Development of an Augmented Reality Interface for Intuitive Robot Programming

Kurzfassung

As the demand for advanced robotic systems continues to grow, the need for new technologies and techniques that can improve the efficiency and effectiveness of robot programming is imperative. The latter relies heavily on the effective communication of tasks between the user and the robot. To address this issue, we developed an Augmented Reality (AR) interface that incorporates Head Mounted Display (HMD) capabilities, and integrated it with an active learning framework for intuitive programming of robots. This integration enables the execution of conditional tasks, bridging the gap between user and robot knowledge. The active learning model with the user's guidance incrementally programs a complex task and after encoding the skills, generates a high level task graph. Then the holographic robot is visualising individual skills of the task in order to increase the user's intuition of the whole procedure with sensory information retrieved from the physical robot in real-time. The interactive aspect of the interface can be utilised in this phase, by providing the user the option of actively validating the learnt skills or potentially changing them and thus generating a new skill sequence. Teaching the real robot through teleoperation by using the HMD is also possible for the user to increase the directness and immersion factors of teaching procedure while safely manipulating the physical robot from a distance. The evaluation of the proposed framework is conducted through a series of experiments employing the developed interface on the real system. These experiments aim to assess the degree of intuitiveness provided by the interface features to the user and to determine the extent of similarity between the virtual system's behavior during the robot programming procedure and that of its physical counterpart.