Applying Vision-Driven Agent to Simulate Pedestrian Trajectory Under Furniture Layouts in Public Squares

Abstract

In recent years, in Japan, public squares have been the focus of downtown revitalization, with the municipal sectors introducing tables, chairs, plantings, and stands that provide refreshments. In considering the efficient operation of public squares, it is important to analyze the influence of these layouts (hereinafter, layout) on pedestrian movement. In this chapter, we investigate pedestrian movement in public squares where furniture layout is arranged, identifying vital characteristics in the process. We also attempt to simulate pedestrian-movement networks that differ according to layout by agent simulation, specifically Exosomatic Visual Architecture (EVA), which is used to represent the Natural Movement. In this chapter, a pedestrian-trajectory survey is conducted in Toyota City Station Square. The results suggest that singularity examples deviating from the shortest-path principle tend to opt for the Longest “Line-Of-Sight” (LLOS) direction. Next, in addition to the LLOS principle, we explain the pedestrian-agent model, which has an OD (Origin-Destination)-tuple in the public square as well as an exit rule. The Toyama Grand Plaza, which is a rectangular square, is examined by matching the result of agent simulation by setting the condition of layout of furniture with the result of the trajectory survey. The simulation results for different layout settings are also examined. Moreover, we also examine how to best handle complex spaces in a general sense.