Essentials of an Integrated Crowd Management Support System Based on Collective Artificial Intelligence

The simulation of the dynamical behavior of pedestrians and crowds in spatial structures is a consolidated research and application context that still presents challenges for researchers in different fields and disciplines. Despite currently available commercial systems for this kind of simulation are growingly employed by designers and planners for the evaluation of alternative solutions, this class of systems is generally not integrated with existing monitoring and control infrastructures, usually employed by crowd managers and field operators for security reasons. This paper introduces the essentials and the related computational framework of an Integrated Crowd Management Support System based on a Collective Artificial Intelligence approach encompassing (i) interfaces from and to monitored and controlled environments (respectively, sensors and actuators), (ii) a set of software tools supporting the analysis of pedestrians and crowd phenomena taking place in the environment to feed a (iii) faster than real-time simulation of the plausible evolution of the current situation in order to support forms of inference providing decision support to crowd managers, potentially directly controlling elements of the environment (e.g. blocking turnstiles, escalators), communicating orders to operators on the field or trying to influence the pedestrians by means of dynamic signage or audible messages.