1367931 - Investigating Road User’s Compliance at Signalized Intersections: A Data-Driven Approach

According to the report published by AAA Foundation for Traffic Safety, disregarding red light signals resulted in more than two fatalities each day in 2019. Furthermore, based on the fact sheet published by the Arizona Department of Transportation (ADOT), in 2020, Arizona experienced the highest number of RLR-related fatal crashes since 2011; with 47 RLR-related fatal crashes only during 2020. Exploring RLR exposure, identifying potentially relevant variables associated with RLR frequency and severity, and proposing technical or enforcement remedies are all essential steps toward reducing the number of red-light violations and maintaining a safe journey for all road users. In this effort, we aimed to identify factors that influence RLR violation (such as yellow light interval, red-clearance interval, and delay). To do so, multiple models were trained based on the severity of RLR. The severity of RLR can be defined based on the time when the violations happened. Thus, the violations that happened within the first few seconds of the signal light turning red are less severe than those that happened later after the red onset. To validate the results obtained from statistical modeling and analysis, several intersections, equipped with smart video sensors capable of collecting various types of data, such as high-resolution event-based, vehicle count, and vehicle red-light violation data in the City of Phoenix were selected. Then, by utilizing a before-and-after study framework, the following questions were examined: 1) whether the ITE 2020 guideline can enhance the safety of signalized intersections by reducing the frequency of RLR violations, and 2) identify the appropriate amount for yellow time and red-clearance time that will result in a reduction in RLR frequency. It is anticipated that the results of this study will support DOTs to understand the impact of different intersection and corridor-based characteristics on the frequency and severity of RLR. In addition, based on the experimental design several countermeasures, including revision of the signal timing plan and increasing/reducing the amount of yellow or red-clearance intervals will be identified.