A hypernetwork-based context-aware approach for design lesson-learned knowledge proactive feedback in design for manufacturing

Abstract

Design lesson-learned knowledge (DLK) clearly describes various design quality problems exposed in past manufacturing stages, solutions and preventive measures. If knowledge management system could proactively feed DLK back into design process, many previous quality problems would be avoided, thus helping designers better implement design for manufacturing (DFM) in a smarter manner. However, since design quality problems are not pointed out in design process, problem-relevant information extracted from design contexts might be inaccurate. In this situation, traditional context-aware approach is prone to acquire design quality problems that designers do not need. Facing these challenges, a hypernetwork-based context-aware DLK proactive feedback approach is proposed to construct hypernetwork-based DLK representation model and predict possible design quality problems in the design process, thus providing corresponding DLK and helping designers reduce the reoccurrence of previous quality problems in DFM. Specifically, hypernetwork-based DLK representation model is first constructed, which consists of designer context network, task context network and DLK network. Based on this model, a context-aware collaborative reason strategy is constructed to predict possible design quality problems according to complex design contexts. To validate the proposed approach, a practical product development case on the shipbuilding design is implemented, and some comparative experiments are conducted. Experiment results show the proposed approach is effective and has a positive performance in DFM. It is anticipated this work opens up a promising way to help designers reuse DLK for reducing the reoccurrence of previous design quality problems in a smarter manner, thus better implementation of DFM.