Evaluation of Confirmation Lights on Red Light Running Violations by Time-into-Red Phase

Abstract

Red light running (RLR) crashes at urban signalized intersections continue to be a serious safety concern for many communities across the United States. Recent national crash data indicated that 676 fatalities were caused by RLR in 2009, which represents approximately 10% of all intersection crashes and 2% of all roadway fatalities. This research studied confirmation lights and their effect on RLR violations’ time into red. Two sets of confirmation lights were installed at two busy urban intersections in Overland Park, Kansas. With the use of a binomial test of proportions, the study found that there was a significant increase in violations within the first second of the red indication 1 month after deployment. Three months after deployment, there was a decrease in violations over 1 s, but this decrease was statistically not significant. Although there was an observed increase in violation rate, the increase did not translate into violations occurring later in the red cycle. In the 3-month after-period, there was a decrease in the violation rate but an increase in violations within the first second of the red indication and a decrease in violations between 1.0 and 3.0 s. The results for this period were statistically not significant and which suggest that the confirmation light had little effect on a violation’s time into the red. In all periods after deployment, there were no violations occurring after 3 s into the red. This finding means that drivers did not use the confirmation lights to traverse the intersection during the all-red cycle.