Air–Ground Cooperative Path Planning for Doorstep Pickup Considering Multiple Uncertainty

Air-ground heterogeneous robots collaboration can efficiently handle first-mile pickups and reduce reverse logistics costs. When an unmanned ground vehicle (UGV) and an unmanned aerial vehicle (UAV) collaborate for doorstep pickups, the UAV handles customer pickups while the UGV acts as a mobile warehouse, providing replenishment and transportation services. This setup enables long-distance, large-scale, and flexible logistics. However, real-world collaborative pickups face uncertainties in pickup times and parcel weights, complicating path planning. To address the challenge, this paper aims to minimize the total travel distance for UAV and UGV and formulates a path planning model with time windows under uncertainty and designs an adaptive genetic large neighborhood search algorithm for solution. Through numerical simulations and comparative analysis, the method’s scalability and adaptivity to uncertainty has been validated extensively. Real-world experiments confirm the method’s effectiveness and feasibility. These findings offer both theoretical guidance and practical insights for air-ground robot path planning, contributing to improved logistics efficiency and cost reduction.