Development of an interactive robot programming method

Today's robot programming methods are not suitable for intuitively teaching robots new tasks. For this reason, the potential of new generations of collaborative robots for production cannot yet be fully exploited. In this work, we propose an active learning framework, in which the robot and the user collaborate in order to incrementally program a complex task. Starting with a basic model, the robot's task knowledge can be extended over time if new situations require additional skills. An on-line anomaly detection algorithm therefore automatically identifies new situations during task execution by monitoring the deviation between measured- and commanded sensor values. The robot then triggers a teaching phase, in which the user decides to either refine an existing- or demonstrate a new skill. The different skills of a task are encoded in separate probabilistic models and structured in a high-level graph, guaranteeing a robust execution and a successful transition between skills. The graph is used as visual feedback for the user, to verify the effects of his demonstrations on the robot's task knowledge. In a user study, our approach is compared to two state of the art Programming by Demonstration frameworks on a real system. An increased intuitiveness and task performance of the method can be shown, allowing shop-floor workers to program conditional tasks for the production with our framework