Abstract

Nowadays, many cities are supplied with one or several fleets of shared vehicles as an alternative method for daily transportation. These fleets could be improved in many aspects if they employ autonomous or electric autonomous vehicles. Zero greenhouse gas (GHG) emission, fewer personnel requirements, automatic self-charge evaluation and process, and smooth driving are some of these improvements. This essay models a fleet comprised of shared electric autonomous vehicles (SEAV), giving one-way transportation service for short-distance travels. The case study of the area around Montreal Olympic Park is suggested to test the model's validity. In this regard, five stations are introduced on the streets surrounding the park. Different combinations of the fleet parameters are considered to generate 105 various scenarios. A multi-agent model is developed in the NetLog toolkit to simulate such a fleet under each scenario. The results show that a fleet of 20 vehicles of a 50km range with 30 parking spots can meet the demand under all considered customer arrivals frequencies with 90% or above performance rates. Another finding from the results suggests that the impact of different ranges on the fleet's performance compared to the fleet size and demand probability is almost negligible. The last part investigates the number of parking spots on fleet performance. The results suggest that the fewer the number of parking spots, the fewer rejected customers but at the same time, the higher empty travels.