Towards a Comprehensive Assessment of On-Demand Autonomous Vehicle Services in Rural Areas in Île-de-France Using Multi-Agent Simulations

Agent-based modeling and simulations are increasingly used to study mobility systems, assess the potential impacts of new transport services and policies, and support their design. Among these innovations, on-demand systems relying on autonomous vehicles have gained attention and are expected to offer greater flexibility for users while increasing operational efficiency. However, their viability heavily depend on design choices, such as fleet size and composition, operational strategies and constraints, service integration with existing offers, and pricing schemes. This is particularly critical in rural areas, where low demand density adds to the challenge of designing effective mobility systems. While existing studies typically focus on one of these aspects, using agent-based transport models and tools, two or more of them are rarely considered jointly.This paper proposes a comprehensive analysis of the potential of autonomous on-demand mobility systems in a rural context, jointly considering the fleet size and composition, operating hours, pricing, and level of integration with public transport. We showcase the methodology on a rural area of the Île-de-France region using a region-wide synthetic population. Various service configurations are evaluated through cost-benefit analysis, providing insights into the most effective and sustainable deployment strategies for autonomous on-demand services in rural areas.