ePrints

Integrating human behaviors in an agent-based model for multi-floor building emergency evacuation

Lookup NU author(s): Dr Sarah Dunn

Downloads

Full text for this publication is not currently held within this repository. Alternative links are provided below where available.

Abstract

© 2025 Elsevier LtdThis study develops a more comprehensive and improved agent-based model (ABM) for simulating emergency evacuation from multi-floor buildings. The model allows agents to make individual decisions by interacting with building environment, while also considering social behaviors that involve interactions between group members (agents within a group with pre-existing relationships), between neighboring agents (unknown agents in each other's field of vision) and between evacuee agents and authority agents (agents acting as authority figures). The ABM implemented is underpinned by theory and practice, in which agent movement on floors and in staircases is validated. Simulation experiments in a virtual four-floor building identified the key location for placing exits. The absence of exits at this location caused some agents to move longer distances to evacuate, increasing their evacuation time, despite the presence of additional exits. Also, although herding behavior is often criticized for causing congestion, it reduced the evacuation time when agents lacked knowledge of complex building layout. Further, higher intimacy level of groups spent more time waiting for group members, thereby increasing their evacuation time. However, groups with ‘low’ intimacy level did not always evacuate faster, since agents in these groups separated and moved to a distant known exit. Finally, placing authority agents at random rather than task-specific locations reduced the evacuation time because they encouraged more evacuee agents to begin their evacuation on the way to the task location. The model supports building designers, safety engineers, facility managers, and emergency planners in assessing building layout designs, evaluating evacuation strategies and enhancing preparedness in complex building environments.

Publication metadata

Author(s): Zhang X, Dunn S, Luo Y, Shen Y

Publication type: Article

Publication status: Published

Journal: Journal of Building Engineering

Year: 2025

Volume: 112

Print publication date: 15/10/2025

Online publication date: 07/08/2025

Acceptance date: 06/08/2025

ISSN (electronic): 2352-7102

Publisher: Elsevier Ltd

URL: https://doi.org/10.1016/j.jobe.2025.113716

DOI: 10.1016/j.jobe.2025.113716

Altmetrics

Share