Nowadays there are many car sharing schemes in operation, where vehicles can be hired for short trips. In most of these schemes the hired vehicle needs to be returned to the same location as where it was picked up. A major challenge is one-way rentals since this introduces the need to redistribute the vehicles to locations where there is high expected demand and/or lack of supply, which is very costly. This is a one-year project funded by Toyota to find a cost effective solution to this problem.
There are several research challenges. First, the aim is to create a user model from the real data currently available from existing Toyota car sharing schemes. The user model should specify the userâs preferred locations, but also alternative locations and how much they would need in terms reward to drop their vehicle at a different location. The second research challenge is to design a payment mechanism based on game theory, which (1) truthfully elicits the user preferences, and (2) optimises the revenue for the car sharing operator (in this case Toyota). The third challenge is to build a simulation environment where different payment schemes are tested. Finally, the aim is for the scheme to be tested at one of the sites.