When Fire Strikes, Can Buildings Be Swept Faster?

Abstract

Efficient emergency sweeps in building environments, such as following fires or toxic gas releases, are critical for ensuring occupant safety and structural clearance. Existing search and rescue planning often relies on static assumptions and struggles to account for dynamic interactions among building layout, responder coordination, and evolving environmental hazards, leading to suboptimal deployment strategies. To address this gap, the study proposes a hybrid Cellular Automata-Agent-Based Model for Search and Rescue (CA-ABM-SAR), designated as CAS. The model integrates enhanced cellular automata for environmental representation with an agent-based framework incorporating an action-priority strategy and an improved A* pathfinding algorithm. This hybrid approach systematically accounts for multiple critical factors, including building structure, occupant distribution, responder characteristics, information uncertainty, and time-varying hazards. The study validated the CAS model across diverse scenarios, from single-story offices to complex multi-story hospitals and warehouses. Monte Carlo simulations demonstrate that optimal responder deployment reduces sweep time significantly. Specifically, the study quantified that a reduction of one responder from the optimal number increases sweep time by 17.9%, while light smoke presence across all rooms increases average time by 32.2%. Furthermore, the model was extended to integrate real-time sensor data for dynamic hazard adaptation. The results provide a robust, quantitative framework for emergency planning committees to optimize resource allocation and sweep strategies.