Colloquium: Raven van Ewijk (C&O)

Using drones with delivery trucks can greatly improve last-mile delivery efficiency. In this approach, drones are dispatched to customers directly from the truck, allowing the truck to continue its route without interruption. This approach gives rise to the widely researched Traveling Salesman Problem with multiple Drones (TSP-mD), a widely studied optimization challenge. Many heuristic models have been developed to address TSP-mD, though these are often static and are vulnerable to disruptions-common in real-world delivery networks. The study introduces a dynamic method for TSP-mD, aiming to improve adaptability to real-time disruptions like truck delays and drone speed variations. Using the simulation tool Bluesky, this new model is compared to a traditional static heuristic algorithm under uncertain conditions. While the existing model achieves shorter completion times in stable conditions, the dynamic algorithm shows significant performance gains under uncertain conditions, suggesting a need to reconsider global optimization strategies for TSP-mD.

Supervisor: Joost Ellerbroek