Scenario-based anticipatory failure determination and patent technology inspiration for product innovation design

Abstract

Exploring opportunities of technologies for product is critical in the highly competitive market. Anticipatory failure determination (AFD) identifies likely failure modes of a product, which provides a tool to explore the technology opportunities for product improvement. The existing methods of AFD are limited in the failure analysis of system components and the acquisition of technical solutions for the failure cause is depending on designer experience. This research proposes an approach to product innovation design based on the scenario-based evolution of AFD and patent technology inspiration. Scenarios obtained based on the proposed scenario generation rules are used to form failure events from the product function level. The failure events are then used to explore technical opportunities based on the anticipatory failure determination. A word vector model is trained with the patent data to extract patent technology terms in combination with the semantic similarity of the functional basis. The Latent Dirichlet Allocation (LDA) is used to cluster patent technologies. The best technical solution is finally selected based on the relative technology maturity and technological distance. The TRIZ tool is used to resolve possible conflicting problems. A complete product innovation design process is proposed from the technology opportunity exploration to concept design formation.