Enhancing design adaptation through an information-enriched reinforcement learning state
Utz Y, Wartzack S, Götz S (2026)
Publication Type: Conference contribution
Publication year: 2026
Journal
Publisher: Cambridge University Press
Book Volume: 6
Pages Range: 867-876
Conference Proceedings Title: Proceedings of the Design Society
Event location: Dubrovnik
Abstract
The applicability and scalability of design adaptations utilizing reinforcement learning can be broadened by using graph-based approaches instead of rigid vector- or grid-based ones. However, graph-based approaches often require a high number of simulations to converge. To reduce the simulation effort in the mechanical optimisation, the reinforcement learning setup is enriched with task-specific causal and physically based information. This work systematically investigates the influence of this additional information on the efficiency of design adaptations using a factorial test design.
Authors with CRIS profile
Yannick Utz
Chair of Engineering Design (KTmfk)
Sandro Wartzack
Chair of Engineering Design (KTmfk)
Stefan Götz
Chair of Engineering Design (KTmfk)
How to cite
APA:
Utz, Y., Wartzack, S., & Götz, S. (2026). Enhancing design adaptation through an information-enriched reinforcement learning state. In Mario Storga, Stanko Skec, Tomislav Martinec, Dorian Marjanovic, Neven Pavkovic (Eds.), Proceedings of the Design Society (pp. 867-876). Dubrovnik, HR: Cambridge University Press.
MLA:
Utz, Yannick, Sandro Wartzack, and Stefan Götz. "Enhancing design adaptation through an information-enriched reinforcement learning state." Proceedings of the 19th International Design Conference, DESIGN 2026, Dubrovnik Ed. Mario Storga, Stanko Skec, Tomislav Martinec, Dorian Marjanovic, Neven Pavkovic, Cambridge University Press, 2026. 867-876.
BibTeX: Download