On Modeling Groups in Crowds: Empirical Evidence and Simulation Results Including Large Groups

Research on pedestrian movement strives to mitigate risks at large events or public infrastructures by better understanding the flow of a crowd. Thus, for evacuation planning it is essential to understand what constitutes a crowd. This includes the crowd’s composition of various groups of people and its influence on the evacuation process. This paper is a joint effort by social scientists and mathematical modelers cooperating in the research project REPKA to shed more light on this aspect. REPKA (Regional Evacuation: Planning, (K)Control, and Adaptation) focuses on open space evacuation of big events, especially the regional evacuation of national soccer matches. Here, larger groups of fans with a spirit of togetherness are eminently present. Social scientists and mathematical modelers work on this task from different perspectives relying on the tools of their trade: Empirical surveys – interviews and observations – have been conducted by social scientists to gather information on the occurrence and relevance of large groups. They are the basic input for mathematical modelers together with first suggestions on consistent distributions for the group composition. The mathematical modelers integrate these results into a pedestrian stream model that includes larger groups composed of subgroups. They demonstrate how the occurrence of larger crowds affects the flow of a crowd at a road crossing.