Understanding human queuing behaviour at exits: an empirical study

Area of Science:

• Civil Engineering
• Computational Social Science
• Human Behavior Modeling

Background:

• Pedestrian modeling and simulation are crucial for facility design and safety.
• Empirical data supporting simulation models for pedestrian egress behavior is limited.
• Understanding exit choice dynamics is essential for efficient crowd management.

Purpose of the Study:

• To investigate pedestrian exit choice behavior in various facility geometries.
• To provide empirical evidence for validating and improving pedestrian simulation models.
• To develop a simplified, effective model for predicting pedestrian egress in congested conditions.

Main Methods:

• Conducted experiments in diverse geometric configurations with varying door widths, pedestrian numbers, and initial locations.
• Collected data on pedestrian perception, distance to exits, and local density levels.
• Analyzed pedestrian behavior using statistical methods and simulation.

Main Results:

• Pedestrians utilize time-dependent information to make optimal exit choices.
• Congested conditions lead to load balancing across available exits.
• A minimal linear model based on exit distance and density accurately predicts egress behavior.

Conclusions:

• Pedestrian exit choice is a dynamic process influenced by environmental cues.
• A simplified linear model offers an efficient alternative to complex optimization methods for simulating load balancing.
• The findings are particularly relevant for non-symmetrical facility designs and emergency egress scenarios.