Man-made climate change has attracted increasing attention in recent years. Possibilities for reducing CO₂ emissions are therefore of great interest. In Germany, car exhaust fumes are responsible for almost 20% of all CO₂ emissions. Particularly in cities, CO₂ emissions are high, since constant braking and acceleration on the one hand and low speeds below 50 km/h on the other hand drive up fuel consumption and thus CO₂ emissions. Due to its geography, Bremen is strongly affected by this, since the surrounding rivers allow only a few connecting roads into the surrounding area, so that these roads have a high traffic density. The traffic management center in Bremen has the ability to control a large number of traffic lights and thus improve the traffic flow.
In the DiSCO₂ project, a so-called digital twin - a simulation of Bremen city traffic - is to be developed. This allows reliable predictions about the traffic flow and the associated CO₂ emissions depending, for example, on the day of the week, the season, the weather situation and events. In addition, anomalies can be detected and the effects of road works and major events on the traffic flow can be captured and reduced.
For about 10 years, traffic density in the entire city area has been recorded with induction loop detectors at approx. 600 measurement locations. These recordings are suitable for the development of a data-based hybrid model, i.e. a model whose parameters are optimized using methods from the fields of Big Data and machine learning. Initially limited to one measuring point, the predictions will be extended to the entire urban traffic in the course of the project. Since the measurements can be used in real time, a further goal is the intelligent control of light signal systems to improve traffic flow and thus reduce CO₂ emissions.