Experiments and Theory of Self-organization in Pedestrian Flow

Self-organization phenomena in pedestrian flow have been a research topic receiving quite some attention. Different microscopic simulation models have been applied to predict different forms of self-organization in pedestrian flows, such as the dynamic formation of lanes in bi-directional pedestrian flows. In this paper, we put forward a theory of self-organization, as well as results from experimental research that provides more insight into these spatial-temporal phenomena. From our experiments, we observe that groups of pedestrians that are homogeneous in terms of desired walking speeds and direction form structures consisting of overlapping layers. This basic pattern forms the basis of other more complex patterns emerging in multi-directional pedestrian flow: in a bi-directional pedestrian flow, dynamic lanes are formed which can be described by the layer structure. Diagonal patterns can be identified in crossing pedestrian flows. This paper both describes these structures, as well as the implications for theory and modeling of pedestrian flows.