Observations of designer behaviors in complex system design

Abstract

The design of large-scale engineering systems requires design teams to balance a complex set of considerations. Formal approaches for optimizing complex system design assume that designers behave in a rational, consistent manner. However, observation of design practice suggests that there are limits to the rationality of designer behavior. This paper explores the gap between complex system designs generated via formal design process and those generated by teams of human designers. Results show that human design teams employed a range of strategies but arrived at suboptimal designs. Analysis of their design histories suggest three possible causes for the human design teams' performance: poorly executed global searches rather than well executed local searches, a focus on optimizing single design parameters, and sequential implementations rather than concurrent optimization strategies.