Weight-categorized truck flow estimation: A data-fusion approach and a Florida case study

Knowledge of spatial distributions of weight-categorized truck flows in a region is critical to the understanding of movements of empty or partially-loaded trucks and devising appropriate strategies to reduce empty or partially-loaded truck flows and improve truck utilization efficiency in the region. However, such disaggregated information cannot be directly obtained from existing data sources and models. In this paper, we propose a compact model for estimating weight-categorized truck origin–destination (OD) flows and link-level truck counts by fusing several freight datasets. The proposed model minimizes the squared errors between the estimated and observed truck OD flows and link volumes considering the flow conservation of trucks and commodity weights. To illustrate a real-world application of this model, a case study is conducted to estimate the spatial distribution of empty or partially-loaded truck flows into, within, and out of the State of Florida. With the case study results, high production- and attraction-zones of empty or partially-loaded truck trips are also identified. Such results can potentially inform freight planning and policy decisions to learn spatial patterns of empty or partially-loaded backhauling truck flows and devise countermeasures to reduce such flows and improve freight transportation efficiency. This is particularly relevant to a terminal state such as Florida that has large volumes of back-hauling truck flow.