Initial Steps Toward an Analogy Retrieval Tool Based on Performance Specification

In design-by-analogy, inventors draw inspiration from natural and man-made systems to create new, often innovative engineering products. Though methods exist to aid concept generation through design-by-analogy, they are not built on fundamental knowledge of how inventors inherently use analogies. Such a foundation is critical for developing effective analogy-finding tools and methods. In this research, we pursue an empirical product study to guide the creation of a computational tool that presents relevant analogies based on designers’ objectives. Using an inductive approach, we studied 57 design-by-analogy products and their inspiring analogs to identify how analogies entered the design process. We developed classification schemes that characterize the analogies and their design contexts, which includes the inventors’ field of work and design objectives. Findings emerged from identifying patterns in the classification results. When comparing academic and commercial inventors, we found differences in how they use analogies to find new product functions or gain better performance. We gained insight into how knowledge-driven and problem-driven analogy usage cases differ. Most intriguingly, we found that products commonly borrow critical functions directly from analog systems, but we also discovered instances of critical function inversion, where the sense of the critical function is reversed between the analog and the product. For example, this occurred with the ECO-Auger Tidal Turbine (critical function: “Convert fluid flow into rotation”) which was inspired by screw conveyors (“Convert rotation into material flow”). This discovery implies that a computational search for analogs should expand from a designers’ function specification to include analogs with inverted critical functions, thus widening the space of possible inspiration sources.Copyright © 2013 by ASME