Computer-Simulated Pedestrian Behavior in Shopping Environment

Abstract

There are as many distinct shopping itineraries in a shopping center as there are shoppers, at some level of discrimination; yet clear patterns of flow and volume can be observed in aggregate movement over the network of pathways. This paper tests the hypothesis that the aggregate pattern is a result of visitors operating according to simple movement heuristics. Itineraries are generated at seven entrances to a shopping mall according to four programmed heuristics. The results are aggregated by network segments and compared with observed itineraries. Generating movement toward stores according to their attraction, where attraction is defined by the size of the store, produces the best fit with actual behavior. Movement toward pathways with high connectivity also describes a significant proportion of the observed spatial behavior. Highly significant preferences are observed at the first and second choice points after the entrance. In general, microscale pedestrian choice in a shopping mall is systematic behavior and is visible in the aggregate itinerary data.