A dynamic approach for evacuees? distribution and optimal routing in hazardous environments

Abstract

In a complex built environment, the situation changes rapidly during an emergency event. Typically, available systems rely heavily on a static scenario in the calculation of safest routes for evacuation. In addition, egress route calculation and evacuation simulations are performed separately from path-finding for rescue teams. In this paper, we propose a state-of-the-art dynamic approach, which deals not only with a 3D environment, shape of spaces and hazard locations, but also with the dynamic distribution of occupants during evacuation. A database of densities and information about hazard influence are generated and used to calculate optimal paths for rescue teams. Three simulation scenarios were rigorously compared in this study, namely static with constant density values determined for subsequent stages of evacuation, semi-dynamic with densities representing an actual people distribution in a building during evacuation simulation, and dynamic with temporal distribution of evacuees stored in a database, and dynamically used in optimal path calculations. The findings revealed that static simulation is significantly different from semi-dynamic and dynamic simulations, and each type of simulation is better suited for the decision task at hand. These results have significant implications on achieving a rapid and safe evacuation of people during an emergency event.