ILM - Institute for Laser Technology in Medicine and Measurement Technique

Poor weather and visibility conditions such as rain, fog, or snowfall are a central challenge for autonomous driving. The lack of sufficient realistic data makes it difficult to train AI-based systems and can, in extreme cases, lead to system failure. The ILM has the expertise to measure, simulate, and physically reproduce light propagation in scattering media. This allows driving scenarios with fog of any quality and visibility range to be realistically augmented. Based on this, AI systems can be specifically trained and lidar and sensor systems validated. This creates an important foundation for safe autonomous driving in limited visibility conditions. Additionally, other applications are conceivable, such as difficult visibility conditions in operating rooms or the detection of objects under scattering media.

At the ILM, light propagation in natural objects can be precisely determined. The basis for this is detailed measurements of the optical material properties, which are transferred into physically based 3D models and simulations of light propagation. Building on this, the models enable highly accurate visual reproduction of biological and technical structures. Algorithms developed at the ILM also allow for extremely fast rendering of these objects, enabling efficient visual prediction of their appearance. A particular application focus is on dentistry: physics-based rendering enables realistic and predictable previews for precise dental restorations. This allows the fit and coloration to be assessed early, reduces rework, and significantly decreases the coordination effort between patients and dental technicians.

Sepsis is a life-threatening systemic infection response where rapid and targeted therapy is crucial for treatment success. In clinical diagnostics, the early identification of the causative pathogen and possible antibiotic resistances poses a major challenge. In the ASTANA project (ILM, Hahn-Schickard – funded by the BMFTR), sepsis pathogens were examined in microfluidic IVD cartridges in a multitude of tiny chambers, each containing different antibiotics. Using holographic methods, the pathogens could be counted individually and their growth precisely tracked. Determining the growth rates enables significantly faster identification of resistances and helps to considerably shorten the time until the appropriate therapy decision. Thus, the method makes an important contribution to accelerated and more precise sepsis diagnostics.