Data-driven Intelligent Computational Design for Products: Method, Techniques, and Applications

Abstract Data-driven intelligent computational design (DICD) is a research hotspot that emerged under fast-developing artificial intelligence. It emphasizes utilizing deep learning algorithms to extract and represent the features hidden in historical or fabricated process data then learn combination mapping patterns of these for solution retrieval, generation, optimization, evaluation, etc. Due its capability automatically efficiently generating solutions thus supporting human-in-the-loop innovative activities, DICD has drawn attention both academic industrial fields. However, as an emerging subject, many unexplored issues still limit development application DICD, such specific dataset building, engineering design-related feature engineering, systematic methods techniques implementation entire product process, In this regard, operable road map from full-process perspective established, including general workflow project planning, overall framework implementation, common mechanisms calculation principles during key enabling technologies detailed three case scenarios application. The can help researchers locate their directions further provide guidance engineers applications.