Mixed-integer linear optimization for full truckload pickup and delivery

This paper is focused on solving an industrially-motivated, rich routing variant of the so-called full truckload pickup and delivery problem. It addresses a setting where the distributor has to transport full truckload shipments between distribution centers and customer locations, yet the distributor’s owned fleet is inadequate to perform the totality of the required deliveries and thus a subset of the deliveries has to be outsourced to third-party carriers. In this work, we propose a novel mixed-integer linear programming formulation to model this problem. Using datasets inspired from industrial practice, we evaluate the computational tractability of this model and demonstrate its potential to serve as a decision-support system for real-life operations. Furthermore,we hypothesize that the distributor may realize cost savings when the later portion of the distribution period is utilized to pre-load cargo for delivery during the following period. To that end, we augment the original model to allow for such cargo pre-loading, and we conduct a rolling horizon-based simulation study to quantify its overall economic effect.