Intelligent traffic flow optimisation at traffic lights

The increasing volume of inner-city traffic causes environmental pollution such as noise and exhaust fumes, longer journey times and higher fuel consumption for road users. In addition to increasing the attractiveness of public transport, the use of bicycles and walking, optimising the traffic flow of motorised private transport is an important step towards reducing emissions and improving the use of space in city centres. More efficient use of roads through better traffic management contributes to improved traffic flow and offers the opportunity to redesign road space in favour of other road users. Recording the current traffic situation and the control of traffic lights and traffic signals based on this are the main challenges here.

The KI4LSA project developed a solution for intelligent traffic flow optimisation. This is based on the lane-based detection of vehicle movements, the determination and optimisation of the traffic flow using artificial intelligence methods and the subsequent switching of the traffic light system. The project was able to show that intelligent traffic light control can achieve shorter journey times for road users and reduced traffic emissions. Potential savings are seen particularly in traffic light control for connected traffic junctions, which will be further analysed in follow-up projects.

The KI4PED project has developed and trialled an innovative approach to demand-based control for pedestrian traffic lights. Here, 3D LiDAR sensor technology is used to recognise people in 3D point clouds in compliance with the GDPR and then switch the pedestrian lights as required. In addition to automatic demand requirements, an adaptive adjustment of the pedestrian green phase can be realised, e.g. for larger groups of people or people with certain vulnerabilities. In addition, the early detection of people and the prioritisation of larger groups of people allows waiting times to be reduced.