A fire navigation model: Considering travel time, impact of fire, and congestion severity

Abstract

The navigation model presented integrates Fire Dynamics Simulation (FDS), a navigation graph generation model, a modified Dijkstra algorithm, Agent-Based Simulation (ABS), and Intelligent Active Dynamic Signage System (IADSS). The FDS evaluates fire impacts on paths’ safety, while ABS captures evacuees’ interactions and congestion. The modified Dijkstra algorithm identifies optimal paths, considering travel time, fire impacts, and crowd density. The IADSS dynamically communicates these paths to evacuees. The contributions include (1) integrating the combined effects of heat and toxic gases on evacuees with congestion and travel time into an evacuation framework, (2) introducing algorithms for integrating signage systems into buildings’ navigation graph, (3) determining recommended and negated signage directions, (4) proposing a central server that updates the signage directions concerning real-time buildings’ conditions and congestion, and (5) developing a tool for minimizing evacuation time, congestion, and fire impacts. Case studies across diverse fire scenarios and building types showcase the framework's adaptability. Results indicate significant improvements over traditional methods, including reduced evacuation time, congestion severity, and cumulative fire impacts. Furthermore, the model computational efficiency enables time-sensitive fire evacuation planning. A validation study comparing it with two established methodologies highlights its superior signage direction determination, heightened evacuation performance, and enhanced output richness.