We are developing Agent-Based In-Store Simulator (ABISS). The original objective of ABISS is to investigate
customer walking flows and purchasing behaviors in a supermarket., however, the simulator is able to be used for any
pedestrian behaviors of people in a museum and/or a city zone. This paper presents the basic principles, the application to museum displays, and future plans of ABISS. The simulator enables us to carry out “virtual experiments” through
changing various parameters of marketing, display, and operation plans of the target place. Intensive experiments with these real and virtual data have suggested the applicability of ABISS to analyze pedestrian behaviors, which are hard to be observed in real data.