Validating predicted rural corridor travel times from an automated license plate recognition system: Oregon's frontier project

This paper summarizes the results of a field validation of a corridor travel time prediction system that uses automated license plate recognition the frontier travel time project. As part of the frontier project, the Oregon Department of Transportation deployed a video image processing system with license plate recognition and privacy-protecting data encryption, a central server, and proprietary algorithms to predict corridor travel times on a 25 mile section of rural highway in northwest Oregon. The system predicted travel times were compared to data independently collected by probe vehicles equipped with Global Positioning System devices. The comparison shows that the predicted travel times were not statistically different than the travel times observed by the probe vehicles. Despite attempts to validate the system under congested conditions, all comparisons were made under essentially free flow travel. Further validation of the system in a congested corridor with alternate routes is recommended.