No abstract

In this work we develop methods to optimize an industriallyrelevant logistics problem using quantum computing. We consider the scenario of partially filled trucks transporting shipments between a network of hubs. By selecting alternative routes for some shipment paths, we optimize the trade-off between merging partially filled trucks using fewer trucks in total and the increase in distance associated with shipment rerouting. The goal of the optimization is thus to minimize the total distance travelled for all trucks transporting shipments. The problem instances and techniques used to model the optimization are drawn from real-world data describing an existing shipment network in Europe. We show how to construct this optimization problem as a quadratic unconstrained binary optimization (QUBO) problem. We then solve these QUBOs using classical and hybrid quantum-classical algorithms, and explore the viability of these algorithms for this logistics problem.