In the KI4PED project, researchers in Lemgo developed a “smart traffic light” for pedestrians. It detects people early on, uses artificial intelligence to assess their intention to cross, and controls the green phases as needed. The goal: to reduce waiting times while increasing traffic safety.

Conventional pedestrian traffic lights often do not take into account that people move at different speeds – for example, older people, children, or people with limited mobility. In addition, long red phases often lead to some people crossing the street despite the red light. KI4PED aims to counteract this.

To this end, Fraunhofer IOSB-INA combines modern 3D LiDAR sensor technology with AI-based data evaluation. The sensors detect the environment at the traffic light without contact and anonymously, while AI models analyze movement patterns and predict whether someone wants to cross the street. On this basis, the traffic light can respond more quickly to actual demand – without the need to press a button. The system has been tested at traffic light systems in Lemgo and Bielefeld, among other places.

“With the KI4PED project, we are demonstrating how AI can help in everyday life. The technology understands people and adapts to them, rather than the other way around,” says Prof. Dr. Jürgen Jasperneite, Director of Fraunhofer IOSB-INA. Project manager Dr. Dennis Sprute adds: "Our goal is to make cities safer and more livable—especially for the most vulnerable road users. "