Edge Artificial Intelligence

23Optimizing Evacuations Using Agent-Based Modeling

Donie Jardeleza

School of Computing and Innovative Technologies, British University Vietnam, Hanoi, Vietnam

Abstract

The increased frequency and intensity of various natural and artificial disasters have brought into focus the need for effective management strategies for evacuating, which could help maintain the safety of the community. While there is a policy on disaster management, and exercises are conducted regularly within many organizations, including educational institutions within Vietnam, how this works is not known given various dynamic variables. There is a big lacuna that exists in understanding and improving evacuation strategies with a real-time data-driven approach. Contemporary methods hardly consider the richness of crowd dynamics and variability of hazardous scenarios; therefore, effectiveness in response measures and preparation activities is reduced. This research paper illustrates the development of a simulation-based framework by using Agent-Based Modeling (ABM) to assess and optimize an evacuation strategy in educational settings. It will consider the impact of several evacuation scenarios on procedural efficiency by modeling real locations and individuals as agents in this simulation framework. The optimum evacuation strategy that reduces the evacuation time, improves safety measures, and distributes resources properly will be the main target. These will also give insight into the phenomenon of bottlenecks ...