A multi-period analysis of the integrated item-sharing and crowdshipping problem

The integrated item-sharing and crowdshipping problem arises from the operational planning of sharing platforms that coordinate diverse services offered by individuals to peers. Item-sharing means to rent-out items whereas crowdshipping describes the willingness of private drivers to conduct deliveries on their planned trips. Integrating both concepts on a single platform can lead to higher profits and better service quality by transferring items through crowdshippers. We investigate a multi-period variant of the problem to facilitate more foresighted assignments of items to requests and crowdshippers. Thereby, items can be used to sequentially serve multiple requests without being returned to their initial locations. This so-called ’request chaining’ cuts transportation effort and improves the availability of items to customers. From a modeling perspective, the planning problem is an example for the rarely investigated many-to-many pickup and delivery problems. We present a binary program and we address problem inherent dynamics due to new incoming announcements by both deterministic and stochastic problem solving approaches that are based on a rolling horizon framework. Furthermore, a simulation environment is set up to investigate the effect of a responsive requesting behavior due to word-of-mouth. Our experiments show that some look-ahead in the planning is particularly useful for scenarios with few requests. The consideration of stochastic information can improve operational performance and word-of-mouth can best be dealt with when consumers interact locally. We also derive managerial insights on the number of items that should be provided in such a community, subject to consumer preferences.