Evaluating the impact of automated vehicles on residential location distribution using activity-based accessibility: A case study of Japanese regional areas

Automated Vehicles (AVs) are expected to disrupt the transport sector in the future. Extensive research efforts have been dedicated to studying its potential implications. However, the existing literature is yet limited regarding long-term impacts. To fill this gap, this paper estimates and validates a residential location choice model to evaluate the impacts of AVs on residential location distribution in the context of Japanese regional areas. Activity-based accessibility, a utility-based accessibility measure, is used to reflect the AV impacts on short-term travel demand. The year 2040 is set as the backdrop for the analyses, where the effects of the decreased population are reflected in the scenario settings, along with variables such as decreased travel impedance to accommodate the uncertainties in the characteristics of AVs. The simulation results of four AV scenarios confirm the potential of urban expansion. The results demonstrate that, compared to the Base Scenario, the median distances between the residences and the closest Dwelling Attraction Areas, representing as the target for attracting residences by the local government, expand by 5.2% to 10.2% for the AV scenarios. Two hypothetical policy mandates are then applied to alleviate the problem. The results indicate that providing a 10% subsidy to the land price might be effective to mitigate urban sprawl for the scenario with relatively conservative AV settings. Nonetheless, the substantial cost associated with implementing such a policy renders it effectively unfeasible and underscores the need for proactive planning in anticipation of a level 4 or 5 Automated Vehicle future.