Scholarlink

<p>This paper explores the concept of integrating generative AI and Monte Carlo Tree Search (MCTS) into facade design processes. The integration is envisioned as a collaboration between the machine and architect, leveraging the unique strengths of each. The proposed workflow uses MCTS to estimate the potential impact of design decisions on embodied carbon and thermal performance. Selected solutions are visually presented to the architect using generative AI. A simplified orthogonal architecture facade design process validates this approach. Our key findings indicate that the MCTS-assisted workflow may facilitate a broader exploration of design solutions and supports informed decision-making in the early design stages. The integration of qualitative visual feedback alongside quantitative data offers architects a more holistic understanding of design options. </p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AI-Augmented Parametric Façade Design: Exploring MCTS for Early-Stage Decision-Making

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年6月26日

查看全文
  • Victor Yu Chieh Lin,
  • Shen-Guan Shih

摘要

This paper explores the concept of integrating generative AI and Monte Carlo Tree Search (MCTS) into facade design processes. The integration is envisioned as a collaboration between the machine and architect, leveraging the unique strengths of each. The proposed workflow uses MCTS to estimate the potential impact of design decisions on embodied carbon and thermal performance. Selected solutions are visually presented to the architect using generative AI. A simplified orthogonal architecture facade design process validates this approach. Our key findings indicate that the MCTS-assisted workflow may facilitate a broader exploration of design solutions and supports informed decision-making in the early design stages. The integration of qualitative visual feedback alongside quantitative data offers architects a more holistic understanding of design options.