Traffic Woes? New Method Allows Traffic Optimization Over Large Geographic Areas

How can traffic be monitored and controlled more effectively? In the ORINOKO project, scientists have developed methods of determining the traffic situation across a wide area, and have refined processes that enable traffic to be optimally channeled.
Traffic jams on the way to work, to the shops or to a holiday destination – a common experience for most of us. Traffic management systems can provide help. Various concepts and measures are being tested, for example in the transport research project ORINOKO (Operative Regional Integrated and Optimized Corridor Control). The project received funding to the tune of almost three million euros from the German federal ministry of economics and technology BMWi over a period of about three years.
The Fraunhofer Institute for Transportation and Infrastructure Systems IVI in Dresden was among the project partners. The IVI team led by Ulf Jung and Georg Förster performed a variety of tasks. “One thing we did was set up a central database containing a digital map of the road network. A vast amount of relevant measurement data flows continuously into this database,” says Georg Förster. “We also provided software interfaces that enable dynamic data from a variety of sources, such as journey times, traffic volume or tailback lengths, to be used for control and information purposes within the scope of the traffic management system.”
The team is particularly proud at having established a sensor system based on video cameras, which was installed and tested on a trial basis at ten different sites in Nuremberg over the past few months. It can automatically determine certain traffic statistics such as the number of vehicles on the roads or the length of a tailback. These values are continuously fed into a central computer system where they are processed and used to control the traffic. For instance, traffic lights are switched to suit the situation observed by the cameras. “This combination of advanced computer technology and the image processing software developed by us delivers data of a similar quality to those of conventional induction loops, but is much cheaper and more flexible to use,” says IVI head of department Ulf Jung.
The video detector can determine the number of vehicles, their speed, the length of a tailback, and other factors. At present, it is able to analyze up to six traffic lanes simultaneously. The recorded images are processed and interpreted in real time on the spot by a small computer connected to the camera module, which then sends the traffic data and live images to a control center. The new system fills the gap between the established but expensive induction loops and the journey time measurements obtained using sensors in taxis. The video detectors are not only cost-efficient but also deliver a continuous stream of reliable data.