Evaluating the impact of private automated vehicles on activity-based accessibility in Japanese regional areas: A case study of Gunma Prefecture

Abstract

Automated Vehicles (AVs) are expected to disrupt the transport sector in the mid- to long-term future, and extensive research efforts have been dedicated to studying its potential implications. However, the existing literature is still limited in the context of Japan. To fill this gap, an integrated travel forecasting approach that combines an activity-based travel demand model and dynamic traffic assignment is estimated and validated to evaluate the impacts of AV-induced travel, and the feedback effects between transport supply and demand. The target area of this study is Gunma prefecture, a regional area in Japan. The planning horizon is set to 2040, where the effects of population decrease are reflected in the scenario settings. Policy scenario analysis that assumes complete substitution of human-driven vehicles with AVs suggests increases between 22.1% and 43.6% in total distances traveled, with a corresponding increase in increased CO2 emissions between 23.6% and 46.0%. While this suggests a worsening network level of service, around 2 to 3 min gains in median activity-based accessibility were also observed, suggesting overall positive effects from AV adoption in the study region.

Keywords

• Automated vehicles
• Autonomous vehicles
• Travel demand modeling
• Travel forecasting
• Large-scale simulation
• Japan