Performance Comparison of Automatic Vehicle Identification and Inductive Loop Traffic Detectors for Incident Detection

Abstract

TransGuide is San Antonio’s multifunctional citywide center that manages traffic through surveillance, incident detection/management, and dissemination of traffic information. This paper reviews automatic incident detection technologies deployed in San Antonio freeways and managed by TransGuide. Traffic and incident data collected from the San Antonio network are used to compare the performance of inductive loop detectors (ILDs) and automatic vehicle identification (AVI) for automated incident detection. California No. 8 and the Texas algorithms were calibrated and tested using the ILD data collected for incident detection. The upper confidence limit algorithm and the Texas algorithm were calibrated and tested using the AVI data collected. When traffic and incident data from the San Antonio network are processed by the four different algorithms, the California No. 8 algorithms applied to ILD data performed best in terms of detection rate and false alarm rate. Automated incident detection (AID) is not currently worth implementing in the AVI system studied, but AID based on AVI data is generally feasible with denser tag penetration and sensor installation.