Two-Echelon Vehicle Routing Problem with Time Windows and Intermediate Facilities for E-Commerce Logistics in Crowdsourcing Model

This paper investigates a novel variant of the two-echelon vehicle routing problem (2E-VRP) within the context of modern e-commerce logistics. The model integrates time windows, occasional trucks, occasional drivers, heterogeneous regular vehicles, and two increasingly relevant types of intermediate facilities: transshipment nodes and parcel lockers. Goods are transported from a first-echelon depot to intermediate facilities via occasional trucks and heterogeneous regular trucks, then delivered to customers through occasional drivers and heterogeneous regular riders at the second echelon, or retrieved by customers from designated parcel lockers. Given significant disparities in service scope, vehicle attributes, delivery rules, and cost structures between the two echelons, the crowdsourcing model for the first echelon is redesigned, and new intermediate facilities are incorporated to expand application scenarios and further enhance the operational advantages of the crowdsourcing mode. The objective aims to minimize transportation costs and minimize customer dissatisfaction costs. This study constructed a two-echelon vehicle routing optimization model under the crowdsourcing mode with time windows and intermediate facilities and developed an improved multi-objective sparrow algorithm (IMOSSA) to solve it. Algorithm comparisons and extensive case studies validate the algorithm’s effectiveness and superiority. Scenario analysis investigates the impact of crowdsourced vehicles on routing decisions, experimentally demonstrates the crowdsourcing model’s advantages, and derives practical managerial insights.