DFKI research projectDeep learning to become more reliable

DFKI scientists want to combine deep learning methods with formal calculation methods to make them more reliable. To this end, they are teaching a robot to juggle - it is supposed to perform calculated movements and at the same time decide on the next arm movement.