Linux Foundation unveils Newton to boost robot learning

The Linux Foundation has added Newton, an open source physics engine, to its collaborative project portfolio. Newton is designed to accelerate robotics research by providing GPU-accelerated simulations that allow robots to learn and perform complex tasks more efficiently. Developed collaboratively by Disney Research, Google DeepMind, and NVIDIA, the engine offers scalable, high-fidelity simulation tools for researchers and developers working on generalist robotics applications.

The platform addresses the growing need for robust simulation frameworks that can replicate the physical challenges robots encounter in real-world environments. By supporting multiple physics solvers and an extensible architecture, Newton enables accurate modeling of robot behaviors that would otherwise require extensive real-world testing.

High-performance GPU-accelerated simulation

Newton leverages GPU acceleration to deliver fast, reliable physics computations, reducing the time needed to train robotic systems. Built on NVIDIA Warp and OpenUSD, it combines parallel processing power with flexible simulation capabilities. This allows developers to test contact-rich robot behaviors such as walking on uneven surfaces like snow or gravel, and handling delicate objects like cups and fruit, all within a simulated environment before deploying robots in real-world scenarios.

According to Rev Lebaredian, vice president of Omniverse and Simulation Technology at NVIDIA, "Robots, like intelligence, must first be trained in simulation before they can safely operate in the real world. Newton brings together GPU acceleration, differentiable physics and open standards like OpenUSD into an open source platform for robotics. Under the Linux Foundation, we’re enabling the global community to accelerate robotics innovation – making general purpose robots faster to develop, more affordable to build and more reliable to deploy."

Collaborative development and governance

Newton’s contribution to the Linux Foundation ensures that the project remains vendor-neutral and community-driven. Under neutral governance, the platform is designed to attract participation from universities, research institutions, and industry partners. Notable early participants include the Technical University of Munich, Peking University, the robotics company Lightwheel, and simulation engine company Style3D.

Jim Zemlin, executive director of the Linux Foundation, stated, "Bringing Newton to the Linux Foundation marks an important step forward for scaling collaborative robotics simulation that accelerates development, reduces costs and gets us closer to the future of sim-to-real robots. We're pleased to welcome Newton and provide the neutral governance its global community needs to build the future of generalist robotics."

This governance model emphasizes accessibility, long-term viability, and the ability to incorporate contributions from a wide array of developers and researchers worldwide. The open source approach encourages innovation and collaboration, allowing simulation techniques and robotics solutions to evolve more rapidly than they might under proprietary systems.

Practical applications in research and industry

Disney Research has leveraged Newton to simulate complex robotic systems for animatronics and character modeling. Moritz Bächer, director at Disney Research, explained, "At Disney, we don’t leave a stone unturned to physically represent our beloved characters as authentically as possible in our Parks. This has led to unique expertise in simulating highly complex robotic systems. Newton’s architecture has made it possible for us to create Kamino, a simulator that unlocks the potential of reinforcement learning for robotic systems of unmatched complexity."

Beyond entertainment, Newton’s capabilities are applicable to a range of robotics research, including autonomous vehicles, industrial automation, and healthcare robotics. By simulating diverse environments and object interactions, developers can train robots on complex tasks while minimizing risk and cost.

Supporting the global robotics community

A key aspect of Newton is its support for a global community of robotics developers. By providing open access to a high-performance physics engine, the platform encourages shared learning and cross-institution collaboration. Researchers can contribute improvements, share datasets, and adopt standardized practices for sim-to-real robot training.

The extensible nature of Newton allows it to integrate with various robotics frameworks, enabling developers to implement new algorithms and solvers efficiently. GPU-accelerated simulations ensure that even highly detailed physical models can be executed at scale, facilitating experimentation and iterative learning.

The Linux Foundation’s stewardship also includes maintaining comprehensive documentation, ensuring long-term project sustainability, and providing a neutral environment for corporate and academic partners to collaborate. This approach strengthens the reliability and adoption of Newton as a foundational tool in robotics research and application development.

Linux Foundation unveils Newton to boost robot learning

By contributing Newton to the Linux Foundation, Disney Research, Google DeepMind, and NVIDIA have made a strategic investment in open, collaborative robotics development. The platform’s combination of GPU-accelerated performance, support for contact-rich robot behaviors, and open governance encourages innovation across research and industry.

Newton demonstrates the potential of open source simulation to reduce development costs, improve learning accuracy, and accelerate the deployment of general-purpose robots. Its adoption by a global community of researchers and developers is likely to influence future robotics standards and practices, supporting a more collaborative and scalable approach to building intelligent machines.

With Newton under Linux Foundation governance, robotics developers have a neutral, high-performance tool that can evolve with community contributions, fostering progress in areas ranging from animatronics to autonomous systems. The project underscores the value of shared technological infrastructure in advancing both academic research and practical applications in robotics worldwide.