The award-winning development by Andreas Blug (center, Fraunhofer IPM), Felix Abt (left, Institute for Beam Tools at the University of Stuttgart) and Leonardo Nicolosi (Institute for Fundamentals of Electrical Engineering and Electronics at TU Dresden) contributes to the quality assurance of laser welding processes: The three project partners presented their novel laser welding sensor technology, on the left on the arm of an industrial robot. A new type of camera is used here, which looks directly into the vapor capillary of the welding process coaxially with the laser beam. It shoots up to 14,000 images per second. A so-called 'cellular neural network' processes the information. For the first time, the laser power adapts to process fluctuations - such as dirty laser optics, feed fluctuations or changes in focus position - in fractions of a millisecond. The quality of the welded joint is improved because there is less excess power and therefore less welding fumes and spatter. Energy consumption is also reduced by around five percent.