Summary

Traffic congestion increases each year on the nation's highways and the annual costs are estimated to be over $67 billion. Intelligent Transportation Systems (ITS) have been developed to operate the network more efficiently by responding to conditions. These active controls rely on sensors to monitor what is occurring on the roadway. The majority of data upon which ITS services are based come from inductive loop detectors. Yet conventional loop detectors have a surveillance region on the order of six feet, and conditions over the rest of the freeway must be inferred from these point measurements. Unfortunately, in the presence of congestion, these local measurements are not representative of the entire freeway and it can take a long time for slow moving signals to propagate through the traffic stream before reaching the sensors. Efforts have begun to address these shortcomings with vehicle reidentification methodologies, in which a vehicle measurement made at a downstream detector station is matched with the vehicle's corresponding measurement at an upstream station on the same facility. The resulting matches allow for accurate travel time measurement between stations and enable delay detection before the effects of queuing behind an incident are (locally) observable at a detector station. The research is driven by the constraints faced by operating agencies. We use the existing hardware normally employed to calculate the point measurements, but replace the software and communications protocols. In conjunction with PATH MOU's 3010 and TO 4107, we have deployed the first operational, real time, vehicle reidentification system on the Berkeley Highway Laboratory (BHL). Having proven the feasibility of vehicle reidentification, the primary goals of this proposal are to:

1. Improve the vehicle reidentification algorithms by extending their operating range across major geometry changes, e.g., the merge or diverge of two freeways.
2. Improve velocity and length estimation at single loop detectors, both to enable the deployment of the vehicle reidentification algorithms on these facilities and to improve conventional traffic monitoring when using single loop detectors.

The first improvement is a necessary step before large-scale deployment of the vehicle reidentification methodology in a district equipped with dual loop detectors. While the second improvement is a necessary step before any deployment of the algorithms in a district equipped with only single loop detectors. Advances in these traffic detection applications would enable better control of traffic flow, improving the movement of virtually all people and goods in California while reducing vehicle emissions and fuel consumption. The algorithms are compatible with the existing detection and communication infrastructure. Furthermore, they would allow continued collection of conventional flow, occupancy and velocity.