Mining Functional Model Graphs to Find Product Design Heuristics With Inclusive Design Illustration

Abstract

Engineering design heuristics offer the potential to improve the design process and resultant designs. Currently, heuristics are empirically derived by experts. The goal of this paper is to automate the heuristics generation process. Functional modeling, a well-established product representation framework, is applied in this research to abstract the intended functionality of a product. Statistically significant heuristics, extracted from a database of functional models, serve as design suggestions or guidelines for concept generation. The heuristics can further be applied to automate portions of the concept generation process. Prior research efforts in automated concept generation rely heavily on the design repository. The repository needs to be appended for broader categories of design problems, and, at the same time, a tool for quick analysis of the expanded repository is required. An automated heuristic extraction process has the capability to efficiently mine the updated repositories and find new heuristics for design practice. A key objective of this research is to develop design heuristics applicable in the diverse and challenging domain of inclusive design. The research applies graph theory for mathematical representation of the functional model, graph visualization for comprehending graphs, and graph data mining to extract heuristics. The results show that the graphical representation of functional models along with graph visualization allows quick updates to the design repository. In addition, we show that graph data mining has the capability to efficiently search for new design heuristics from the updated repository.