Analysis of Travel Times and Routes on Urban Roads by Means of Floating Car Data
R-P Schäfer, K-U Thiessenhusen and P Wagner, German Aerospace Center (DLR), DE
Heavy traffic including congestion can be observed on urban roads all over the world. In view of limitedspace for new roads, it is necessary to improve the exploitation of the existing network. In order to solve this problem the number of telematics applications has been rapidly grown over the last couple of years. This development was mainly pushed by new wireless communication and location sensing technology. Consequently, traffic management and mobility services providing online travel information for road users were established over the last years. Traffic information are usually broadcasted via radioor TV as well as via modern channels like the internet or cellular phone networks. Conventional data sources for traffic information are based on traffic volume measurements by inductive loops or infrared sensors. The major problem of this approach is the reconstruction of travel time and the routes of the vehicles in the road network.
In this paper a more reliable method is discussed analyzing the travel time and routes of vehicles on urban roads. The approach is based on floating car data (FCD) using the global positioning system (GPS). Nowadays an increasing number of commercial vehicles are equipped with GPS devices. In that context, we established a project with a Berlin Taxi company. 300 Taxis equipped with a GPS receiver and a wireless communication device are dealing as FCD data supplier. Once per minutes each Taxi sends its current position to the Taxi headquarters, where the data are processed for the online Taxi client disposition. Using synergies from the Taxi disposition at our institute, a real-time FCD traffic information system was developed. In May 2001 the real time operation of our FCD system wasstarted, exploiting the GPS data of the Taxi disposition software .
Our FCD traffic sensor has big advantages. It is ready for use and there are no additional on-board hardware and software costs.
Since May 2001 several millions of position data from the associated Taxi fleet have been collected. These data are an excellent basis for data mining analyzing the daily variation of travel time on almost all major roads of Berlin.
In the paper we compare the travel times on different types of roads. Additionally, the reasons for fluctuation patterns are discussed. There are several factors why the travel time on a certain time can vary from day to day. The general traffic demand, weather condition, seasonal effects, events or road work influence the vehicles behavior. With this input we are able to generate travel time maps for the Berlin road network.
The results of our research are the basis for real-time traffic monitoring and management as well as for real-time travel time forecast in urban networks. Furthermore, our analyses are used for fleet management of a Berlin waste disposal company. Our routing algorithms are implemented for the optimization of routes of the vehicles as well as for the reduction of fuel consumption.
Association for European Transport