Crowdsourcing Incentives for Multi-hop Urban Parcel Delivery Network

Efficient and economic parcel delivery becomes a key factor to the success of online shopping. Addressing this challenge, this paper proposes to crowdsource parcel delivery tasks to urban vehicles to utilize their spare capacities, thus improving the efficiency while reducing traffic congestions. The delivery is planned as a multi-hop process, and participating vehicles will carry parcels from one shipping point to the next until they arrive at the destination, following the routes learned from historical traffic statistics. The major contributions include an incentive framework to motivate the vehicles to participate in the delivery tasks by preserving the interests of the platform, the sender, and the crowd vehicles. Two incentive models are designed from platform-centric and user-centric perspective, respectively. The platform-centric model first assesses an optimal reward R for parcel delivery with the principle of Stackelberg game, which enables the platform to maximize its profit. The user-centric model then applies a reverse auction mechanism to select the winning bids of vehicles while minimizing the sender cost, with truthfulness guarantee. Theoretical analysis and extensive experiments on a real urban vehicle trace dataset are provided to validate the efficacy of the proposed framework.