visual description.
But hguratively making the invisible visible also means to partly unravel the mysterious,
the unknown and the unsaid and pass it on to architects, building technologists,
industrial designers and to students as a new knowledge and insight. The mysterious
brings along some uncertainty about objectives. Mysteries are challenging, they are a
motivation to go and do research and therefore, as far as I am concerned, they never
need to be solved completely. When one mystery is solved, new mysteries will have to
appear, new challenges, ever further on the way to the future. Yet, in the meantime
knowledge grows, the skill, the insight and hopefully also the vision on the specialism
of product design and development.
Dutch Design and Dutch Architecture are internationally appreciated for its powerful
value-for-money quality and its surprises within the set limitations of the challenges.
Dutch architects and industrial designers often have to dance on the rope. Solid
design approximations have contributed to this quality of Dutch Design and Dutch
Architecture.
This Conference Design Processes is dedicated to the methodology and processes of
designing, developments and research of building and industrial products, systems and
components, as well as to the applications of industrial products in buildings.
Therefore, it is of importance to product designers and building product developers,
who are mainly concerned with developing products and components at the side of
producers, as well as to materializing architects and component designers who, at the
architect's ofhce, are concerned with the materializing of the functional and spatial
building concept as a whole and in parts.
The conference hrst and foremost meant to bridge the knowledge helds of Architecture
and Industrial Design Engineering, but also for professionals and students in the
professional held of both faculties.
Design Methodology
Design Methodology has a long and thorough history at industrial design Engineering,
thanks to the books of Norbert Roozenburg and the late Johannes Eekels.
In architecture the situation is more varied. There is a lot of talk on designing in the
architectural world, but there seems to be little openness and uniformity when it comes
to the process of designing and what design methods are being used.
Nowadays, the computer became an accustomed medium in every design ofhce and
even conceptual design possibilities are being carefully explored.

110
But the systematics and methodology of design have to go through a renaissance before
the full fruits of the computer in the conceptual designing process can be gathered. In
my observation design methodologies in architectonical designing are only reluctantly
used and there is hardly any systematical and methodical account for the originating
process of the design.
Indeed, the bridge between the non-cognitive intuitive design process and the ultra-
systematic computer as a potential design medium, is missing. So then the computer
cannot be used other than a current medium for the hnal development of the design:
It facilitates the drawing, but not the thinking. And, therefore, it cannot be inserted as
a full valued reciprocal design medium which is stimulating from self-esteem. To make
considerations explicit, as is done with methodical designing, does not just advance
insight and clarity in ones own activities. In practice it stimulates the communication
between the ever growing group of professionals which has to co-operate in a building
team, aimed at realizing a specihc building (complex).
Design Phases
Methodologists speak of a hrst phase of conceptual design because of the 3-D concept
with its degree of abstraction, leaving many liberties to choose materials and sub-
systems the architect has at his disposal.
Compared to designers in related technical specialisms (like ship- and aeroplane
designers) the architect has an enormous freedom, through the given freedom of
choosing structural systems, constructions, structures, building components with their
specihc shapes and production techniques, the topological placing of components and
geometrical freedom, and with all that to attain a purposeful sculptural quality of the
building. Seldom we realize how jealous other designers could be of him in this respect.
In order to make a whole new design concept of his building, the architect has (almost
too) many possibilities at his disposal.
The second phase of the process is the materialization design concerns choice of
materials, structural schemes and structural composition up to details. The second
phase is as important as the hrst conceptual phase. As compared with this luxurious
situation, the (poor) aircraft designer knows only one or a few degrees of liberty
of designing every part of the aeroplane because of the high functional and safety
demands. We call this parameter designing: the degree of freedom is only one variation
on one single parameter.
The leap from the conceptual design to the materialized design mainly takes place in
the mind of the designer: sometimes it will be intuitive, often routinely and sometimes
methodical. The execution of an intuitive and non-argumented choice and its perfection
can, nevertheless, very well be done methodically.
After the functional and spatial building concept, a purposeful and efhcient design
process and the development of materialized and technical building components have
become of fundamental importance for the design process of the building. Like the
product designer, who usually operates at the side of the producer, a good project
architect also knows how far he can go as a consumer of building products in the
market and how far he can develop new one-off components to be specially ordered.
He should have insight in the iterative development processes for building products,
systems and components. The interchangeable relation between technical components
and architecture is indispensable for the materialization of the architectonic conceptual
design in an inspiring manner.
Closing Speech - A.C.J.M. Eekhout

