Quality of Private Sector Travel-Time Data on Arterials

Abstract

Accurate traffic state information is essential for both travelers and transportation agencies. In the past, traffic condition data were usually collected by a government agency using its own sensors. Recently, a number of private sector companies have started selling travel-time and speed data collected using probe vehicles, which provides a viable opportunity to outsource traffic data collection. Because these data sources and their related algorithms are proprietary, the reliability and accuracy of this private sector data is often an important issue for transportation agencies. Previous studies have examined the accuracy of private sector data on freeways, but arterials have not been examined extensively. Arterials represent a fundamentally more challenging environment for probe vehicle data given the larger variance in travel times created by traffic signals and other intermediate access points. In the research, the quality of private sector data on arterials is evaluated by utilizing Bluetooth travel-time data as the ground truth. The evaluation is conducted from two perspectives: the ability to track real-time conditions, and the ability to identify long-term traffic state changes. The study sites are three signalized arterials in the state of Virginia. The results indicate that the private sector data evaluated were not suitable for real-time applications, but could be used to measure long-term traffic state changes for performance measurement programs.