%0 Conference Paper %K Design %K Computer %K Decision %K Intelligence %K Judgement %K Knowledge %K Plan %K Quality theory %K Tool %A Konstantinos M Papamichael %A Jean-Pierre Protzen %B Focus Symposium on "Computer-Assisted Buildong Designs Systems", of the Fourth International Symposium on System Research, Informatics and Cybernetics %C Baden-Baden, Germany %D 1993 %G eng %T The Limits of Intelligence in Design %2 LBL-31742 %3

474701

%4

August 3-4, 1993

%#

Design & Retrofit Tools

%$

Commercial Building Systems

%]

y

%8 05/1993 %X

A new, comprehensive design theory is presented, applicable to all design domains such as engineering and industrial design, architecture, city and regional planning, and, in general, any goal-oriented activity that involves decision making. The design process is analyzed into fundamental activities that are characterized with respect to the nature of knowledge requirements and the degree to which they can be specified and delegated to others, in general, and to computers in particular.

Throughout the history of research in design theories and methods, from the 1940s with operations research and optimization, through the 1960s with the characterization of design problems as "wicked," or "ill-defined," design has been understood as a rational activity, that is "thinking before acting." The new theory presented in this paper suggests that design is "thinking and feeling while acting," supporting the position that design is only partially rational. Intelligence, "natural" or "artificial," is only one of two requirements for design, the other being emotions. Design decisions are only partially inferred, that is, they are not entirely the product of reasoning. Rather, design decisions are based on judgment that requires the notion of "good" and "bad," which is attributed to feelings, rather than thoughts.

The presentation of the design theory extends to the implications associated with the limits of intelligence in design, which, in-turn, become constraints on the potential role of computers in design. Many of the current development efforts in computer-aided design violate these constraints, especially in the implementation of expert systems and multi criterion evaluation models. These violations are identified and discussed in detail. Finally, specific areas for further research and development in computer-aided design are presented and discussed.