Abstract
Compact wireless magnetometers offer attractive vehicle detection ability at signalized intersections because their installation requires minimal pavement cutting and the detectors are less likely than saw-cut inductive loops to malfunction because of pavement failure. A study was done at an instrumented intersection to evaluate the performance of wireless magnetometers at operating signalized intersections. A test bed was constructed with colocated inductive loop and wireless magnetometer detection zones. A 5-day analysis period was conducted for each of two left-turn pockets at an actuated, coordinated signalized intersection. Discrepancies between the detection and nondetection states were quantified with highresolution log data of traffic events, and 240 h of data collection that was ground-truthed by visual inspection of video recordings of the detection zones. Behavior of detector state changes was also characterized. Wireless magnetometers were found to perform similarly to loops in relation to missed calls and had a slightly higher tendency to generate false detection calls. Detection state changes in the wireless magnetometers had typical (85th percentile) reporting latencies of 0.2 s or less for activation and 0.5 s or less for state termination. The paper concludes by recommending 8-ft spacing of the sensors adjacent to the stop bar to minimize missed calls.
Get full access to this article
View all access options for this article.