Estimating Link Travel Time with the Mobilizer Video Image Tracking System

Abstract

Travel time is currently believed to be one of the best, if not the best, indicators of the quality of flow and operation on a transportation facility. Travel time is such a good measure because it is easy for both commuters and transportation engineers to understand. The problem with travel time, however, is that there currently does not exist a method that is capable of collecting this measurement automatically, in real-time, and without statistical bias. These elements should certainly be considered as basic requirements for traffic management centers as they progress to meet the future demands of Intelligent Transportation Systems applications. A new technology for measuring travel time is being tested and evaluated by researchers at the University of Washington. This new technology is a video image detection system based on vehicle tracking, called the Mobilizer. This system offers the promise of being able to track individual vehicles passively, and in real-time, along any route monitored by video cameras. This system also provides all of the standard traffic measures reported by previous generation point processing video image systems, but is also capable of reporting link travel times, which cannot be measured directly by the point processing systems. This paper presents the preliminary findings of the Mobilizer system based on results from studies performed at local roadway segments within the Puget Sound region. These preliminary results indicate that the Mobilizer is capable of matching vehicles in successive fields-of-view with a reasonable degree of accuracy and that the travel time estimates provided by the system are statistically valid.