Machines with a “brain”

Industrial robots are strong, lightning-fast and precisely in time. But they are so "dumb" that they have to be painstakingly reprogrammed when the workpiece is changed. Service robots, on the other hand, are an example of technical systems with cognitive capabilities. They can recognise their environment, learn and react flexibly to new situations. They are fitted with sensors and devices for gripping or machining objects on mechanical arms, which themselves can be precision-guided towards other objects. But what really matters is how they react to faults or changing environments. Although easy for humans, such occurrences present real challenges for machines. In a bid to develop machines of this kind, research organisations in and around Munich have come together to pool their best scientific expertise in engineering, the natural sciences, psychology, neurosciences and computer sciences in the CoTeSys cluster of excellence (Cognition for Technical Systems), headed by the Technische Universität München (TUM). The other institutions partnering with the TUM are Ludwig-Maximilians-Universität München, the Universität der Bundes­wehr, the Institute for Robotics and Mecha­tro­nics at the German Centre for Aerospace (DLR), and the Max Planck Institute for Neurobiology.