NVIDIA Accelerates Humanoid Robotics Development

Developers Gain Access to New NVIDIA NIM Microservices for Robotics Simulation in Isaac Lab and Isaac Sim, OSMO Robot Cloud Compute Orchestration Service, Teleoperated Data Capture Worklow and More

To accelerate humanoid development on a global scale, on Aug 1, NVIDIA announced it is providing the world’s leading robot manufacturers, AI model developers and software makers with a suite of services,  models and computing platforms to develop, train and build the next generation of humanoid robotics.

Among the o仟erings are new NVIDIA NIM™ microservices and frameworks for robot simulation and learning, the NVIDIA OSMO orchestration service for running

multi-stage robotics workloads, and an AI- and simulation-enabled teleoperation worklow that allows developers to train robots using small amounts of human   demonstration data.

“The next wave of AI is robotics and one of the most exciting developments is humanoid robots,” said Jensen Huang, founder and CEO of NVIDIA. “We’re advancing the entire NVIDIA robotics stack, opening access for worldwide humanoid developers and companies to use the platforms, acceleration libraries and AI models best suited for their needs.”

Accelerating Development With NVIDIA NIM and OSMO

NIM microservices provide pre-built containers, powered by NVIDIA inference software, that enable developers to reduce deployment times from weeks to minutes. Two new AI microservices will allow robotic ists to enhance simulation worklows for generative physical AI in NVIDIA Isaac Sim™, a reference application for robotics simulation built on the NVIDIA Omniverse™ platform.

The MimicGen NIM microservice generates synthetic motion data based on recorded teleoperated data from spatial computing devices like Apple Vision Pro. The Robocasa NIM microservice generates robot tasks and simulation-ready environments in OpenUSD, a universal framework for developing and collaborating within 3D worlds. NVIDIA OSMO, available now, is a cloud-native managed service that allows users to orchestrate and scale complex robotics development worklows across distributed  computing resources, whether on premises or in the cloud.

OSMO vastly simplifies robot training and simulation worklows, cutting deployment and development cycle times from months to under a week. Users can visualize and

manage a range of tasks — like generating synthetic data, training models, conducting reinforcement learning and implementing software-in-the loop testing at scale for humanoids, autonomous mobile robots and industrial manipulators.

Advancing Data Capture Workiows for Humanoid Robot Developers

Training foundation models for humanoid robots requires an incredible amount of data. One way of capturing human demonstration data is using teleoperation, but this is becoming an increasingly expensive and lengthy process. An NVIDIA AI- and Omniverse-enabled teleoperation reference worklow, demonstrated at the SIGGRAPH computer graphics conference, allows researchers and AI developers to generate massive amounts of synthetic motion and perception data from a minimal amount of remotely captured human demonstrations. First, developers use Apple Vision Pro to capture a small number of teleoperated demonstrations. Then, they simulate the recordings in NVIDIA Isaac Sim and use MimicGen NIM microservice to generate synthetic datasets from the recordings.

The developers train the Project GR00T humanoid foundation model with real and synthetic data, enabling developers to save time and reduce costs. They then use

Robocasa NIM microservice in Isaac Lab, a framework for robot learning, to generate experiences to retrain the robot model. Throughout the worklow, NVIDIA OSMO seamlessly assigns computing jobs to di仟erent resources, saving the developers weeks of administrative tasks. Fourier, a general-purpose robot platform company, sees the benefit of using simulation technology to synthetically generate training data.

“ Developing humanoid robots is extremely complex — requiring an incredible amount of real data, tediously captured from the real world,” said Alex Gu, CEO of Fourier. “ NVIDIA’s new simulation and generative AI developer tools will help bootstrap and accelerate our model development worklows.”

Expanding Access to NVIDIA Humanoid Developer Technologies

NVIDIA provides three computing platforms to ease humanoid robotics development: NVIDIA AI supercomputers to train the models; NVIDIA Isaac Sim built on Omniverse, where robots can learn and refi ne their skills in simulated worlds; and NVIDIA Jetson™ Thor humanoid robot computers to run the models. Developers can access and use all — or any part of — the platforms for their specific needs.

Through a new NVIDIA Humanoid Robot Developer Program, developers can gain early access to the new o仟erings as well as the latest releases of NVIDIA Isaac Sim,

NVIDIA Isaac Lab, Jetson Thor and Project GR00T general-purpose humanoid foundation models.1x, Boston Dynamics, ByteDance Research, Field AI, Figure, Fourier, Galbot, LimX Dynamics, Mentee, Neura Robotics, Robot Era and S kild A I are among the fi rst to join the early-access program.

“ Boston Dynamics and NVIDIA have a long history of close collaboration to push the boundaries of what’s possible in robotics,” said Aaron Saunders, chief technology o代cer of Boston Dynamics. “We’re really excited to see the fruits of this work accelerating the industry at large, and the early-access program is a fantastic way to access best-in-class technology.”

Availability

Developers can join the NVIDIA Humanoid Robot Developer Program now to get access to NVIDIA OSMO and Isaac Lab, and will soon gain access to NVIDIA NIM microservices.

Learn more about the latest in generative AI and accelerated computing by tuning in to Huang’s fireside chats at SIGGRAPH, the premier computer graphics conference,  running through Aug. 1 in Denver.

About NVIDIA

NVIDIA (NASDAQ: NVDA) is the world leader in accelerated computing.

Source:https://nvidianews.nvidia.com/news/nvidia-accelerates-worldwide-human oid-robotics-development