Agent-Based Simulations for Disaster Decision Support

Abstract Agent-based simulation is a potentially useful tool to examine and understand how observed social entities respond efficiently to interventions for defense against sudden and destructive events. Agents can be coded as objects with properties (e.g., age, gender) to mimic the behavior patterns of individuals during an emergency. In this paper, I discuss the potential application of agent-based models on decision making related to evacuation plan. Specifically, I focus on using such models to learn how at-risk populations interact with each other and respond to interventions during evacuation, so that scenarios can be built to study how interventions might optimize evacuation process.