TR-10-13: The Microscopic Human Factors methodology for modelling cognition in crowds and swarm systems

Abstract

Like in other species there has been an interest in modelling behaviours of human aggregates: crowds. Despite the important influence of cognition on individual human behaviour, it has been a challenge to integrate human factors into swarm-principled crowd models (termed: microscopic crowd models). To improve this situation we offer the Microscopic Human Factors methodology, which (a) guides modelling of cognitive factors, and (b) increases the transparency of resulting models to non-modelling crowd researchers. Through (a) resulting models become models of people, with better relevance to real crowds. Through (b) resulting models can better contribute to the larger field of inquiry into crowd behaviour. The methodology proposed can also assist in modelling collective behaviour in other species, particularly those species with significant cognitive components in their social behaviour. The methodology is general, calling for three complete descriptions of the collective behaviour regardless of the implementation formalism to be used: The specification delineates the crowd behaviour of interest. The reduction identifies the abstract, local behaviour of individuals that underlies the specified collective effect. Formal modelling and implementation details are isolated to the implementation. By considering the methodology’s layers in reverse order it is possible to make arguments concerning the validity of resulting models.

TR-10-13.pdf