Flexible 3D Trajectory Teaching and Following for Various Robotic Applications

This paper addresses an actual problem regarding complex industrial robot applications. Based on the fact that no CAD model for the processed parts is available, the application presented here consists in a 3D accurate path following, applicable in various robot tasks. Using a sensor-based 3D path learning procedure available in automatic or manual mode, the 6 d.o.f industrial robot will be able to reproduce in real-time the learned trajectory. Calibration and synchronization aspects are presented, and experimental results are provided and analyzed.