Privacy-Preserving Time-Sharing Services for Autonomous Vehicles

Sharing the Autonomous Vehicles (AV) has the potential to be the ultimate solution for overcoming the cost problem of these type of vehicles to fundamentally change the transportation systems. AVs enable the time sharing services where AV owners share their AVs at the times they don't need them. Such sharing could reduce the cost by enabling the owner to share the cost of the vehicle with other users. However, these services raise a severe privacy concern as the shared location and route data of the users are considered highly private and sensitive. In this paper we propose a privacy- preserving time-sharing scheme for AVs. Our approach enables the owner and the requester to perform a privacy-preserving matching on their transportation needs over the server without disclosing their routes to the server. To do so we use a set of Points of Interest (POI) locations as intermediate destinations in travel paths. Only if the matching is conflict-free and efficient, the owner and the requester share the details of the routes. We also show the accuracy of the proposed approach through extensive simulations on real data. It is shown that our enhanced POI selection scheme, with consideration of the traffic information and patterns in the area, outperforms the baseline scheme where the POIs are chosen randomly. Furthermore, it shows that our scheme achieves high accuracy in terms of resulting in false negatives compared to the ground truth.