Driver Behavior Characteristics at Urban Signalized Intersections

Kerrie L. Schattler, Ph.D.
Assistant Professor
Department of Civil and Environmental Engineering
Wayne State University
Telephone: 313/577-9163
Fax 313/577-8126
kerrie@ce.eng.wayne.edu

and

Tapan K. Datta, Ph.D., P.E.
Professor
Department of Civil and Environmental Engineering
Wayne State University
Telephone: 313/577-9154
Fax 313/577-8126
tdatta@ce.eng.wayne.edu

This paper was submitted for presentation and publication at the 83rd Annual Meeting of the Transportation Research Board, January 2004, Washington, D.C.

Submission Date: November 15, 2003

SUMMARY AND CONCLUSIONS

This study included the evaluation of the all-red interval on red light violations and late exits at nine intersections in Detroit, Michigan. Four approaches of four test intersections (total of 16 approaches) were used as the test sites, which had all-red intervals with lengths determined by the ITE equation. Four approaches of five intersections (total of 20 approaches) were used as control sites, which had no all-red intervals, or all-red intervals with lengths less than that determined by the ITE equation. Specifically, 12 of the 20 control sites had no all-red intervals present in the signal timing, four sites had all-red intervals of 0.5 seconds and four sites had 1.0-second all-red intervals. The results of this study indicated the following:

• The safety consequences of drivers who violate the red signal within a fraction of a second after it turns red, is dependent on the proper use of the all-red interval. If an allred interval of adequate length is used, drivers will have a lower probability of being involved in a right angle crash. If an all-red interval is not present or does not have an adequate length, these drivers have a higher probability of being involved in an angle crash.
• A comparison of the test and control groups in Detroit indicated that the late exit rates (LE per hour, LE per 1,000 through vehicles, and LE per 10,000 vehicle-cycles) at the test sites with all-red intervals of adequate length were significantly lower as compared to the control sites, which had no or inadequate lengths of all-red intervals. However, the statistical test on RLV rates (RLV per hour, RLV per 1,000 through vehicles, and RLV per 10,000 vehicle-cycles) indicated that there were no significant differences in the red light violation experience at the test and control sites in Detroit.
• A comparison of the angle crashes involving red light running indicated that on average, these crashes were significantly lower at the test group, than at the control group of intersections.

Providing adequate all-red clearance intervals can significantly impact intersection safety by reducing the probability of occurrence of right angle crashes, even if drivers run the red light (within a fraction of a second after the onset of the red signal). Thus, even though the presence or absence of an all-red interval does not influence drivers' decisions to violate the red signal, the provision of all-red intervals of adequate length provide significant safety benefits. The impacts of implementing all-red intervals of adequate length (as per the ITE equation) will allow the intersection to operate safely.

Click here for a full copy of the study.

Go Back To Red Light Cameras Studies Page