Institute for Anthropomatics and Robotics - Intelligent Process Automation and Robotics Lab

RoSylerNT: Learning and robot-based systems for neuro-muscular training

Regular body training has a high potential regarding the prevention and the therapy of aging-related diseases. Assistance and trainings system using optimal stimuli while minimizing the risk of injury caused by overstraining would provide a high benefit in the context of rehabilitation. A robust and safe technical implementation of such a system comprises multiple safety-critical interaction concepts and is extremley challenging. In contrast to current systems focused on compliance in human-robot-interaction, this project robots will actively apply higher forces on humans for the purpose of physical training. For this the deep understanding of the bio-mechanics of the human body is crucial.

The goal of this project is to implement safe and robust physical interaction skills for robot systems. Machine learning techniques will be used to empower robotic system to provide robust interaction mechanisms and learn interaction behavior.

The realization of this basic and fundamental interaction skill will open up a broad spectrum of new application scenarios outside of the classical industry sector.