Simulating stakeholder collaboration : generating synthetic BCF data with multi-agent systems

Tools

Kandler, Benedikt Justin and Brötzmann, Jascha and Rüppel, Uwe; Moreno-Rangel, Alejandro and Kumar, Bimal, eds. (2025) Simulating stakeholder collaboration : generating synthetic BCF data with multi-agent systems. In: EG-ICE 2025. University of Strathclyde Publishing, GBR, pp. 496-505. ISBN 9781914241826 (https://doi.org/10.17868/strath.00093270)

[thumbnail of Kandler-etal-EG-ICE-2025-Simulating-stakeholder-collaboration-generating-synthetic-BCF]
Preview
 Text. Filename: Kandler-etal-EG-ICE-2025-Simulating-stakeholder-collaboration-generating-synthetic-BCF.pdf
Final Published Version
License: Creative Commons Attribution 4.0 logo
Download (1MB)| Preview

Abstract

This paper presents a novel framework for generating synthetic Building Collaboration Format (BCF) data through Large Language Model (LLM)-driven Multi-Agent Systems (MAS). Addressing data scarcity in BIM-related issue tracking, the approach simulates realistic stakeholder communication based on role-specific agents and enriched prompts. The workflow combines semi-automated issue extraction, dynamic agent orchestration, and structured evaluation. Results demonstrate that MAS can emulate authentic construction discourse, though limitations in realism and context integration persist. The study outlines pathways for enhancing tool integration, memory structures, and agent behavior, paving the way for future simulations of issue impact, knowledge management, and risk mitigation in construction projects.

  • Item type:Book Section
    ID code:93270
    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 10:53
    Last modified:06 Aug 2026 00:05
    URI:https://strathprints.strath.ac.uk/id/eprint/93270
CORE (COnnecting REpositories)