Impacts of Autonomous Vehicles on Parking and Congestion

Autonomous Vehicles (AVs) are expected to improve mobility, safety, and parking accessibility. To capitalize on these benefits, all major car manufacturers are testing autonomous technology, and surveys show that people want to embrace autonomous features in their next vehicles. It is estimated that AVs will be available on the market in the next decade. Hence, this thesis investigates AVs impacts on congestion and parking, and provide insights that will help transportation planners/managers derive the greatest benefit from AVs. As car-parks are closed environments, they are among the first locations to expect influenced with AVs. AVs can be stacked in multiple rows like valet parking due to their self-parking capability. To release blocked vehicles, car-park operators relocate certain vehicles in driverless mode to create a clear pathway for cars that need to leave the facility. The problem of finding the optimal car-park layout design is investigated that minimizes relocations to fit a given number of vehicles in the car-park. Results show that autonomous vehicle parking facilities can decrease the need for parking space by an average of 62%. The operation of AV parking facilities is also investigated, and policies for choosing a parking spot for each vehicle are proposed based on the arrival and departure time information. The analyses show that parking in the spot with the lowest blockage probability can decrease the number of relocation movements. AVs will not only change the car-park facilities, but also where travellers park. AV users can exit from their vehicle and send their vehicle to park in a strategic location. AVs can park farther from destination in cheaper parking lots or cruise for the whole activity time. Results show that while the same parking price across all the locations would exacerbate the congestion by motivating more AVs to cruise, a toll for zero-occupant AVs would decrease the congestion. Since AVs adoption will be gradual, there would be a mix of human-driven and AVs on streets for a while. Results show that simple policies such as AV exclusive links can improve network performance in mixed traffic of AVs and human-driven vehicles.Ph.D