A computation-implementation parallelization approach to the vehicle loading and routing problem.

In this article, we address a version of the capacitated vehicle routing problem where there is a constraint on the total time that can be spent on computing delivery routes and loading the vehicles. This problem, which we call the vehicle loading and routing problem (VLRP), can arise, for example, in the delivery of small, quick-turnaround orders from a warehouse. We propose a computation-implementation parallelization (CIP) approach to solving large VLRP instances, and present computational results showing that CIP improves upon the conventional compute-first-implement-later method. Our approach allows us to find lower-cost solutions in the same amount of time, or alternatively to find equally good solutions that allow improvements in order cutoff time or truck dispatching time to increase customer satisfaction.