Technical University of Munich (TUM)
With increasing use of robotics in manufacturing environments, systems to ensure the safety of humans near robots are moving into focus. While physical barriers such as cages ensure safe human-robot interaction, they restrict movement and are take up valuable factory space.
The invention is a human localization system, consisting of various sensor technologies, including vision (point-cloud and 2D), LIDARs, and touch sensors. Analysis is performed by several algorithmic techniques, such as probabilistic methods (Bayesian filtering), neural techniques (e.g. convolutional networks), and nonlinear observers.
We are seeking companies with manufacturing facilities which use automated production lines, but still require humans for the execution of complex tasks.
We are looking for new partners to aid us in the improvement of our system, Proof of Concept testing, and eventual certification of our system.