Video Friday: Humanoid Learns Tennis Skills Playing Humans
IEEE Spectrum's Video Friday highlights recent advancements in robotics, featuring a humanoid robot that learns tennis skills through the LATENT system, which utilizes imperfect human motion data to replicate dynamic athletic behaviors. The edition also showcases a wind-powered robot inspired by Strandbeests from Cranfield University, designed for exploring hostile environments. Additionally, Sharpa demonstrates a breakthrough in bimanual manipulation, where a robot peels an apple using dual dexterous hands powered by the MoDE-VLA AI model, combining vision, language, and touch data for precise control. The Robotics and AI Institute presents its Universal Mobile Vehicle (UMV), trained via NVIDIA Isaac Lab to perform complex maneuvers like jumping and flipping. Finally, a collaborative project between Tesollo and Hanyang University ERICA introduces Finger-Tip Changer technology, enhancing robotic gripper versatility. These developments underscore significant progress in humanoid agility, dexterous manipulation, and autonomous mobility, illustrating the growing integration of advanced AI and reinforcement learning in robotic systems.
Wire timeline
Video Friday: Humanoid Learns Tennis Skills Playing Humans
IEEE Spectrum's Video Friday highlights recent advancements in robotics, featuring a humanoid robot that learns tennis skills through the LATENT system, which utilizes imperfect human motion data to replicate dynamic athletic behaviors. The edition also showcases a wind-powered robot inspired by Strandbeests from Cranfield University, designed for exploring hostile environments. Additionally, Sharpa demonstrates a breakthrough in bimanual manipulation, where a robot peels an apple using dual dexterous hands powered by the MoDE-VLA AI model, combining vision, language, and touch data for precise control. The Robotics and AI Institute presents its Universal Mobile Vehicle (UMV), trained via NVIDIA Isaac Lab to perform complex maneuvers like jumping and flipping. Finally, a collaborative project between Tesollo and Hanyang University ERICA introduces Finger-Tip Changer technology, enhancing robotic gripper versatility. These developments underscore significant progress in humanoid agility, dexterous manipulation, and autonomous mobility, illustrating the growing integration of advanced AI and reinforcement learning in robotic systems.
IEEE Spectrum