Development of a Participatory Policy Planning Tool Based on Multi-Agent Social Simulation
Naoki Sugie, Mamoru Yoshizoe, Hiromitsu Hattori · 2024
Cities are intricate systems where diverse social structures interact, leading to the continuous emergence of complex social phenomena. Analyzing and optimizing these interactions, as well as efficiently managing urban systems, present significant challenges. Multi-Agent Social Simulation (MASS) offers promising solutions for these challenges. However, MASS is not a user-friendly technology for stakeholders involved in urban issues, such as residents and local government officials. This paper presents the development of a participatory policy planning tool based on MASS. We developed a prototype system to explore potential traffic policies, integrating traffic simulation, data visualization, and a web-based interface. A social experiment conducted with residents in Japan demonstrated the potential of the MASS-based planning tool. Additionally, in the paper, we discuss an approach to simulate and evaluate urban traffic policies on the developed MASS-based system, focusing on optimizing the combination of traffic signal timing adjustments and policy implementations.