Summary

Red light running is among the top causes of motor vehicle crashes. In 2001, there were nearly 218,000 RLR crashes in the United States, which resulted in 181,000 injuries and 880 fatalities; estimated economic loss is over $14 billion. In the City and County of San Francisco, red light violations cost the local San Francisco economy approximately $40 million each year not including property damage costs.

Countermeasures to reduce RLR crashes fall into three categories: Engineering, Enforcement, and Education. Different types of measures may be more appropriate to address the variety of causes of red light running, for instance, law enforcement and education measures might be more effective to intentional violations while engineering approach appears to be able to address the majority of causes of red light running.

Many states have implemented Red-Light-Camera (RLC) program to enhance the RLR problem. Although there are evidences shown that RLC programs can reduced the total number of RLR crashes, they also increase rear-end and injury accidents.

California PATH Program, in partnership with Caltrans and San Francisco Department of Parking and Traffic (DPT), are conducting a research program to investigate an Adaptive RLR Collision Avoidance concept. This concept intends to address the RLR collision problem at two levels: first, developing an off-line proactive signal timing strategy that provides better design of signal timings in consideration of RLR mitigation and can be easily implemented by updating existing signal parameters; and second, developing an online signal timing refining strategy to adaptively adjust the signal timing to avoid a potential RLR collision.

Objectives

(1) develop off-line signal timing strategies to further increase progression ratio and decrease number of arrivals during yellow and all-red interval so that the probability of RLR occurrences can be reduced proactively in the stage of signal timing optimization;

(2) define the system architecture of an adaptive RLR collision avoidance system, and identify surveillance requirements, hardware/software requirements, and communication requirements for the system;

(3) develop online adaptive RLR collision avoidance algorithms which are able to reduce the possibility of RLR occurrences as well as react to predicted RLR collisions in real-time such that the collisions can be avoided; and

(4) prepare a guide that provides recommendations on applicability, effectiveness, development and deployment of signal operation countermeasures of RLR based on extensive field data collection and analysis.

Red-light running behavior varies from location to location and from person to person, therefore our first efforts will be to further investigate the local red light violation and accident data for further understanding of the problem. Two data sources will be used in the proposed study including the crash data provided by our research partner, the San Francisco DPT and real-time traffic and signal timing data on our El Camino Real test-bed arterial. Data analyses will be conducted to further investigate the correlations among the occurrence of RLR and the influential factors and define RLR crash scenarios, which are fundamental inputs for developing corresponding solutions.

The "proactive" signal timing strategy will provide valuable references for better timing of traffic signals with consideration of RLR mitigation. The end product of the adaptive approach is the adaptive RLR avoidance algorithm that identifies the high-risk vehicles and automatically triggers the traffic signal to extend green, yellow, all-red interval for a needed period of time.

These efforts will culminate in engineering, designing, and testing a demonstration system that integrates the "adaptive" RLR control algorithm with existing closed-loop traffic signal control system; documentation of the developed control philosophy and guidelines on system implementation to facilitate possible statewide deployment.

Scope of Work

Task 1. In-depth investigation of the RLR and crash data obtained from the San Francisco Department of Parking and Traffic and real-time data collected from the El Camino Real test-bed arterial corridor for in-depth understanding of red-light-running and associated crash at signals

Task 2. Develop system architecture for RLR system and implementation means

Task 3. Develop the proactive signal timing strategy, conduct simulation and field evaluation, document guidelines for bettering design of signal timings with consideration of RLR mitigation

Task 4. Develop, field test, and evaluate the adaptive RLR collision avoidance algorithm