Abstract
The Bavarian Road Administration (BRA) is responsible for all extra-urban traffic signals in the German state of Bavaria. BRA is instructed by ministerial decree to establish a quality management system for its traffic signals, including periodic reviews of relevant performance indicators at the signals. A proposed practice is to interview the participating 19 state building offices. However, such an approach would be time-consuming and not sufficiently consistent. This paper presents an alternative multi-step approach to support the required signal reviews in a large-scale generic analysis. The procedure is based solely on anonymized GPS probe data from navigation devices and works in the following way. A special fine-granular digital map with the locations of traffic signals is used as the starting point. Then, in a generic geofencing step, the areas of influence around the signalized junctions are determined, and the probe data are map matched onto that qualified network. From these map-matched data, paths are identified that represent the maneuvers of the probe vehicles through the intersection area. Finally, delay times and other indicators of all the traveled paths at a junction are calculated through comparison of all individually measured path travel times with reference travel times on the paths. These reference travel times also were calculated from GPS probe data that analyzed path-specific speed distributions. This is the first time an approach has been developed that allows automatic analysis of a huge number of traffic light control systems at intersections at one time. The data and figures for 2,300 Bavarian traffic signals were provided to the BRA in support of the evaluation of their traffic signals.
Get full access to this article
View all access options for this article.