Optimization of the drone-assisted pickup and delivery problem

The proposition of utilizing unmanned aerial vehicles (UAVs), colloquially known as drones, for the purpose of delivery was ushered into the mainstream by Amazon in 2013. Since then, UAVs have quickly gained traction as a feasible option for last-mile operations in transport and logistics. In the context of this research, we construct a mathematical framework aiming to depict the collaborative interaction between trucks and drones in coordinating pickup and delivery tasks with the objective of minimizing operational costs. The drone-assisted pickup and delivery problem (DAPDP) is a variant of the problem in which a single truck departs a depot with parcels and a UAV on board. As the truck picks up packages and makes deliveries, the UAV can also be used to make deliveries to customers near the truck's position. As the unmanned aerial vehicle (UAV) embarks on its delivery task, the truck continues on its route, making further deliveries along the way and retrieving the UAV at another customer location different from the launch point. The model is presented as a mixed integer linear program (MILP), and a novel heuristic solution approach based on the classic Clarke–Wright savings heuristic is proposed. Our heuristic's efficacy is evaluated in comparison to a scenario involving only trucks, with a comprehensive series of numerical experiments conducted to underscore the advantages of incorporating UAVs into the pickup and delivery problem. Finally, we perform a comprehensive sensitivity analysis of key drone parameters in order to demonstrate their impact.