Innovative Dynamic Queue-Length Estimation Using Google Maps Color-Code Data

Queue length is an important parameter for traffic-signal priority systems emergency vehicles. Instead of using conventional detector data, this paper investigates the feasibility queue-length estimation Google Maps color-code data via random forest (RF) and gradient-boosting machine (GBM) methods. Alternative ways specifying independent variables from are also investigated. Additionally, models separated by peak or off-peak periods presence absence adjacent upstream signalized intersections. The results show that performance predicted RF GBM methods similar in all cases. Although error values both relatively high, they considerably lower than those obtained estimates historical data. variable-importance analysis importance red band near intersection significantly higher other a direction without prior intersection. For with intersection, varies, depending on period (peak off-peak). Since available cover most world intersections, proposed approach provides cost-effective option cities no detectors installed.