Toward a Method for Achieving Synergy between Heuristic Rules of Thumb and Quantitative Methods in Engineering Design

University Of Illinois At Urbana-Champaign, Urbana IL

Investigators

Abstract

Engineering design experts' heuristic "rules of thumb" usually save time and money when making design decisions. Through years of experience with product design, materials, and manufacturing, experts can perform many design tasks quickly, such as narrowing the set of possible design options to investigate further. Sometimes, however, these shortcuts result in poor design decisions that waste time and money. This happens especially when new materials, manufacturing technologies, or types of products are being considered. The old rules of thumb no longer apply, and even experts can be vulnerable to "cognitive biases," or decision traps that we all sometimes fall into. Although mathematical models and approaches can alleviate this problem, they often require significant cost and effort to implement and therefore counteract the desired benefit. This project will benefit society by combining the best of both worlds, working toward a methodology for determining the best balance between rapid heuristic strategies and more rigorous quantitative methods for making design decisions. The intellectual significance of this project lies in building a foundation for a new, formal design methodology that exploits the synergy between heuristic and normative methods. This project will analyze a design process for one advanced systems engineering problem, where the designer is required to expand the frame of analysis to include issues never before considered. Normative methods such as optimization and decision-based design consider efficiency and tradeoffs under uncertainty with the same degree of analytic rigor that was previously applied only to physical system modeling. While this approach has been successful to some extent, there are times when the cost of this approach (formulation, data gathering, solving) is simply not worth the effort. Conversely, the heuristic "rules of thumb" employed by design experts are faster, but can sometimes lead to suboptimal results because of systematic cognitive biases that are inadvertently embedded in the heuristic and lead the design choice away from the optimal solution. The problem addressed here is that strictly heuristic design methods can lead to inferior solutions, while strictly normative methods can be too time consuming and costly to implement. Observed heuristics will be analyzed to determine their effect on both the design process and the resulting product. Normative methods will be analyzed in the same manner. Then, a method for defining the best combination of the two approaches will be developed. This project will provide the foundation for developing a full set of "rules for rational design analysis" for determining the best combination of normative and heuristic approaches.