Professional architects’ design trade-offs in reinforcement learning-powered agent-based simulation

Abstract

While emphasising play-based approaches in early childhood education, there is a deficiency in the practical understanding of play integration into children’s learning environment design. To support methodological limitations, this study developed a play behaviour model using reinforcement learning agents to generate children’s unpredictable behavioural responses during physical and social play, including unforeseen injuries. In comparative laboratory experiments, 14 professional architects engaged in design tasks before and after using the simulation. The design results were assessed through external expert evaluations, with a focus on the optimal trade-offs between facilitating play behaviours while ensuring safety considerations and on the level of effective design decision making and completeness. The statistical analysis demonstrated that the simulation facilitated optimal trade-offs, enhanced decision making and improved the overall completeness of children’s learning environment design. This research addressed the capabilities of simulation to support a child-centric design, with potential implications for improving early childhood education environments.

KEYWORDS:

• Agent-based simulation
• Reinforcement learning
• Design trade-offs
• Child-centric design performance
• Play-based learning environment design
• Expert evaluation

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data available on request from the authors.

Funding details

This work was supported by the National Research Foundation of Korea (NRF) grants funded by the Ministry of Education [grant number NRF-2021R1A2C1004608]; and the Korea Agency for Infrastructure Technology Advancement (KAIA) grant funded by the Ministry of Land, Infrastructure, and Transport [grant number RS-2021-KA163269].