Augmented reality for programming industrial robots
暂无分享,去创建一个
Existing practice for programming robots involves teaching it a sequence of waypoints in addition to process-related events, which defines the complete robot path. The programming process is time consuming, error prone and, in most cases, requires several iterations before the program quality is acceptable. By introducing augmented reality technologies in this programming process, the operator gets instant real-time, visual feedback of a simulated process in relation to the real object, resulting in reduced programming time and increased quality of the resulting robot program. This paper presents a demonstrator of a standalone augmented reality pilot system allowing an operator to program robot waypoints and process specific events related to paint applications. During the programming sequence, the system presents visual feedback of the paint result for the operator, allowing him to inspect the process result before the robot has performed the actual task.
[1] Ulrich Neumann,et al. Extendible object-centric tracking for augmented reality , 1998, Proceedings. IEEE 1998 Virtual Reality Annual International Symposium (Cat. No.98CB36180).
[2] Gregory D. Hager,et al. Fast and Globally Convergent Pose Estimation from Video Images , 2000, IEEE Trans. Pattern Anal. Mach. Intell..
[3] Eric Foxlin,et al. Circular data matrix fiducial system and robust image processing for a wearable vision-inertial self-tracker , 2002, Proceedings. International Symposium on Mixed and Augmented Reality.