Sanctuary AI Leverages NVIDIA Isaac Lab to Accelerate Dexterous Learning
Sanctuary AI Leverages NVIDIA Isaac Lab to Accelerate Dexterous Learning
Sanctuary AI has demonstrated industry-leading sim-to-real transfer of dexterous manipulation policies using their high-degree-of-freedom hydraulic hands. The company leverages NVIDIA Isaac Lab, an open-source framework for robot learning, to simulate dexterity-focused training environments. This approach enables efficient training of autonomous robots, with the ability to train thousands of hands simultaneously and accelerate the learning process through online reinforcement learning.
Sanctuary AI's research highlights the use of NVIDIA Isaac Lab for sim-to-real transfer in dexterous manipulation, emphasizing the advantages of high-degree-of-freedom hydraulic hands and the role of simulation in accelerating reinforcement learning.
Source: Sanctuary AI — Research and announcements · Read original article ↗
Sanctuary AI leverages NVIDIA Isaac Lab –an open-source, unified framework for robot learning that enables training robot policies using high-fidelity simulation–to simulate dexterity-focused training environments. Built on NVIDIA Isaac Sim, Isaac Lab leverages PhysX for physics simulation and RTX rendering to bridge the gap between simulation and perception-based robot training, helping researchers and developers build autonomous robots more efficiently.
No media selected
Sanctuary AI uses Isaac Lab to train our highly capable hydraulic hands to perform cutting edge manipulation tasks, first in simulation, and ultimately in reality. Sanctuary AI’s hands have kinematics beyond human capability, which cannot be accessed using analogous teleoperation. Online reinforcement learning in a simulated environment allows the learning algorithms to fully leverage the hands’ capabilities. Additionally, we can use simulation to train thousands of hands at once, drastically accelerating the learning process.
Our proprietary robotic hand is differentiated by a high number of active degrees of freedom, allowing for finger abduction and therefore in-hand manipulation, a feature not offered by other commercially available hands. We are unique in our use of hydraulic actuation, providing high strength, speed and controllability. Our miniaturized hydraulic valves provide a viable path to achieving human-level dexterity.
About Sanctuary AI
Founded in 2018, Sanctuary AI is an award-winning general purpose robotics and physical AI company based in Vancouver, Canada, widely known for “Phoenix,” its general purpose robot unveiled in 2022. Sanctuary AI is focused on creating dexterous intelligence for autonomous labor to help address the labor challenges facing many organizations today. The commanding IP portfolio of Sanctuary AI–ranked third globally–along with their proprietary dexterous hand technology, uniquely positions the company as a leader in the field.
Sanctuary AI's growing list of customers and investors represents a wide variety of industries across Canada, the US, Japan and other countries around the world.
The post Sanctuary AI Leverages NVIDIA Isaac Lab to Accelerate Dexterous Learning appeared first on Sanctuary AI.
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S3
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
n NVIDIA Isaac Sim , Isaac Lab leverages PhysX for physics simulation and RTX rendering to bridge the gap between simulation and perception-based robot training, helping researchers and developers build autonomous robots more efficiently. No media selected Sanctuary AI uses Isaac Lab to train our highly capable hydraulic hands to perform cutting edge manipulation tasks, first in simulation, and ultimately in reality.
Open source S3
Source：Sanctuary AI — Research and announcements · sanctuary.ai