Tuesday 1 April 2025 12:04 GMT

Agility Robotics Expands Relationship With Nvidia


(MENAFN- Robotics & Automation News) Agility Robotics expands relationship with Nvidia

March 28, 2025 by David Edwards

Agility Robotics , creator of the market-leading humanoid robot Digit, is expanding its collaboration with Nvidia, one of the world's leading companies in accelerated computing.

As part of the relationship, Agility Robotics is expanding adoption of Nvidia Isaac Sim and Nvidia Isaac Lab robot simulation and learning frameworks to train and test behaviors on Digit, as well as collaborating to make Digit models available to partners through“Mega”, an Nvidia Omniverse Blueprint.

Digit demonstrated its innovative and autonomous workflows at the recent Nvidia GTC AI Conference.

NVIDIA Isaac Lab is an open-source modular framework for robot learning designed to simplify how robots learn and adapt to new skills in simulation, while Nvidia Isaac Sim is a reference application that enables developers to simulate and test AI-driven robotics solutions.

Isaac Lab and Isaac Sim are both used by robotics developers to train new policies using Digit with reinforcement learning in physically-based, virtual environments.

This has already led to improvements in areas like step-recovery, improving the stability of Digit by helping to manage physical disturbances such as being pushed or bumped by using AI models trained across billions of instances.

Pras Velagapudi, Agility Robotics' chief technology officer, says:“When it comes to hardware for running AI models and the software for creating them, we continue to benefit from our strong collaboration with Nvidia.

“Nvidia's focus on technologies to enable physical AI is helping us advance the capabilities of Digit faster than ever.”

Agility makes use of Nvidia's AI acceleration platform for real-time perception and reinforcement-learned controllers onboard Digit. This platform allows humanoid robots to host robust models that can process sensor data and make decisions in real time – crucial for interactions with humans in dynamic environments.

Rev Lebaredian, vice president of Omniverse and simulation technology at Nvidia, says:“Agility's work with Digit represents a significant breakthrough in humanoid robotics.

“By leveraging simulation and reinforcement learning, they are developing Physical AI capable of understanding and interacting with the real world – an essential step in advancing Digit's intelligence.”

Agility also revealed that it has been provided with early access to Nvidia Cosmos and the Mega Nvidia Omniverse Blueprint.

Nvidia Cosmos is a platform of generative world foundation models, and Mega is a blueprint for simulating, testing, and optimizing physical AI and robot fleets at scale in digital twins. The use of the Mega blueprint is already underway with Agility and automotive manufacturing customer Schaeffler.

Agility has thus far been able to represent assets such as common facility items in digital form. These digital twins of assets, such as bins and shelves, are used by Agility to coordinate customer deployments such as Schaeffler's.

Schaeffler provides digital assets and Nvidia's Omniverse provides a common platform for efficient testing and training, an ideal way to optimize operations with physical AI.

Digit is the most advanced Mobile Manipulation Robot (MMR) on the market and recognized as the only humanoid autonomous robot actually getting paid to work in warehouses and factories today.

Designed to navigate our world, Digit can walk into existing facilities and address the hardest-to-automate workflows. Digit is already deployed at multiple customer sites, solving some of the biggest issues in several different industries.

MENAFN28032025005532012229ID1109368867


Legal Disclaimer:
MENAFN provides the information “as is” without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the provider above.

Search