Authors: Alberto Tremori, Enrico Bocca, Federico Tarone, Francesco Longo, Simonluca Poggi
Over the last decade the authors have been working on several researches concerning the use of intelligent agents for managing simulation of critical situation in security, such as managing civil disorders, riots or terrorist actions. The adoption of intelligent agents to reproduce human behavior is a key factor to implement realistic models for security and civil protection. Human behavior is very complex phenomena to be simulated and it is important to anticipate final results to get consistent models. In this paper will be described the use of early testing procedures to anticipate results before the VV&A phase. It is, in fact, useful to adopt this kind of early testing activities in application of intelligent agents with human behavioral models.