Gaussian Mixture Model to Characterize Payload Distributions for Predominant Truck Configurations and Body Types

Abstract

This article develops and applies a Gaussian mixture modeling approach to characterize payload distributions and calculate mean payloads for the 14 predominant axle configuration–body type pairs operating in Manitoba, Canada. The methodology uses readily available truck mass data to distinguish empty and partially loaded trucks. The results demonstrate that, regardless of axle configuration, the load distributions for trucks with van, flat deck, and container body types comprise trucks with gross vehicle masses throughout the available mass range, reflecting the diversity of commodities being hauled. In contrast, trucks with dump, hopper, and tanker body types which haul heavy bulk commodities, tend to travel either fully loaded or empty. Comparisons with findings reported in the literature reveal that mean tare masses may be insensitive to jurisdiction-specific industrial and regulatory conditions. However, the availability of a variety of axle configurations enables carriers to select trucks well-suited to their freight transport task. This finding emphasizes the need for jurisdiction-specific characterization of payload distributions and means, and supports better understanding of the relative impacts of industries on pavement and highway network performance.