Fair Stochastic Vehicle Routing with Partial Deliveries.

This article focuses on the fair stochastic vehicle routing problem with partial deliveries (FSVRP-PD), a variant of the vehicle routing problem that addresses uncertain customer demands and allows for partial fulfillment of these demands while ensuring fairness. The authors propose a solution framework combining route planning with a sequential resource allocation policy based on Rawlsian (max-min) fairness, which guarantees that each customer’s expected fill rate meets a predefined threshold. To solve the FSVRP-PD, an exact branch-price-and-cut (BPC) algorithm is developed, incorporating problem-specific bounding techniques and heuristic methods to efficiently generate routes that satisfy fairness and cost-efficiency criteria. Computational experiments on 75 benchmark instances demonstrate that the proposed max-min allocation policy outperforms alternative policies in balancing routing costs, service equity, and vehicle utilization. The paper concludes by suggesting future research directions including alternative fairness metrics, fleet coordination during execution, and scalability improvements for pricing algorithms.