This thesis presents a new approach for a skill framework which is based on BPMN. Graphical modelling of complex processes simplifies their understanding and opens up new possibilities for robot programming. Interactive parametrization combines human robot interaction and traditional interface based configuration to provide a good usability.The skill framework consists of three layers with different abstraction levels thatsupport reuse, scalability and modularity. A real industrial assembly task is used for application of the framework andto evaluate strengths and weaknesses. Embedding the skill framework in anunified automation environment shows its potential to be used in production environments with steadily changing requirements for a flexible production. The skills were tested on a lightweight robot especially designed for collaborative robotics, the KUKA iiwa. As a consequence fundamentals for a basic safety level for human robot collaboration have been considered and implemented.

Universität Bielefeld, Technische Fakultät, Masterarbeit, 2017