Humanoid robots have the potential to make a significant impact in various areas, including healthcare, education, and entertainment. Low-cost and open-source humanoid robots have made them more accessible and affordable for these applications, enabling researchers, educators, and enthusiasts to experiment with them without the need for a large budget or extensive technical expertise. However, deploying whole-body control on open-source humanoid robots  poses several challenges due to limited sensing capabilities and computational resources, inaccurate dynamics models, and simplified mechanical designs. Besides advancing performance boundaries, it is important to acknowledge such limitations and work within them. This can lead to unique and effective solutions, as shown by the Researchers at University of Bonn.
Apart from winning the X-Prize teleoperation challenge, they won several RoboCup championships with their open-source humanoid robot NimbRo-OP2X. The presentation will focus on the experience and insights gained from overcoming these challenges and enabling effective whole-body control on low-cost/open-source humanoid robots.