2023: Testing AI algorithms for Yanu service robot

Yanu is a service robot for modern bars. Although it already is a cool fully autonomous AI-powered robotic bartending unit there are still a lot of fantastic ideas and developments to add to this machine. This demonstration project here entailed the development of a ROS-based digital twin to facilitate trajectory planning for a robotic manipulator, integrating AI with control theoretic approaches for algorithm enhancement. The digital twin served as a simulated environment that accurately mirrored the real robotic cell, complete with environmental noise and characteristics, allowing for extensive testing and benchmarking of various path-planning algorithms. This simulation-driven approach significantly accelerated the testing process, minimized risks associated with direct robotic deployment, and ensured the reliability and efficiency of algorithms prior to their application on the physical robot. Upon successful validation within the digital twin, the most effective algorithms were selected for deployment in the actual robotic cell, bridging the gap between theoretical planning and tangible execution. The project was carried out in cooperation with the University of Tartu, Institute of Technology.