An Assessment of Loop Detector and RTMS Performance

Traffic detectors support most traffic management applications, so it is important that a detector performs as expected. This study evaluates the performance of four loop sensor models and the Remote Traffic Microwave Sensor (RTMS), adding to the body of sensor performance knowledge through the use of new analytical techniques. The study collected contact closure data from all five of the detectors and concurrent video data. Each loop sensor was deployed following Caltrans guidelines for at least 24 hours across dual loop detectors in each lane of I-80, north of Oakland, CA. The research examined various distributions of the individual vehicle actuations from each of the detectors. This exercise found two loop sensor models locked up and not provide any data to the controller although they appeared fully functional from the front panel. The median on-time varied between the RTMS and the loop sensors, as well as from one loop sensor model to the next, which means the occupancy measurements will also change. The video was used to manually validate each vehicle passage over extended periods, preselected at random, with the errors classified by type (e.g., non-detected vehicle) and source (e.g., due to a lane change maneuver). The RTMS exhibited problems due to occlusion and reflections, while two of the loop sensors exhibited non-negligible problems. Then, the detector data are aggregated and used to evaluate the performance of each sensor under standard operation. Next, the results of two earlier evaluations of the RTMS are presented. The reports have not been formally published and to address the omission, this section recaps the relevant portions of those two studies in the context of the results presented above. This first effort considered the aggregated data reported by the RTMS using its internal controller emulation in comparison to dual loop detectors. The RTMS occupancy and flow measures are shown to be noisier than loops, though the velocity estimates are almost as good as those from single loops. The second study considers aggregate measurements from the contact closure data, comparing the RTMS against the dual loop detectors. For reference, the work also compared one loop against another in a dual loop detector, with the spacing between loops being greater than the spacing between the reference loops and the RTMS detection zone. As is illustrated in this study, the RTMS shows lane dependent results