Artificial intelligence revolutionizes traffic flow: Regensburg woman wins!
Leonie Weiß from the University of Regensburg wins the state competition Jugend forscht 2025 with an AI project for traffic light optimization.

Artificial intelligence revolutionizes traffic flow: Regensburg woman wins!
On April 7, 2025, the state youth research competition took place in Klingenberg, in which early student at the University of Regensburg, Leonie Weiß, won first place. The competition, which is celebrating its 60th anniversary this year, brought together the most successful participants from the Bavarian regional competitions.
Leonie Weiß impressed the jury with her innovative project to optimize construction site traffic lights. This groundbreaking work uses object recognition-based decision-making processes of a trained artificial intelligence (AI). The adaptive control of the traffic lights not only reacts to the current traffic flow, but can also detect objects such as cars, motorcycles, trucks, bicycles and pedestrians in different weather conditions such as rain and fog.
Innovative technologies in transport
The project aims to avoid traffic jams, reduce emissions and optimize transport routes in urban areas. These technological approaches are particularly relevant as the challenges in traffic flow are constantly increasing. The jury, consisting of professors, business specialists and computer scientists, was impressed by the professional implementation and real-time simulation of the AI-supported traffic light control.
Leonie Weiß's success now leads her to the final round of the competition in Hamburg, which will take place at the end of May. With these outstanding achievements, she has also received several awards in the field of artificial intelligence. Her work focuses on developing technologies that address societal challenges.
Intelligent traffic light systems
The developments in the area of traffic control are not limited to Leonie Weiß. The “KI4LSA” project at the Fraunhofer Institute for Optronics, System Technology and Image Analysis IOSB pursues a similar goal. The aim here is to achieve intelligent, predictive traffic light switching using artificial intelligence. The project is supported by partners such as Stührenberg GmbH and Cichon Automationstechnik GmbH as well as Stadtwerke Lemgo GmbH.
A central concern of this project is to replace the current, rule-based traffic light controls, which often cannot be adapted to complex traffic situations, with high-resolution camera and radar sensors. These sensors enable more precise traffic detection and enable real-time analysis of vehicle numbers and waiting times.
| Project name | Main goal | technology | Project partners |
|---|---|---|---|
| KI4LSA | Intelligent traffic light switching | Artificial Intelligence, Deep Reinforcement Learning | Fraunhofer, Stührenberg GmbH, Cichon Automation Technology GmbH |
| KI4PED | Demand-based control of pedestrian traffic lights | Artificial intelligence, LiDAR sensors | Including Stadtwerke Lemgo |
The “KI4PED” project also focuses on pedestrian safety and the automation of traffic light systems through the use of LiDAR sensors that can capture pedestrians as 3D point clouds. This could significantly shorten waiting times when pedestrian traffic is high and significantly reduce the risk of dangerous crossings.
The combination of these technologies shows how artificial intelligence can be used in the transport sector to optimize traffic flow and improve the quality of life in cities. Leonie Weiß's successes and the results of the research projects illustrate the increasing need for intelligent solutions to the challenges that our society has to overcome in the area of transport.