Increasing Design Search Relevancy using Statistical Methods in Conjunction with Ontologies

Abstract

In engineered products, one of the most critical business processes is managing and reusing product design information within a product and across a product family (PF). Reuse of existing component designs plays an important role in reducing product design lead-time while lowering cost and increasing quality. Towards this end we propose a ontology-based design search and reuse method that combines statistical methods with ontologies to increase the component design search relevancy when having existing product design repositories. This proposed ontology-based design search and reuse method is performed in three stages: (1) computation of statistical similarity between the desired design and existing designs, (2) efficiency assessment of the remaining component designs based on a predefined component ontology, and finally (3) configuration of the selected design(s) for reuse. This method searches and ranks existing component designs in a design repository using statistical similarity and/or uses ontologies to expand the design search. This method is illustrated with a case study involving single-use cameras