Analysis of Driver and Vehicle Characteristics Involved in Red-Light Running Crashes: Isfahan, Iran

Abstract

Red-light running is a prevalent sight at road intersections that vehicles pass without caring for the light. A red-light runner not only risks his life, but also the safety of other users. Elaborating features of at-fault drivers that issue crashes could result in recognizing important factors and lead decision-makers to find out actions they could start taking to prevent them. This paper aims to identify the driver and vehicle characteristics affecting red-light running crash occurrence at signalized intersections of Isfahan province in Iran. Based on the 2012–2016 Isfahan crash database, the classification and regression tree technique along with Quasi-induced exposure concept were used in this research for five independent variables and a target variable of “driver status” with two levels of at-fault and not-at-fault. According to the dataset, 9765 drivers were involved in red-light running crashes during 5 years. It was determined that vehicle type, license type, driver age, and education, respectively, are significantly associated with the red-light running crashes occurrence. The results revealed that vehicle type is the most important variable that affects red-light violations; this could be due to their various skid marks and maneuverability. Also, drivers with an age range of 18–22.5 years are usually at-fault as they have more risky behaviors and less driving experience. The tree model separated drivers into three age groups: younger than 22.5, 22.5 to 51.5, and older than 51.5. The youngest and oldest driver groups have the largest probability to be at-fault. Besides, drivers with vehicle type two-wheel and “other”, drivers who are younger than 22.5 with the automobile, and drivers whose driving license is type 2 are more likely to be at-fault in red-light running crashes. Some countermeasures such as education programs in driving skills and hazard concept could be helpful. The education level of the driver is a relatively new variable in traffic safety research used in this study, which indicated that it has a significant effect on drivers being at-fault. That is, by increasing the education level, they are less likely to be at-fault. Efficient countermeasures could be separating theoretical certificating classes according to the education level of the driver with more emphasis on violations in lower education levels, increasing the amount of traffic fines and prohibiting drivers with poor driving history for a period of time.