POPULATION-LEVEL IMPACT ANALYSIS OF AUTOMATED VEHICLE SERVICES FOCUSING ON THE CONCENTRATION AND DISPERSION OF URBAN STRUCTURE USING AGENT-BASED SIMULATION

Abstract

 In this study, population-level travel demand data were generated assuming different urban structures with concentrated or dispersed population and facility distribution in addition to the current distribution in the target area. Comparative evaluation of the convenience and transportation efficiency of shared automated vehicle (SAV) services in different urban structures and impact analysis of regional implementation of SAV were conducted. Agent-based simulations were conducted based on a mode choice model estimated from a stated preference survey that asked about the intention to use SAVs. For this analysis, a regional city in the northern part of Gunma prefecture was set as the target area. Results revealed that as the distribution was concentrated in the centeral city, the convenience and transportation efficiency increased, and the positive influence on the operation of automated driving services became stronger. On the other hand, regarding the regional disparity in convenience, the results showed that the waiting time was considerably longer in the low population areas far from the central city area.