Drone-aided mobile blood collection problem: A rolling-horizon-based matheuristic

Abstract

This study introduces the drone-aided mobile blood collection problem, which integrates mobile blood donation vehicles with drones to improve operations related to the blood collection in urban areas. Each vehicle, carrying multiple drones, travels to several collection sites to conduct blood collection operations within a working day. Drones fly between vehicles to pick up collected blood bags and deliver them to the blood center. This collaborative framework enhances the performances of the collection system and ensures the freshness of collected blood upon arrival to the blood center. We develop a novel mixed-integer linear programming model to optimally synchronize the routes and collection schedules of mobile units and drones to ensure the timely delivery of collected blood to the blood center. We also develop a rolling-horizon-based matheuristic to solve large-scale instances of the problem. This algorithm combines a rolling horizon approach, which divides the problem into manageable subproblems solved sequentially, with a local branching technique that enhances solutions by exploring promising neighborhoods. To evaluate the algorithm’s performance, we conduct a comprehensive computational study. Our results show that the proposed algorithm not only finds better solutions than those obtained by Gurobi but also outperforms other matheuristics, including the rolling horizon, relax-and-fix, and fix-and-optimize algorithms. Finally, we demonstrate the real-life applicability of the problem through a case study in Quebec City, Canada.