Evacuation simulation with communication for anti-disaster planning

Crowd evacuation simulations are a tool for analyzing and assessing the safety of occupants in buildings. Agent-based simulations provide a platform for computing individual and collective behaviors in crowds. During evacuation, it is well known that trained leaders or evacuation guides have a key role to play in saving human lives. We propose an evacuation simulation system where agents are guided by evacuation orders from authorities or where agents can acquire information from other agents in the vicinity. We considered two different situations to test the effects of evacuation guidance acquired by agents themselves during evacuations. The simulations captured typical behaviors observed in crowd evacuations. For example, the total evacuation time was reduced when most of agents followed the guidance, although the evacuation times of individual agents were different. A specific agent involved in the movements of agents moving to another place took a longer time to evacuate. The simulation appeared to capture real situations, which shows that simulations can be used to estimate evacuation situations prior to emergency evacuation drills.