NVIDIA推出全新的服務、模式及運算平台，加速推動人形機器人發展

技術的快速發展，正在讓人形機器人從形態的「擬人化」逐步轉向功能的實用性，以此來加速產品商業化和市場擴展。 NVIDIA一直在積極研究和開發人工智慧和機器人技術，尤其是在人形機器人領域，NVIDIA致力於透過 NVIDIA Isaac 、NVIDIA Jetson等平台加速全球人形機器人的發展和創新。為了加速全球人形機器人的發展，在SIGGRAPH 2024上，NVIDIA正式宣佈為全球機器人製造商、AI模型開發者和軟體製造商提供一套服務、模型以及運算平台，以開發、訓練和建構下一代人​​形機器人。

The complete suite of products includes new NVIDIA NIM™ microservices and frameworks for robotics simulation and learning, NVIDIA OSMO orchestration services for running multi-stage robotics workloads, and AI and simulation-enabled remote operations workflows that allow development Researchers use small amounts of human demonstration data to train robots.
Simplify training and simulation workflow, shorten deployment and development cycle
NVIDIA NIM (NVIDIA Inference Manager) is a tool provided by NVIDIA for managing and optimizing the inference process of AI models. NIM is designed to help developers and enterprises deploy, expand and maintain their AI inference applications more efficiently.
To help developers accelerate product development, NVIDIA provides pre-built containers powered by NVIDIA inference software in NIM microservices, enabling developers to reduce deployment time from weeks to minutes. Robotics experts will be able to enhance generative physics AI simulation workflows in NVIDIA Isaac Sim™m, a robotics simulation reference application built on the NVIDIA Omniverse™ platform, through two new AI microservices.
In addition, MimicGen NIM can generate synthetic motion data based on remote operation data recorded by spatial computing devices such as Apple Vision Pro. RobocasaNIM generates robotic tasks and simulation-ready environments in OpenUSD, a universal framework for developing and collaborating in 3D worlds.
Now available, NVIDIA OSMO is a cloud-native managed service that allows users to orchestrate and scale complex robotics development workflows across distributed computing resources, whether on-premises or in the cloud.
OSMO greatly simplifies robot training and simulation workflows, shortening deployment and development cycles from months to within a week. Users can visually manage a variety of tasks, including synthetic data generation, model training, reinforcement learning, and large-scale software-in-the-loop testing of humanoid robots, autonomous mobile robots, and industrial manipulators.
It is not difficult to find,
Use real data and synthetic data to save time and reduce costs
Training the basic model of humanoid robots requires a large amount of data. Remote operation is one way to obtain human demonstration data, but the process is becoming increasingly expensive and lengthy.
With the NVIDIA Al and Omniverse remote operation reference workflow demonstrated at the SIGGRAPH computer graphics conference, researchers and AI developers can generate large amounts of synthetic motion and perception data from a very small number of remotely captured human demonstrations.
First, the developer used Apple Vision pro to capture a small number of remote operation demonstrations, then simulated these recordings in NVIDIA lsaac Sim, and used MimicGen NIM to generate a synthetic data set based on the recordings.
Developers can use real data and synthetic data to train the Proiect GROOT humanoid robot basic model to save time and reduce costs. They can then use Robocasa NIM in Isaac Lab, a robot learning framework, to generate experience and retrain the robot model. Throughout the entire workflow, NVIDIAOSMO seamlessly allocates computing tasks to different resources, saving developers weeks of administrative workload.
Fourier, a general robotics platform company, sees the advantages of using simulation technology to comprehensively generate training data.
Fourier Alex Gu said that the development of humanoid robots is extremely complex, and this work requires tedious acquisition of large amounts of real data from the real world. NVIDIA’s new simulation and generative AI developer tools will help guide and accelerate our model development workflow.
Expand developers’ technology access channels and simplify the development of humanoid robots
NVIDIA provides three computing platforms to simplify the development of humanoid robots, namely: NVIDIA AI supercomputer for training models; NVIDIA lsaac Sim built on 0mniverse to enable robots Skills can be learned and perfected in a simulated world; and the NVIDIA Jetson™ Thor humanoid robotics computer used to run models. Developers can access and use the entire platform or any part of it based on their specific needs.
Through the new NVIDIA Humanoid Robot Developer Program, developers can gain early access to these new products as well as new versions of NVIDIA lsaac Sim, NVIDIA lsaac Lab, Jetson Thor and Project GR00T universal humanoid robot base models.
The first companies to join the early access program are 1x, Boston Dynamics, ByteDance Research, FieldAl, Figure, Fourier, Galaxy General, Zhuji Dynamics, Mentee, Neura Robotics, Star Dynamics and Skild Al.
Boston Dynamics’ Aaron Saunders said: “Boston Dynamics and NVIDIA have a long history of close collaboration in advancing robotics technology. We are very excited to see that the results of this work are accelerating 33 developments across the industry.
According to reports, currently Developers can now join the NVIDIA Humanoid Robot Developer Program to access NVIDIA OSMO and lsaac Lab, and will soon be able to access NVIDIA NIM microservices
Written at the end:
The development of humanoid robots is constantly improving, as technology improves. , their application in the above fields will become more extensive and in-depth. With the further development of artificial intelligence, machine learning and robotics, the capabilities and application scope of humanoid robots will continue to expand.
As NVIDIA Huang Renxun said, the next step of AI. One wave is robotics, and one of the most exciting developments is humanoid robots.NVIDIA is advancing the entire NVIDIA Robotics stack, opening access to humanoid robot developers and companies around the world, giving them access to the platforms, acceleration libraries and AI models that fit their needs.

以上是NVIDIA推出全新的服務、模式及運算平台，加速推動人形機器人發展的詳細內容。更多資訊請關注PHP中文網其他相關文章！