Acquisition and Utilization of Real-Time District-Wide Truck Traffic Volume Data from Single Loop Detectors

We deployed a recently proposed single loop detector based, real-time freeway truck traffic volume estimation algorithm to a County-level traffic database. The scale-up of the algorithm is achieved by adding an extra layer, consisting of a spatial filtering component for removing noise and an aggregation component for efficient analysis. All calculations could be done in real-time to produce a real-time, district-wide snapshot of the spatial distribution of truck traffic volume. Quantitative comparison with an independent truck volume data shows 31% discrepancy, which is decent considering the limited nature of the latter [???] data source. Qualitatively, the map of the estimated truck traffic volume enables us to instantly identify known freight truck rich routes and capture interesting temporal patterns in truck traffic. We also show that the output of the algorithm can be used as an input for a simplified emission factor model. In another application, the study of the relationship between the estimated truck traffic volume and freeway crash rate shows that the former significantly affects crash rate. As illustrated by these applications, the proposed algorithm and visualization scheme could provide valuable tool and data source for transportation planners and practitioners in diverse fields.