111
This conference is the fruit of joining hands between the faculties of Industrial Design
Engineering and Architecture. Although they started their relationship as a mother
and an unwilling daughter, they now seem more like sisters under the skin to quote
Rudyard Kipling.
In the months after the Fire of Architecture the faculty of Industrial Design Engineering
is hosting a number of staff, researchers and students from Architecture, for which
they are extremely thankful. Hopefully also this regrettable cause will contribute to the
intensihcation of the relationship between both worlds, both faculties.
Closing Speech - A.C.J.M. Eekhout

IDE+A
Appendix 1

113
2 Chairmans impressions
Prof. dr. ir. T.M. de Jong (Chair)
Introduction
The main issue of the conference was:
1 the contemporary interrelationship of Industrial Design and Architecture
2 a confrontation of contemporary design practice in both domains with
academic theory and education
Details about eight cases of design processes in practice (four industrial design and four
architecture) were collected by students. The students were stimulated to be not too
strict in handling their query, because practice can raise new questions that may be lost
in a poll structured by academic suppositions. The query of the interviews contained
questions about:
1 the project in general
2 social complexity in collaboration
3 design process
4 decision making
5 visualization
6 project management
7 knowledge diffusion
Specialists regarding the topics 2 till 7 were invited to analyze the cases and to write a
paper from their point of view.
Although the speakers of the conference each rehected on the outcomes of the
interviews, it appeared that some of the experts reviewing the cases missed questions..
This was more than compensated by the resulting rough material which provided some
interesting details beyond the chosen themes.
As an urban designer and ecologist ! learned a lot from the interviews and the rehections
of the speakers. However, from both also raised many remaining questions. Some of
these questions ! addressed the speakers after their lecture. ! will come back on these
topics, but let me hrst cite some statements from the interviews that triggered me
most.
The interviews
From the interviews I selected some interesting propositions to introduce the projects
at the conference. They are listed below in the hrst column with small modihcations for
the sake of readability. In the second column I try to analyze why they triggered me.

114
propositions refIection of the chairman
CePeZeds Westraven building interview
Every advisor has solutions. n an urban design team you often have to
disappoint the advisors, because they
answer partial problems in a whole field of
locally connected, context sensitive
problems.
The architect has to take all ideas to a
higher level.
To create surplus value at a higher level of
scale you need an encompassing concept.
A concept like that changes the whole field
of problems, the field of aims and the
direction of solution.
The architect introduces problems, the
advisor provides solutions.
That is an interesting statement contrary to
the idea that design is always problem
solving. t sometimes creates problems.
Copies are compliments. Here the returning debate about patents
emerges.
Ahrends A230 chair
Clients have questions. That is seldom the only inspiration for the
architect. After all, such questions seldom
survive the life time of a building with
changing users.
Decision making mostly means: 'how large
is the demand'.
For the architect the next users are
unknown. For the urbanist the users are
numerous. For both the stakeholders and
specialists are different for every project.
The sales agency is our antenna. Architects and urbanists seldom do have a
sales agency.
The purchasing agency is an interesting
source.
Architects consult the building materials
documentation and references. Urbanists
their numerous specialists.
We write our program of requirements
ourselves.
The program of requirements develops in
interaction with the design sketches since a
concept changes the whole field of
problems, the field of aims and the
direction of solution for all participants.
n an office, meeting may be more
important than work.
t resulted in a table and chairs suitable for
quick meetings at the table of the host.
Styling is 10% of our work. Does that mean education has to give 90%
of the time to other item?
Neutelings-Riedijks Image and sound
building
The scale of a project is not relevant for
the way of communicating.
But the scale has a relation with the
number of possible participants.
Steps are similar to those taught at TUDelft
+ geographical centered communication.
The plus-sign indicates the many always
different contexts architecture and
urbanism have to operate in:
administrative, cultural, economic,
technical, ecological and spatial contexts at
different levels of scale and changing in
different periods of time.
Chairmans impressions - T.M. de Jong

