Modeling order guidelines to improve truckload utilization

Abstract

Freight vehicle capacity, whether it be road, ocean or air transport, is highly underutilized. This under-utilization presents an opportunity for companies to reduce their vehicular traffic and reduce their carbon footprint through greater supply chain integration. This thesis describes the impact of ordering guidelines on the transport efficiency of a large firm and how those guidelines and associated practices can be changed in order to gain better efficiency. To that end, we present three recommendations on improving the guidelines based on the shipment data analysis. First, we discuss the redundancy of one of the company's fill metrics based on a scatter plot analysis and a chi-square independence test. Second, we explore the impact of using linear programming to allocate SKUs to different shipment, highlighting the reduction in the number of shipments through better truck mixing. Finally, we divide the SKUs into three groups: cube-constrained, neutral, and weight-constrained. Based on this segmentation, we present a basic model that mixes different SKUs and helps a shipment to achieve a much higher utilization rate. The application of the last two findings can be further explored to address under-utilization in freight carriers across different industries.