Towards automated BIM conflict resolution using reinforcement learning

Tools

Jiang, Yuye and Du, Changyu and Wu, Jiabin and Nousias, Stavros and Borrmann, André; Moreno-Rangel, Alejandro and Kumar, Bimal, eds. (2025) Towards automated BIM conflict resolution using reinforcement learning. In: EG-ICE 2025. University of Strathclyde Publishing, GBR, pp. 97-107. ISBN 9781914241826 (https://doi.org/10.17868/strath.00093289)

[thumbnail of Jiang-etal-EG-ICE-2025-Towards-automated-BIM-conflict-resolution-using]
Preview
 Text. Filename: Jiang-etal-EG-ICE-2025-Towards-automated-BIM-conflict-resolution-using.pdf
Final Published Version
License: Creative Commons Attribution 4.0 logo
Download (1MB)| Preview

Abstract

The established practice in construction planning is based on design activities of the different disciplines being conducted in parallel, leading to potential conflicts that need to be resolved in coordination sessions. In BIM projects, this principle is termed federated modeling approach when referring to the process of combining multiple individual discipline-specific models into a single, coordinated model for reference, clash detection, and decision-making. Although BIM coordination tools can detect conflicts automatically, resolving them remains a time-consuming manual process requiring iterative design coordination. To address this challenge, this paper proposes a Reinforcement Learning (RL)-based method for automated geometric conflict resolution. We implement a Proximal Policy Optimization (PPO) algorithm in a BIM environment, training the RL agent to resolve conflicts using real-time feedback from a rule-based model checker. The method's feasibility is evaluated across scenarios of varying complexity. The results demonstrate that the agent learns effective conflict resolution strategies, offering a valuable step beyond model checking towards automatic conflict resolution. The code is available at https://github.com/YuyeJ48/Towards-Automated-BIM-Conflict-Resolution-Using-Reinforcement-Learning

  • Item type:Book Section
    ID code:93289
    Dates:
    Date
    Event
    1 July 2025
    Published
    9 June 2025
    Accepted
    Subjects:Fine Arts > Architecture
    Department:Faculty of Engineering > Architecture
    Depositing user: Pure Administrator
    Date deposited:27 Jun 2025 15:03
    Last modified:01 Aug 2026 00:16
    URI:https://strathprints.strath.ac.uk/id/eprint/93289
CORE (COnnecting REpositories)