Robot Motion Control Offloading in 5G Network Using Trajectory Interpolation

Abstract

5G technology in manufacturing can enable a more flexible, versatile and efficient usage of production assets, including robots, by shifting the intelligence to the cloud and controlling the devices wirelessly. However, achieving very low cycle times in the range of milliseconds for real-time control of such applications is still a major challenge in currently available 5G networks. In this work, we present a test setup that uses a control-split approach which allows to operate a robotic arm with a variable cycle time by using a piece-wise spline interpolation on the motion trajectory. We verified functionality of this approach over a private 5G network deployed in a factory. Measurement results on the network performance and the robot trajectory accuracy are presented. Our results show that the proposed robot control implementation operates reliably even under high network utilization due to background traffic with only moderate tracking error.