{"id":221581,"date":"2025-01-17T11:09:45","date_gmt":"2025-01-17T15:09:45","guid":{"rendered":"https:\/\/inspenet.com\/noticias\/exbody2-the-humanoid-robot-that-imitates-the-fluidity-of-human-movements\/"},"modified":"2025-01-17T14:23:36","modified_gmt":"2025-01-17T18:23:36","slug":"exbody2-the-humanoid-robot-that-imitates-the-fluidity-of-human-movements","status":"publish","type":"noticias","link":"https:\/\/inspenet.com\/en\/news\/exbody2-the-humanoid-robot-that-imitates-the-fluidity-of-human-movements\/","title":{"rendered":"ExBody2: the humanoid robot that imitates the fluidity of human movements"},"content":{"rendered":"\n

A team of researchers from the University of San Diego<\/strong> , UC Berkeley<\/strong> , MIT<\/strong> , and NVIDIA<\/strong> has developed ExBody2<\/strong> , a robotic platform that has the ability to replicate complex human movements with great precision, integrating advanced simulations and reinforcement learning.<\/p>\n\n

Control and adaptability of the ExBody2 robot<\/h2>\n\n

The robot uses a decoupled control approach that separates point tracking from speed adjustment, ensuring stability in its movements. It also allows it to perform actions such as dancing<\/em> , running<\/em> and crouching<\/em> fluidly and without losing balance, even when faced with dynamic situations. <\/p>\n\n

To train the robot, developers used virtual simulations<\/a><\/strong> and a teacher-student model. This allows it to learn digitally and then transfer its skills to the real world, where it adapts them using advanced calculations and motion sensors. <\/p>\n\n

The ExBody2 robot<\/strong> represents a technical breakthrough for practical applications in areas such as medical rehabilitation, education and entertainment. For example, in physiotherapy, the robotic platform could accurately guide patients through personalized exercises, while in the educational sector, it would be a valuable resource for teaching anatomy or body mechanics. <\/p>\n\n

\n