MODELING A STRATEGIC HUMAN ENGINEERING DESIGN PROCESS: HUMAN-INSPIRED HEURISTIC GUIDANCE THROUGH LEARNED VISUAL DESIGN AGENTS. (May 2020)

• Record Type: Journal Article
• Authors: Puentes, L.; Raina, A.; Cagan, J.; McComb, C.
• Is Part Of: Proceedings of the Design Society: DESIGN Conference. Volume 1(2020)
• Page Start: 355
• Page End: 364
• Publication Date: 2020-05
• Subjects: artificial intelligence (AI) -- engineering design -- design automation -- deep learning -- heuristics; Engineering design -- Congresses; Industrial design -- Congresses; 620.0042
• DOI: 10.1017/dsd.2020.42
• Languages: English
• ISSNs: 2633-7762

Abstract

Abstract: Human designers often work in a visual design space, projecting step-by-step design progression through evolving mental images. The strategic evolution of that design leverages heuristics based on experience and domain knowledge. The methodology presented in this paper brings together the visual nature of design problem solving and design heuristics in a deep learning computational agent framework that emulates and enables human-mirrored design. When applied to a truss design task, results demonstrate superior results to those of human designers who provided the initial data.