Innovation in graduate projects: Learning to identify critical functions

Design-by-analogy is considered to be a powerful tool for engineering design. The difficulty of finding suitable analogies for solving a given design problem gives rise to the current efforts on computational tools for analogy-based design. In searching for analogies, the critical functions in a design problem are potential search criteria. The study described in this paper investigates whether novice designers identify an expert-derived critical function in a design problem under three scenarios: when they are asked to report the important functions in the problem, when they are directly asked to report the critical function and when they are asked to use design-by-analogy. It is observed that student designers do not identify the expert-derived critical functions when directly asked or when asked to list important functions. However, they inherently use the expert-derived critical functions in their analogical mapping process. This suggests that, during analogical reasoning, designers tend to identify and use the same critical functions regardless of experience, and also that critical functions are valid search criteria for deriving analogies from a computational database. This insight is highly valuable for current efforts to develop computational tools for analogical reasoning.