Physics-Based Simulation for Construction Activity Sequence Planning

Physics-Based Simulation for Construction Activity Sequence Planning

2026

DOI: https://doi.org/10.60164/ojo96vp61

Authors: Mohammad Rezaul Karim, Yasser Mohamed

Citation:

Karim, M. R., and Mohamed, Y. (2026). Physics-Based Simulation for Construction Activity Sequence Planning. Lean Construction Journal 2026 Special Issue CRC-CSCE pp 170-189 https://doi.org/10.60164/ojo96vp61 (Submitted 10Jan2026; Accepted 11Sep2026) www.leanconstructionjournal.org

Abstract:

Research Question: How can construction activity sequences be generated automatically based on physical constructability constraints, rather than relying on predefined precedence logic embedded in conventional scheduling methods?

Purpose: To develop a physics-based simulation framework in which construction activity sequences emerge from physical constraint satisfaction.

Research Method: A BIM-derived industrial module is transformed into a physics-enabled representation by exporting component geometry and enriching it with construction metadata for rigid-body simulation. A PyBullet-based environment models gravity, contact, collision, support, and spatial clearance constraints. Activity sequences are generated using a baseline brute-force search in which component placements are iteratively evaluated for physical feasibility and reordered when violations occur. Seven module configurations of increasing complexity are tested to assess feasibility and computational behavior.

Findings: The results show that physically executable construction sequences can be generated through physics-based evaluation of stability and installation clearance without predefined precedence rules. Brute-force search scales poorly with increasing complexity, indicating limited practicality at larger scales.

Limitations: The implementation is limited by the poor scalability of brute-force search and simplified construction assumptions, including prefabricated components, idealized installation paths, and simplified connection behavior. Temporary works and non-physical planning constraints such as trade coordination, productivity, safety, and regulatory requirements are not modeled. In addition, missing physics-relevant attributes in BIM models require assumptions and manual data augmentation during BIM-to-simulation transformation.

Implications: The results suggest that physics-based simulation can function as a practical feasibility layer between design models and detailed schedules, allowing sequencing logic to be tested against real physical constraints before being finalized. This creates a pathway for integrating constructability checks earlier in planning, rather than addressing issues during execution.

Value for Practitioners: The approach offers a way to test and refine proposed sequences in a virtual environment, supporting early exploration of alternative construction strategies and reducing reliance on trial-and-error during planning. It can be used alongside existing tools to improve confidence in sequence feasibility prior to detailed scheduling and execution.

Keywords: Activity sequencing; physics-based simulation; constraint-satisfaction process; brute-force search

Paper type: Full paper