Joint Optimization Of Incentive And Routing Strategies In Crowdsourced Last-Mile Delivery

Abstract

<p>The growing demand for efficient last-mile delivery has given rise to a new urban logistics paradigm – crowdshipping – which involves engaging the public in parcel transportation by offering incentives to the "crowd" during their daily trips. There is uncertainty regarding whether participants will accept an assigned delivery order, along with its accompanying reward from the platform. This uncertainty must be considered when the platform formulates real-time decisions pertaining to order allocation and route optimization. This study addresses the design of incentive strategies for crowd carriers considering this uncertainty and explores collaborative routing decisions with professional couriers to minimize overall platform costs, including both incentive and routing expenses. This problem is formulated as a Mixed Integer Nonlinear Programming (MINLP) model that determines the optimal incentive costs for each crowd carrier, which in turn influences their probability of participation. Additionally, the model optimizes order allocation and routing strategies involving both crowd carriers and professional couriers. The detour distances for crowd carriers are considered, which further impacts their actual participation probabilities. As the model is NP-hard, we develop an Adaptive Large Neighborhood Search (ALNS)-based algorithm that significantly improves solving efficiency. Through numerical experiments, we identified the optimal incentive and routing strategies under different relationships between incentive costs and initial participation probabilities of crowd carriers. Our findings emphasize the varied roles of different crowd carriers in the delivery ecosystem and highlight the importance of tailored incentive strategies to enhance overall delivery efficiency.</p>