Travelling Thief Problem with Drone: New Model for Logistics

A recent study presents the Travelling Thief Problem with Drone (TTP-D), a combinatorial optimization challenge that depicts a ground vehicle gathering items while an accompanying drone collects distant items to mitigate the total penalty related to travel time. The goal is to optimize the profit collected after accounting for a time-dependent rental expense by effectively coordinating item selection, vehicle routing, and drone flight timing. The researchers create a mixed-integer linear program for smaller cases and devise metaheuristics along with an attention-based Deep Reinforcement Learning (DRL) approach for larger scenarios. Additionally, they introduce a hybrid solver that utilizes the DRL policy to kickstart solutions for the metaheuristic. This research can be found on arXiv with the identifier 2608.16435.