Wireless Interference Estimation Using Machine Learning in a Robotic Force-Seeking Scenario

Abstract

Cyber-physical systems are systems governed by the laws of physics that are tightly controlled by computer-based algorithms and network-based sensing and actuation. Wireless communication technology is envisioned to play a primary role in conducting the information flows within such systems. A practical industrial wireless use case involving a robot manipulator control system, an integrated wireless force-torque sensor, and a remote vision-based observer is constructed and the performance of the cyber-physical system is examined. By using readings from the remote observer, an estimation system is developed using machine learning regression techniques. We demonstrate the practicality of combining statistical analysis with machine learning to indirectly estimate signal-to-interference of the wireless communication link using measurements from the remote observer. Results from the statistical analysis and the performance of the machine learning system are presented.