Market Design Approaches for Fractional Ownership of Autonomous Vehicles

This research aims to design two new markets for fractional ownership of autonomous vehicles (AVs): matching and auction. The introduction of AVs to consumer markets will reshape how we currently own and operate vehicles. Since AVs would be allowed to travel without passengers, a big change is expected: it will enable sharing vehicles and expedite co-owning vehicles. Currently, co-owning a car with friends is not easy, although it can certainly reduce the cost of owning a car. With AVs, the co-owned car can travel autonomously from my location to my friend’s location, which has not been possible with conventional vehicles. Therefore, we envision that AVs will be co-owned widely and new markets will be created accordingly. AVs are impactful when carrying passengers; they are even more impactful when traveling without passengers. In the first part of this talk, we use stable matching to help customers find an appropriate group to share an AV and present a generalized stable matching model that allows flexible sizes of groups as well as various alternative objectives. We also present a computational method for the resulting combinatorial optimization problem. In the second part, we propose a novel auction market design, namely a combinatorial auction with bidder defined items, in which customers submit bids for the time periods they want to use an AV. Considering spatial information of bidders, we formulate the winner determination problem for factional ownership under both discrete- and continuous-time settings. We propose a clique-based method for solving the optimization problem.