Existing truck-drone delivery research often neglects dynamic traffic and inter-route flexibility. To address this, we investigate the Time-Dependent Vehicle Routing Problem with Drones and Inter-route Recovery (TD-VRP-D-IR). We formulate a Mixed-Integer Linear Programming (MILP) model incorporating FIFO-based piecewise speed functions and spatiotemporal synchronization to enable flexible cross-truck operations. An Improved Adaptive Large Neighborhood Search (I-ALNS) algorithm is proposed, featuring cascade removal and cooperative induction operators, along with a “Cluster-first, Route-second” strategy for scalability. Experiments using real-world data demonstrate that I-ALNS matches optimal solutions for small instances and achieves 95% cost optimization for large-scale instances within 600 seconds. The results validate that the proposed inter-route mechanism effectively circumvents congestion by spatiotemporally decoupling heterogeneous fleets, thereby significantly reducing delivery costs while ensuring robustness against traffic uncertainty.
https://doi.org/10.1145/3808707.3808863Cite as:
@inproceedings{Fu_2026,
series={DMIT 2026},
title={Truck-Drone Routing with Inter-route Recovery in Time-Dependent Networks: An Improved ALNS Approach},
url={http://dx.doi.org/10.1145/3808707.3808863},
DOI={10.1145/3808707.3808863},
booktitle={Proceedings of the 2nd International Conference on Digital Management and Information Technology},
publisher={ACM},
author={Fu, Jialan},
year={2026},
month=Feb,
pages={1025–1031},
collection={DMIT 2026}
}