Reinforcement Learning (RL) in Robotics
Reinforcement Learning (RL) is one of the three major areas of machine learning. It is based on the premise that an agent acts in an environment, thereby changes the current state of the environment and receives a corresponding reward (or punishment) in return. These steps are repeated several times, with the agent's goal being to maximize its overall resulting reward.
The TTZ Günzburg is involved in the practical application and research of RL in the field of robotics. It has dedicated itself to the adaptation of human behaviour (Imitation Learning) in the field of industrial robotics. Given the current shortage of skilled workers and demographic change, which predicts a future decline in the workforce, our field of research is proving to be particularly relevant.
This means that supply chain mechanisms need to be adapted to ensure the supply maintenance. The use of robots in industry, in combination with RL, offers a promising approach to this.
Our goal is to automate, optimize and accelerate production processes with the help of RL. We are driven by the research question of how RL systems in combination with Robot Learning can be developed so that robot systems adapt human behaviour more effectively and be quickly adapted and used for multi-task applications in industrial production.
