Big Techday 26: Let's dance! - Teaching your robot some moves with reinforcement learning - TNG

Let's dance! – Teaching your robot some moves with reinforcement learning Has this ever happened to you: You just spent €50,000 on a humanoid robot, but it's not very fun at parties, so you have to train it to dance using reinforcement learning in a simulation environment on your €100,000 GPU cluster? What do you mean "NO"? In this talk, Thomas Endres and David Winderl will teach you how you can train your humanoid robot to dance using Reinforcement Learning (RL) in simulation environments. They're gonna kick things off with an introduction into RL, covering the basics of policies and rewards. After that, they’ll take their first steps in simulation environments with the introduction of NVIDIA Isaac Sim and the Unitree RL Lab. Our speakers will run through the steps of shaping a walking policy, one hilarious fall at a time. Finally, they're going to shake things up with a dancing policy: You will learn how you can retarget human motion from videos to the Unitree G1 robot and train an imitation policy that can dance just as poorly as you. About the speakers: Thomas Endres: Thomas Endres is a Managing Partner at TNG Technology Consulting in Munich. Besides his "normal" work for the company and the management of client projects, he develops various prototypes together with the Innovation Hacking Team - including an augmented reality application that shows the world from an artist's perspective, real-time deepfakes, or an AI for generating presentations. He works on applications in the field of AR/VR and AI and gesture control, in order to make, for instance, quadrocopters fly autonomously or to control them touch-free. He is also involved in various open source projects. Thomas holds a degree in computer science (TU Munich) and is a passionate software developer. As Intel Software Innovator and Black Belt, he presents new technologies such as AI, AR/VR and robotics around the world. His achievements include, among others, a JavaOne Rockstar Award and several Best Speaker Awards. David Winderl: David Winderl works as a Software Consultant at TNG, where he implements generative AI solutions and cloud-based data infrastructure. He currently contributes to an agent-based chatbot integration that performs data analysis for enterprise data lakes. Beyond client projects, David develops proof-of-concept implementations in TNG's Innovation Hacking initiatives. Before joining TNG, he earned his bachelor's degree in Informatics at LMU Munich and his master's degree at TU Munich. Fabian Lübbe: Tobias Lübbe is a Software Consultant in TNG’s Innovation Hacking team. His work focuses on Physical AI and the development of innovative robotics showcases. Currently, he is exploring the use of Robot Foundation Models (RFMs) across different robotic systems. Before joining TNG, he studied Mechatronics at the Karlsruhe Institute of Technology (KIT) and Robotics, Cognition, Intelligence at the Technical University of Munich (TUM). #tngbtd