Experimental and agent-based simulation analysis of collective escape in mice: Validation and density-dependent effects of obstacles.

Understanding collective movement under stress provides insight into the mechanisms governing escape dynamics. This study investigated how multiple environmental and group factors jointly affect the escape dynamics of mice under emergency-like conditions induced by electrical stimuli. The tested factors included stimulus intensity (40-55 V), the presence or absence of an obstacle near the exit, the guide walls at different exit angles (15°, 45°, and 75°), and group size (15 vs. 50 mice). Evacuation time, average velocity, and headway were quantified to assess escape efficiency. To validate and extend the experimental findings, an agent-based modeling framework-previously verified in similar studies-was used to reproduce and generalize the tested conditions. Results show that the effects of obstacles and guide walls depend strongly on both group size and stimulus intensity: under high-density and moderate-stimulus conditions, obstacles and guide walls facilitated smoother flow and shorter evacuation times, whereas under low-density or extreme stimuli, they hindered movement. These results demonstrate the density-dependent nature of obstacle effects and confirm that headway serves as a robust mechanistic indicator of flow stability, revealing that sufficient inter-agent spacing is essential to prevent physical interlocking and clogging at the bottleneck. Overall, this study integrates experiment and simulation to establish a validated framework for analyzing collective evacuation dynamics in animals under controlled conditions.