Humans do not always act selfishly: social identity and helping in emergency evacuation simulation

To monitor and predict the behaviour of a crowd, it is imperative that the technology used is based on an accurate understanding of crowd psychology. However, most simulations of evacuation scenarios rely on outdated assumptions about the way people behave or only consider the locomotion of pedestrian movement. We present a social model for pedestrian simulation based on self-categorisation processes during an emergency evacuation. We demonstrate the impact of this new model on the behaviour of pedestrians and on evacuation times. In addition to the Optimal Steps Model for locomotion, we add a realistic social model of collective behaviour.