In modern vehicles, more and more driving functions are taken over by the vehicle itself; these currently include automatic distance and lane-keeping assistants, automatic emergency braking assistants, and other functions intended to relieve the driver. The goal is the complete automation of the vehicle, so the driver does not have to intervene while driving. The Society of Automotive Engineers (SAE) divides the various levels of automation into five levels. From level 0 – no automation to level 5 – full automation. Various sensors such as RADAR, LIDAR, cameras, ultrasonic sensors, etc., are used to automate vehicles.
The Laboratory of Lighting Technology is researching the optimization of light distributions for camera systems, as these, like the human eye, are also sensitive in the visual wavelength range. To implement the higher levels, a robust and reliable perception of the environment under different weather and light conditions must be guaranteed. The laboratory’s research focuses on optimizing light distributions to ensure the detection of objects in traffic areas (pedestrians, traffic signs, other vehicles, etc.) under different light and weather conditions. In comparison to adaptive headlight systems, the investigation focuses not on the human being but on various camera sensors that are also characterized in the laboratory. The research results from adaptive headlight systems are also used to design the camera-optimized light distribution for the human driver.
In addition to the reliable detection of objects, communication with weaker road users is also one of the laboratory’s research areas. This includes, for example, recognizing the intention of whether a pedestrian wants to cross the street and has not seen the autonomous vehicle or conveying the intention of the highly automated vehicle to other road users utilizing light signals. In order to find the optimal signaling device, the Laboratory of Lighting Technology relies on real test subjects with real signaling devices and, additionally, on simulations using various VR headsets.