Agent-based simulation of shared autonomous electric vehicle fleets: application to Montreal Olympic Park

Abstract Many cities now offer shared vehicle fleets as an alternative mode of daily transportation. These fleets can be further enhanced through the use of autonomous or electric autonomous vehicles, which reduce greenhouse gas (GHG) emissions, lower staffing requirements, enable automated charging, and provide smoother operation. This paper proposes an agent-based simulation framework to model a fleet of shared electric autonomous vehicles (SEAVs) providing one-way, short-distance trips. A case study is conducted for Montreal’s Olympic Park, encompassing five strategically located stations. By varying key fleet parameters, 105 scenarios are generated and analyzed. Results indicate that a fleet of 20 vehicles with a 50 km range and 30 parking spots can meet demand across all tested customer arrival frequencies, achieving service performance rates of 90% or higher. Vehicle range is shown to have minimal impact on performance compared to fleet size and demand probability. Additionally, reducing the number of parking spots decreases the number of rejected customers but increases the number of empty trips.