Time Estimation and Hotspot Detection in the Evacuation of a Complex of Buildings: A Mesoscopic Approach and Case Study

Evacuation drills are critical to ensure safety and to evaluate emergency preparedness. However, drills are overly expensive in terms of the involved resources. Before conducting drills, it is possible to test the infrastructure under stress using computer simulation (combined with optimization) to identify critical points (hotspots) that could improve the flow dynamics during a real evacuation or emergency. This article proposes a mesoscopic (optimization-simulation) framework that models the evacuation dynamics from a strategic perspective considering the infrastructure of the evacuated area and the effect of congestion on the evacuation times. The mesoscopic framework iteratively interweaves two models: (i) a macroscopic network-oriented optimization model that finds flows of evacuees toward the safe zones, and (ii) a microscopic discrete-event simulation model that evaluates the effect of congestion. To illustrate the applicability of the proposed framework, the article presents a case study that recreates a real evacuation drill that took place in a university campus. The results of the case study support infrastructure investment decisions that enhance the evacuation dynamics.