Experimental Automated Incident Detection Study Achieved 30 Percent Faster Incident Detection and 25 Percent Faster Congestion Dissipation in Simulations.
Simulation-Based Study Tested Automated Incident Detection Techniques Using Vehicle-to-Infrastructure (V2I) Communications.
An Improved Automatic Traffic Incident Detection Technique Using a Vehicle to Infrastructure Communication
Summary Information
Accurate and timely detection of traffic incidents can play an important role in improving roadway safety and understanding typical traffic behaviors. In this study, a methodology was developed to estimate traffic incidents using a Hybrid Observer (HO) approach. The detection process leveraged an enhanced Automatic Incident Detection (AID) technique based on lane-changing speed patterns on highways. Experimental results and analysis were derived from simulation-based evaluations.
METHODOLOGY
This study first developed the connection between vehicles and Roadside Units (RSUs) by using a beacon mechanism. This would facilitate information exchange once the vehicles get access to a wireless medium. Second, the researchers utilized a probabilistic approach to collect the traffic data, by using a Vehicle-to-Infrastructure (V2I) communication. Third, the traffic incident was estimated by using an HO method, providing an accurate estimation of an event occurring. Finally, to detect the traffic incident accurately, the probabilistic data collected through V2I communication was applied based on a lane-changing speed mechanism.
FINDINGS
- Results showed that the proposed method led to around 30 percent faster detection of traffic incidents and 25 percent faster dissipation of traffic congestion.
