Intelligent Building Evacuation: From Modeling Systems to Behaviors

Disaster risk management requires new approaches and mechanisms to improve citizens’ safety in disasters. The Internet of Things (IoT) is among the technologies that could enhance awareness by providing real-time information. When an emergency happens, building occupants need to be evacuated to safe areas in the shortest possible time. Optimization algorithms could receive humans’ mobility data from IoT resources and calculate the best route to follow. The algorithm we present in this chapter formulates and solves a linearized, time-indexed flow problem on a network that represents feasible movements of people at a suitable frequency. We evaluate the performance of the IoT system, including the algorithm, to confirm compliance with real-time use. While the optimization method gives a best case scenario, it does not reflect actual human behavior in evacuation. Humans may stay calm and follow our IoT system’s instructions, but they may also have different characteristics and contexts or experience panic attacks, or emotional and social attachment. Thus, we recreate our scenarios with agent-based social simulations, which model occupants as computational agents in an artificial society. The simulations give insights towards a more efficient IoT infrastructure design. We apply our approach to a real location with actual data to prove its feasibility.