Optimal route selection model for fire evacuations based on hazard prediction data

Abstract

Providing accurate information on available evacuation routes is critical during the time-sensitive emergency situation of a building fire, particularly when it occurs in a large-scale facility with a complex layout. Timely access to safe and efficient egress paths helps minimize exposure to hazardous fire effluents such as toxic smoke during evacuation. The following study develops a computational model which uses hazard prediction data to identify optimal evacuation routes, the safest and shortest paths to the nearest exit, during the event of a building fire. It uses the Fire Dynamics Simulator to provide prediction data on smoke propagation inside a structure and the A* algorithm to search for the fastest escape path. The algorithm is modified to consider whether the ensuing nodes in the route are in a normal or hazardous state. The test simulations demonstrate that the model is both accurate and effective in guiding evacuees to a place of safety while minimizing direct exposure to smoke. These results enable a more informed approach to safety management during indoor fires and reduce the likelihood of evacuees impeding the evacuation process by entering a dangerous area unprepared.