AI & SOCIETY

, Volume 30, Issue 4, pp 493–507

Simulating effects of signage, groups, and crowds on emergent evacuation patterns

  • Mei Ling Chu
  • Paolo Parigi
  • Jean-Claude Latombe
  • Kincho H. Law
Original Article

DOI: 10.1007/s00146-014-0557-4

Cite this article as:
Chu, M.L., Parigi, P., Latombe, JC. et al. AI & Soc (2015) 30: 493. doi:10.1007/s00146-014-0557-4

Abstract

Studies of past emergency events have revealed that occupants’ behaviors, egress signage system, local geometry, and environmental constraints affect crowd movement and govern the building evacuation. In addition to complying with code and standards, building designers need to consider the occupants’ social characteristics and the unique layout of the buildings to design occupant-centric egress systems. This paper describes an agent-based egress simulation tool, SAFEgress, which incorporates important human and social behaviors observed by researchers in safety and disaster management. Agents in SAFEgress are capable of perceiving building emergency features in the virtual environment and deciding their behaviors and navigation. In particular, we describe four agent behavioral models, namely following familiar exits, following cues from building features, navigating with social groups, and following crowds. We use SAFEgress to study how agents (mimicking building occupants) react to different signage arrangements in a modeled environment. We explore agents’ reactions to cues as an emergent phenomenon, shaped by the interactions among groups and crowds. Simulation results from the prototype reveal that different designs of building emergency features and levels of group interactions can trigger different crowd flow patterns and affect overall egress performance. By considering the occupants’ perception about the emergency features using the SAFEgress prototype, engineers, designers, and facility managers can study the human factors that may influence an egress situation and, thereby, improve the design of SAFEgress systems and procedures.

Keywords

Crowd simulation Egress simulation Building egress Social agents Social behavior Collective behavior Simulated perception 

Copyright information

© Springer-Verlag London 2014

Authors and Affiliations

  • Mei Ling Chu
    • 1
  • Paolo Parigi
    • 2
  • Jean-Claude Latombe
    • 3
  • Kincho H. Law
    • 1
  1. 1.Department of Civil and Environmental EngineeringStanford UniversityStanfordUSA
  2. 2.Sociology DepartmentStanford UniversityStanfordUSA
  3. 3.Kumagai Professor Emeritus, Computer Science DepartmentStanford UniversityStanfordUSA

Personalised recommendations