A user study of interpretability of engineering design representations

Abstract

In order to save time and money, designers need to understand the utility of different engineering representations. To this end, designers often rely on past experience to decide which model to construct; yet students without this experience have no help. Interestingly there are noticeable gaps in the research literature with respect to how and when to select representations for engineering design. This paper examines the differences between three types of engineering representations, specifically sketches, drawing packages, and physical prototypes. The amount of information designers can extract, in addition to the correctness and confidence, from these representations is studied. Design reviews of concepts with respect to requirements verification serves as the design task of this investigation. The data from this user study are analysed, using descriptive and non-parametric statistics. The results reveal that designers are more confident and correct in making conclusions about whether a design meets requirements when using high-fidelity representations and physical representations, specifically high-fidelity prototypes. Low-fidelity representations appear to be useful for determining if a design meets functional requirements, but not geometric or manufacturing requirements. The relationship between drawing packages and low-fidelity prototypes is still unclear and thus is an area for further research.

Keywords:

• prototyping
• representations
• interpretability
• user study