115
Different mock ups to simulate different
research questions.
nstead of 'mock ups' would like to speak
about 'models', including mathematical
ones. Models are partial representations to
analyze and evaluate partial effects.
All knowledge in architecture is common
knowledge.
Anyone is 'hands-on' expert, but what
about the technical details and changes?
MMIDs Beertender
Beertender is produced in very large
series.
Except in chains of shops architectural
evaluation is avoided by the owner,
because it can harm the value of its
property. Urban evaluation happens after
the retirement of the designer. So, (s)he
does not care so much about the
performance.
My own style isn't important in this project. How different form architectural designers!
Style is work method f-d-p (Functionality &
technology, Design (look & feel),
Production & assembly)
t would be interesting to study the impact
of the work method on style. However, the
work method of an architect changes by
her or his experience. Unfortunately
experienced architects are not aware of
their tacit knowledge, they forget to
mention hidden suppositions beginners do
not share. So, experienced architects often
are bad teachers. However, there are
exceptions where architects are aware of
the roots of their experience ,
e.g.Hertzberger. He saved and numbered
all his sketches, including the instructive
mistakes.
cannot recall decisions that explicitly. Even the movements of my hands are
decisions. To make them all explicit would
hamper designing.
But there have been moments like that
during the project. Time, Money and
Quality.
That is the selection of decisions to make
explicit as asked by the client.

116
Frank Vermeulens Spark
A car consists over more than 1200
components.
Shortly after my chairmanship designed a
device to keep the pages of differently
sized books straight for scanning from
above (see Fig. 1). t contains 22 Meccano-
components and it took 5 hours to restore
all mistakes of putting the rightly-sized
screws at the right place, to change the
size and order of the components and to
understand the rotational behaviour of
adaptable paralellograms in three
dimensions.
Small steps have to restrict high risks. Equal steps do not have equal risks, but
smaller steps are easier to evaluate on
risks. However, steps in designing are
often related to each other making next
steps (im)possible. To evaluate the
possibilities of next steps is not easy, it
supposes design.
Architecture is a specified direction in
product designing.
n Delft, product design was split off from
the Faculty of Architecture into a separate
Faculty. This proposition reverses history,
but upscaling a profession will meet many
problems. The separation was not without
reason.
Media like to attach a name of an architect
to a building.
Architecture and urbanism are public, local
and long term. They are often part of a
public debate, asking for publicly
responsible names. The value of your
property is affected by your neighborhood,
you cannot buy the product without
influencing your neighbors.
A mass product has a lifecycle of one year,
but a building has a lifecycle of 50-100
years.
And if your neigbours sell their property it
still affects the value of your property.
Verheijen and Tabbers' ID building
Small series, big scale difference. Fons Verheijen gave an exellent summary
of differences between ndustrial Design
and Architecture. refer to that interview
below, concluding that the most important
factor is scale.
The first big decision was to decide to do
such a big renovation project, then
deciding upon the final amount of square
meters and where to place which function.
'To preserve or to change' is a crucial
decision, often taken before design. n
urban design and architecture it is more
precise: 'what to preserve and what to
change'. These decisions are often the
consequence of a program of
requirements. This proposition suggests
they precede it.
The whole idea, to create one big space in
which everybody would be able to enjoy
what others are doing, was one big risk.
n general, connecting causes more risks
than separating.

