Goal of the research topic
Today, the use of robot systems is essentially restricted to highly optimized industrial scenarios with large lot sizes and precisely specified production guidelines. This is linked to a high effort for integration and programming of these systems. On the other hand, small and medium sized enterprises in particular have a need for flexible automation due to small lot sizes, short product cycles, and a great variety of products. This drives a need for easy programming without expert knowledge as well as intelligent solutions that are adapted to a particular domain.
Challenges arise in the transfer of classic programming methods to adaptive process descriptions, that can be understood by domain experts and which can be transferred to different systems. The integration of human workers into the production process requires an improvement in terms of efficiency and ergonomics in order to increase the acceptance of this type of cooperation.
Research topic outcome
The focus is on the improvement of usability, flexibility, and integration of robot systems in order to be able to adapt them to changing requirements without a need for expert knowledge. In order to accomplish this, new operating concepts are being researched and evaluated regarding their applicability for human-robot cooperation.
A robust interaction with a human requires the integration and evaluation of different input and output modalities. New and intuitive operating concepts for specific domains are created through the integration of cognitive skills in connection with semantic descriptions of objects, processes, tasks, and robot components. The programming of the systems is now based on a task description and is adapted to the language of the domain expert. With the help of process and common knowledge, underspecified instructions can be complemented automatically through logical inference. For the full realization of such flexible systems, control concepts for actuators and sensors of the robot system have to be extended as well. Through this, a holistic architecture is realized that combines task description, planning, and execution on the robot.