Abstract
Despite their impacts on traffic performance, infrastructure, environment and safety, the travel behavior of commercial vehicles is not well understood. This is largely due to limited data available to distinguish between various types of commercial vehicles and their travel patterns. This study presents a new vehicle inductance sensor technology that obtains both the undercarriage profile and axle configuration of each vehicle, the fusion of which provides the potential for distinguishing between commercial vehicles at a more detailed level. The commercial vehicle classification model developed in this study yielded encouraging results of 96.7% and 89.6% classification accuracy for the calibration dataset and test dataset respectively, and has the potential to serve as the basis for developing a more powerful commercial vehicle classification model which is able to provide further insight into the travel behavior and characteristics of commercial vehicles on roadways today.