117
Indes' care bed
Not much attention was given to
aesthetics.
With such a firm statement not many
architects would survive as designers.
Users played an important role, from the
start they were consulted and later they
were involved when prototypes had to be
tested.
n architecture the expensive testing in a
prototype is replaced by referring to
preceding examples ('precedents').
The people involved in the engineering
phase are already looking over the
shoulder during the concept development
stage.
That practice penetrates in Architecture,
but its scale forces to choose cheap, less
mobile materials in less complicated
mechanisms.
For an industrial designer a mock-up is a
category of models
Conclusion?
Extremely interesting: difference between
historical relations and causal relations
s that the reason?
These propositions already raise many subjects discussed in the conference
Scale (frame and grain)
The larger scale of architecture and urbanism causes may other differences from
product design:
1. a prominent role of gravity: vertical structures with horizontal
hoors and connecting spaces;
2. a ground bound unique, always different and specihc
governmental, cultural, economic, technical, ecological and
spatial context;
3. small series, many external parties, different by context;
4. boundaries of prefabrication by transport possibilities;
5. many solutions for the same overall problem: to climatize,
separate or combine activities;
6. changing scale changes terms and legend units of the drawing;
7. upscaling in space and time affects the composition of the team;
8. upscaling decreases decision making based on the size of the
demand and pay-back time.
Shortly after my chairmanship I designed a device to keep straight for scanning
from above the pages of differently sized books (see Fig. 1). As an urbanist with a
task to teach technical ecology, I wanted to understand the difference of designing
at the largest scale form product design by doing:
1. In this product gravity plays a role, but not a prominent one
much inhuencing the construction. There are two horizontal
planes, but the vertical structures are mobile.
2. !t is not bound to different and specihc governmental, cultural,
economic, technical, ecological and spatial contexts.
3. There were no external parties, but if it had to be produced in
large series much effort still would have to be done and more
parties would have to be involved.
4. There are no boundaries of prefabrication.

118
5. The problem has a limited number of solutions.
6. The character of the legend units are in the range of architecture,
but drawing an urban plan requires an other kind of legend, other
categories and other ways of thinking.
7. There was no team, but see 3.
8. There was no decision making based on the size of the demand
and pay-back time, but see 3.
Figure 1: A device to keep the pages of differently sized books straight for scanning from above,
designed by an ecological urbanist.
The reections of the speakers
I will not summarise the contributions of the speakers here, but I will make some
additional remarks probably useful for further inquiry.
Social complexity in collaboration
The integration of a group compared to its integration in a larger context is proportional
to the time budget they spend internally and exernally. If management askes for many
external contacts, the result is sprawl of effort increasing internal entropy.
Design process
A short term goal is a long term means. A goal is a design. So, design cannot be goal
directed. Design directed design does not say much. Engineering is design driven
research to solve problems risen by design. So, design also raises problems to be solved
by engineering. Engineering is the problem solving activity.
Design creates improbable possibilities. So, it changes desirable futures, changing the
expected undesirable futures, the held of problems. There is never one goal or one
problem. There is always a held of problems and goals.
Solving one problem creates new problems. Reaching one aim creates new goals.
Wicked problems are not wicked they are simply helds of related problems. Design
does not solve

119
Decision making
!s everything a decision?
Visualisation
Schetching is another language.
Project management
Internal integration causes external disintegration and the reverse at any level of
scale.
Lack of time causes specialisation. Specialisation saves time, integration saves space.
Knowledge diffusion
!ntuition is experience to tranferable by words. What is knowledge?
Discussion
See questions

IDE+A
Appendix 2

121
Program
10.00- 10.10 Introduction
Prof. dr. C.J.P.M. de Bont
Dean of the Faculty of Industrial Design Engineering, TU Delft
10.10- 10.30 Explanation Cases
Professor, Chair of Environmental Planning and Ecology
Faculty of Architecture, TU Delft
10.30- 11.00 Design Processes
Dr.ir. H.H. Achten
Assistant Professor, Architectural Modeling
Faculty of Architecture, TU Eindhoven
11.00- 11.30 Coffee/Tea
11.30- 12.00 Visualization
Prof. G. Goldschmidt
Professor,The Mary Hill Swope Chair in Architecture & Town Planning
Faculty of Architecture and Town Planning, Israel Institute of Technology
12.00- 12.30 Project Management
Prof. dr. ir. J.W.F. Wamelink
Professor Design- and Construction
12.30- 13.30 Lunch

122
13.30- 14.00 Social Complexity in Collaboration
Prof. dr. P.G. Badke-Schaub
Professor Design Theory and Methodology
Faculty of Industrial Design Engineering, TU Delft
14.00- 14.30 Decision Making
Dr. ir. P.P.J. van Loon
Associate professor Design and Decision Systems
14.30- 15.00 Tea/Coffee
15.00- 15.30 Technology Diffusion
Dr. ir. W.A. Poelman
Associate Professor Product Development
15.30- 16.00 General Discussion
16.00- 16.10 Afterword
Prof. ir. W. Patijn
Dean faculty of Architecture, TU Delft



Architect

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Architect

Uploaded by

Copyright:

Available Formats

Design processes

This page intentionally left blank

You might also like