.Ted Hisokawa.Aug 08, 2024 13:53.Body has unveiled its own next-gen Figure 02 humanoid robot, making use of NVIDIA Omniverse as well as GPUs for boosted artificial intelligence functionalities and also autonomous functions.
Silicon Valley-based Body has revealed its own next-generation Figure 02 humanoid robotic, which leverages NVIDIA Omniverse and NVIDIA GPUs to conduct completely independent jobs. According to the NVIDIA Blog, the Design 02 model has actually been checked for records compilation and also use-case instruction at BMW Team's Spartanburg assembly-line production in South Carolina.Technological Advancements.Amount 02 is actually a notable upgrade from the preliminary version released simply ten months ago. The new model combines a second NVIDIA RTX GPU-based element, supplying three times the inference gains matched up to its own precursor. This improvement facilitates dealing with totally self-governing real-world AI duties. The provider used NVIDIA Isaac Sim, a recommendation treatment improved the NVIDIA Omniverse system, to design, learn, and test AI-based robots utilizing synthetic data.Business Requests.Figure intends to commercialize industrial humanoid robotics to deal with labor lacks all over a variety of markets. The company also intends to produce consumer models. Brett Adcock, CEO of Number, highlighted the fast improvement in speech, concept, dexterity, and also computational electrical power, which delivers all of them closer to providing humanoid robots for several markets.Alliances and also Backing.Founded in 2022, Figure has actually partnered with OpenAI to build personalized artificial intelligence designs qualified on NVIDIA H100 GPUs. These styles improve the robots' informal AI functionalities. The startup just recently protected $675 thousand in financing from leading technology providers, featuring NVIDIA.Deepu Talla, bad habit president of robotics and advantage computing at NVIDIA, highlighted the significance of the NVIDIA robotics stack, from upper hand to cloud, in driving innovation in humanoid robotics. "Creating self-governing humanoid robots demands the combination of three pcs: NVIDIA DGX for AI training, NVIDIA Omniverse for simulation, and NVIDIA Jetson in the robot," Talla said.Robot Hands Efficient In Managing Real-World Tasks.Figure 02 features brand-new human-scale hands, six RGB cameras, and also impression AI designs trained along with man-made data generated in Isaac Sim. These capabilities make it possible for the robot to conduct high-precision pick-and-place jobs necessary for wise manufacturing applications.Figure is amongst the initial members to participate in the new NVIDIA Humanoid Robot Programmer Course, which uses very early accessibility to the current devices and computing technologies for humanoid robotic growth. This features the most up to date releases of NVIDIA Isaac Sim, Isaac Laboratory, NIM microservices (RoboCasa as well as MimicGen), OSMO, Jetson Thor, and also Project GR00T general-purpose humanoid structure models.Image source: Shutterstock.