Gesture Control of a Lightweight Industrial Robot Supported by Augmented Reality

This paper presents an approach to intuitive robot control by gestures made by a human operator supported by augmented reality. Unlike to the concepts of different robot manufactures to guide the robot using the manual control pendant, gesture control would be a universal method to drive the manipulator arm, where no special knowledge is necessary. For this purpose, we develop a set of gestures to move the robot in Cartesian space and to operate the gripper. In order to that, it is possible to program handling tasks, which are very relevant concerning robotic tasks. Furthermore, our scenario may be enhanced towards teleoperation. Augmented reality is then used for verification and modification of the robot program. In contrast to several approaches which use expensive hardware, for example head mounted displays, our concept needs only a monitor or a video projector. We compare the performance of gesture based robot control with conventional methods, such as moving the manipulator with the teach pendant or by hand guidance. For this purpose, a simple handling task is configured and the programming times are analyzed.