Energy-Efficient Management of Collaborative Robots using Data-Driven Models

This research addresses energy management of collaborative robots that may replace some types of specialized service robots, such as lawn mowers, snow plows, vacuum cleaners, etc. those chaotic movement during work execution is often associated with significant power overruns. To minimize losses, a new data-driven approach to robot modelling and analysis is introduced. It is aimed at reducing energy expenditure at different arrangements of robot joints (joint configurations) needed to locate the tool in the target poses. Presented model-based procedures consider various acceptable locations of the processed parts, travel trajectories, arm properties, and velocity profiles based on ABB and MATLAB libraries and applications. In the case study, a conventional window washer is replaced with a dual-arm IRB14000 (YuMi) robot able to purposefully paving the way, thereby avoiding unnecessary tool movements and power consumption.