Evaluation of Microwave Radar Vehicle Detectors at a Signalized Intersection under Adverse Weather Conditions

Abstract

Two microwave radar systems for vehicle detection, manufactured by Wavetronix and MS SEDCO, were evaluated at stop bar and advance zones of a signalized intersection under three adverse conditions: (a) high wind speeds, (b) rain, and (c) fully or partially snow-covered roadway. Data under normal (favorable) weather conditions were also analyzed and used for comparison. Results show that the performance of the two systems deteriorated during adverse weather, and the type of detection errors and their frequency were system dependent. In general, wind increased false calls at the advance detection zone by more than 50%, depending on the location and sensor type; snow-covered roadway had more widespread effects and increased false calls at stop bar and advance zones, depending on the system, and also increased missed calls at stop bar zones; and rain affected the two systems by increasing false calls and stuck-on calls at the stop bar and false calls at the advance zones. Data also showed that frequencies of false and missed calls were likely to increase as precipitation intensity and wind speeds increased. Results from this evaluation show significant effects of adverse weather on microwave radar detection systems, providing information to researchers and practitioners regarding the potential performance of similar setups under such conditions. Details on the detection errors offer insight on situations that should be monitored and could be used to improve performance under adverse weather.