Dipartimento di Ingegneria informatica, automatica e gestionale

Agent-based Modeling and Simulation is a powerful technique which allows to study the interactions in complex systems, and allows to explore or even foresee the emergence of more complicated properties or behaviors related to the interaction among the simpler agents in the environment. In the context of emergency or crisis scenarios, Agent-based Modeling and Simulation can allow to effectively study emergency plans, with the goal of assessing their viability, also with respect to the number of possible fatalities. In this paper, we analyze Agent-based Modeling and Simulation for crisis scenarios from a methodological and empirical point of view, with the goal of identifying what are the behavioral parameters that a model should encompass, in order for the results of the simulation to be useful for emergency plan assessment and/or compilation. We also experimentally provide a characterization of the effects of such behavioral parameters.