Advancing Safety Performance Monitoring at Signalized Intersections through use of Connected Vehicle Technology

Abstract

Crash-based safety evaluation is often hampered by randomness, lack of timeliness, and rarity of crash occurrences. This situation is particularly true for technology-driven safety improvement projects that are frequently updated or replaced by newer ones before adequate crash data for a reliable and defensible before-and-after evaluation can be gathered. Surrogate safety data are commonly used as alternatives to crash data; however, the current practice is resource intensive and prone to human errors. The advent of connected vehicle technology allows vehicles to communicate with each other and with the infrastructure wirelessly. Through this platform, vehicle movements and signal status at facilities can be automatically and continually monitored in real time. This study explored the viability of long-term monitoring of connected vehicle data for evaluation of safety performance. A safety monitoring application that used connected vehicle data to detect potential safety indicators at signalized intersections was proposed. As limited saturation of onboard equipment (OBE) was expected in near-term evolution, development focused on a roadside equipment application to process data elements from OBE by way of vehicle-to-infrastructure communications through standard message sets. A microscopic simulation was designed to evaluate effectiveness of the proposed safety indicators in detection of degrading safety performance. A signalized intersection test bed was created in VISSIM, while the wireless communications capability and the application were implemented in a car-to-devices application programming interface. Evaluation results indicated that the application could effectively detect changes in safety performance at full market penetration. Future research needs to quantify the combination of penetration rates and monitoring periods that can yield effective detection of changes in safety performance at intersections with varying operating characteristics.

Safety Evaluation of the Advanced Stop Assist System in Connected Vehicle Environment

“…A study by Songchitruksa ( 7 ) proposed a methodological framework designed to extract and compute safety indicators from vehicle movement, intersection description, and signal data available within the RSE at an isolated signalized intersection. The study classified the safety indicators as either single-OBE or dual-OBE, depending on the number of OBE-equipped vehicles that needed to be monitored.…”

“…The frequency of these events on a particular roadway section can be related to disturbances in traffic flow, and can indicate the risk of rear-end crash occurrence. According to Songchitruksa ( 7 ), MaxD-based events are not always a precursor for rear-end crashes. Mixed results of potential indicators can occur with unsafe scenarios.…”

“…Several operational assumptions were made to model V2I communications in this study. Some of these assumptions were derived from a study by Songchitruksa ( 7 ). These assumptions are listed below:…”

“…The history of CV can be traced back to 2003 when the U.S. DOT first launched the Vehicle-Infrastructure-Integration (VII) program ( 7 ). The initial objective of VII was to address the traffic safety problems through high-speed wireless communications among Vehicle-to-Vehicle (V2V) and V2I.…”

“…Some of these applications include Cooperative Intersection Collision Avoidance System (CICAS) framework and its related applications such as CICAS-V (traffic signal violation), CICAS-LTA (signalized left-turn assist) and CICAS-TSA (traffic signal adaptation) ( 14 ). Songchitruksa ( 7 ) proposed safety performance monitoring using V2I data. However, the study proposed a framework for monitoring safety performance at signalized intersections by means of CV technology.…”