Engineering Design is a Decision-making Process

Sundar Krishnamurty
Department of Mechanical and Industrial Engineering
University of Massachusetts-Amherst
NSF Open Workshop on Decision Based Design
13-14 September 1997

Engineering design is essentially a part of the product realization process by which new or modified products are conceived, developed, designed, manufactured, and brought to market. It then becomes apparent that design engineers must be knowledgeable about and competent to contribute, not just to one aspect of design or design based primarily on the principles of engineering science, but also to the series of decisions and their consequences associated with the entire product realization process. As such, engineering design is essentially a decision making process that requires rigorous evaluation and comparison of design alternatives from a global perspective on the basis of different classes of design criteria.

Therefore, engineering design should stress a fundamental and scientific understanding of design based on the entire product realization process. The research agenda for engineering design should address all aspects of design including modeling, configuration and parametric design, design for manufacturing, and material selection. Issues such as tolerances and uncertainty in data need to be accounted for in some manner. Design strategies must be devised to specifically address the inherent complexity arising from representing physical design problems using idealized computer-based abstractions (computer models). In addition, the research agenda should also focus on the strategic integration of these different design stages. Such a research agenda can then lead to the identification and development of a consistent and recognized body of design principles with vital and respected research components that will play a crucial role in enabling engineers make intelligent decisions towards improving the overall quality of the products designed. Currently, our research group is working towards establishing such a rigorous framework for decision making in the context of engineering design by exploring concepts from multiattribute utility theory and through its integration with statistical exploration based robust optimal design strategies. Our goal is to develop a sound, practical trade-off based robust modeling and design methodology to quantitatively incorporate qualitative knowledge and preferences of multiple, conflicting attributes without loss of generality and accuracy under conditions of risk and uncertainty. We expect the results from this research to advance the state of knowledge by which decisions can be made in engineering design and to lead to a better understanding of the consequences of such design decisions from an overall perspective.