A Simulation-Based Process Model for Managing Drone Deployment to Minimize Total Delivery Time

Managers in many countries are exploring innovative solutions for last mile delivery, including using delivery drones. In fact, drone delivery is already improving logistics service in many countries. There are, however, unaddressed nuances when comparing drones with surface vehicles. An assumption often made is that delivery time is independent of delivery processes that are not typically associated with drive time, such as loading. This can be acceptable for comparing alternate types of surface delivery vehicles, but not when comparing surface delivery vehicles with drones. To address this issue, empirical store-to-consumer deliveries are collected from a retail establishment in the United States and used to design and calibrate a discrete event simulation model to match the retailer's delivery process. Using this model, drones are compared with surface vehicles under alternate changes of state. The model is given a range of post hoc tests to further define the dimensions of the model and to generalize it for research applications in multiple contexts.