In on-demand ridesharing, optimizing the quality of supply-demand matches and minimizing the number of unfulfilled trip requests is an important business problem. One approach is to match supply with demand based on greedy local heuristics, which is sub-optimal at the market level. Another approach is to find the globally optimal matching over the whole marketplace. However, because the computation has high latency, and usually requires aggregate supply and demand data over time, instead of only the supply and demand at the time of individual trip requests, it is unfeasible to perform the global optimization for every trip request in realtime. This paper proposes a solution that performs a global optimization offline periodically based on forecasted supply and demand data, and uses the offline results to guide realtime supply and demand matching. The solution is a novel two-stage robust optimization scheme for supply-demand matching in on-demand ridesharing. We implemented the proposed solution and evaluated it using simulated trips based on public industry data. It reduces the number of unfulfilled trips significantly compared to a state-of-the-art approach.

Mobile sensor networks are a great source of data. By collecting data with mobile sensor nodes from individuals in a user community, e.g. using their smartphones, we can learn global information such as traffic congestion patterns in the city, location of key community facilities, and locations of gathering places. Can we publish and run queries on mobile sensor network databases without disclosing information about individual nodes?