NVIDIA speeds up open-source robotics with Isaac ROS 5.0

NVIDIA is extending the power of AI agents to robotics development with Isaac ROS 5.0, a release that brings GPU-accelerated computing and physical AI models to the open-source ROS ecosystem and its nearly 1.3 million users. The new platform introduces agentic workflows, capabilities previously used for software tasks, directly into the robotics development process, allowing developers and AI to build robots together. NVIDIA states that the release also adds support for ROS Lyrical and Ubuntu 24.04, providing a clear upgrade path for developers.

Agentic Workflows Expand Robotics Development with Isaac ROS 5.0

The integration of agentic workflows into NVIDIA Isaac ROS 5.0 is accelerating object pose estimation and tracking, now delivering performance up to 5.5 times faster thanks to the FoundationPose model’s agent-ready inference library. This advancement allows robots to perceive and track the position and orientation of objects with significantly improved speed, a critical capability for dynamic environments and complex manipulation tasks.

The MenteeBot humanoid robot uses Isaac ROS to scale its perception capabilities across Jetson hardware platforms
The MenteeBot humanoid robot uses Isaac ROS to scale its perception capabilities across Jetson hardware platforms. — Source: blogs.nvidia.com

Prefix.dev’s Pixi package-management tool further enhances this flexibility, simplifying the creation of reproducible robot development environments by bringing together ROS with the NVIDIA CUDA platform. This integration allows developers to more easily set up and share accelerated robotics workflows, fostering collaboration and accelerating innovation within the open-source community.

The introduction of NVIDIA Isaac skills for setup and manipulation represents a shift towards AI-assisted robotics development, providing reusable workflows that both developers and AI agents can utilize, the company says. These skills automate repetitive tasks, allowing developers to focus on higher-level problem-solving and customization. Agent-ready documentation further facilitates this collaboration, enabling AI agents to understand Isaac ROS tools and workflows, ultimately translating developer intent into functional applications more efficiently.

NVIDIA states that AI agents are changing how software is built, highlighting the potential for automation and accelerated development cycles. Magna is already using this combined approach, pairing Isaac ROS with Isaac Sim for hardware-in-the-loop testing, a method designed to accelerate deployment and reduce the time required to validate robotic systems in realistic scenarios. This testing process allows for early identification and correction of potential issues, minimizing risks and ensuring reliable performance in real-world applications.

Ekumen is also demonstrating the benefits of GPU acceleration, utilizing isaac_ros_cumotion to map a collision-free path for a warehouse arm in just 2 to 5 milliseconds, a speed critical for maximizing throughput and efficiency in automated logistics environments. The impact of Isaac ROS 5.0 extends beyond individual components, fostering a broader ecosystem of integration and collaboration.

Ouster’s integration of Stereolabs ZED stereo cameras with NVIDIA Isaac ROS delivers GPU-accelerated perception for robotics applications, simplifying the development of real-time object detection, mapping, and navigation, according to the company. This interoperability with the wider ROS ecosystem ensures that developers can use existing tools and libraries while benefiting from NVIDIA’s accelerated computing capabilities.

NVIDIA’s commitment to bringing the complete physical AI stack to the robot is exemplified by its Jetson platform, a scalable computing solution designed to run AI models and applications at the edge with real-time performance. Isaac ROS 5.0 supports a range of Jetson devices, from the entry-level Orin Nano to the high-performance Thor, providing developers with a clear path from development to deployment as robotics workloads become increasingly sophisticated.

Seeed Studio is using NVIDIA Isaac ROS with reBot Arm, combining accelerated perception, spatial understanding and motion planning on NVIDIA Jetson Thor. This integration gives developers a practical platform for building adaptable physical AI applications, from object localization to collision-aware manipulation and autonomous pick and place.

The release of Isaac ROS 5.0, available now as free and open-source software, signifies a continued investment in the ROS community and a commitment to accelerating innovation in the field of robotics. Developers can access resources and begin working with NVIDIA Isaac ROS on GitHub, further solidifying the platform’s position as a central hub for open-source robotics development.

ROS Lyrical & CUDA Integration Streamlines Open Source Tools

The integration of ROS Lyrical and NVIDIA’s CUDA platform within Isaac ROS 5.0 establishes a standardized data interface that facilitates efficient operation of robotics software across diverse computing hardware, including GPUs. This advancement addresses a critical need for interoperability as developers increasingly use heterogeneous computing resources to accelerate demanding robotics workloads. NVIDIA worked with the Open Source Robotics Alliance to contribute this interface, underscoring a commitment to fostering a more cohesive and adaptable open-source ecosystem, the firm reports.

The resulting consistency in acceleration methods, exemplified by CUDA, provides a practical framework for developers seeking to optimize performance. Isaac ROS 5.0’s support for Ubuntu 24.04 further streamlines the development process by ensuring compatibility with a widely adopted operating system. This compatibility is not merely a matter of convenience; it provides developers with a clear upgrade path, allowing them to benefit from the latest software improvements and security features without disrupting existing workflows.

RealSense’s optimization of its AI-native 3D stereo depth cameras, including the D585 Pro, alongside an open-source software development kit for Isaac ROS and the NVIDIA Jetson Thor edge AI platform, is expanding the possibilities for perception, navigation, and manipulation applications. This collaboration demonstrates a broader trend within the robotics community towards using specialized hardware and software components to achieve optimal performance. The integration of RealSense’s cameras with Isaac ROS allows developers to tap into advanced 3D perception capabilities, enabling robots to better understand and interact with their environment, NVIDIA reports.

Intrinsic’s Open Machine Tending Solution, built on the newly released Intrinsic Core, exemplifies how Isaac ROS 5.0 is accelerating industrial robotics applications. The solution’s compatibility with NVIDIA FoundationPose for object registration, tracking, and pose estimation simplifies the process of automating complex manufacturing tasks.

By reducing the need for rigid physical fixtures and specialized systems integration, Intrinsic’s solution lowers the barriers to entry for deploying robots in industrial settings. The use of FoundationPose enables robots to dynamically detect and handle parts, increasing flexibility and reducing downtime. This is important for collaborative projects and for ensuring that applications can be consistently deployed across different hardware configurations. The ability to create reproducible environments reduces the risk of errors and inconsistencies, leading to more reliable and robust robotic systems.

Foxglove’s integration with Isaac ROS, as an Isaac ROS Partner, provides developers with powerful visualization and debugging tools. These tools, accessible through both web and desktop interfaces, are integrated throughout Isaac ROS tutorials and support critical data types such as 3D topics, nvblox meshes, and rosbags. This seamless integration allows developers to gain deeper insights into the behavior of their robotic applications, facilitating faster iteration and problem-solving.

By building on ready-to-use AI and perception capabilities, Noble Machines is able to focus on innovation rather than reinventing the wheel, the company’s account states. This approach not only speeds up development but also reduces costs and risks. The combination of Isaac ROS, Jetson, and pre-built AI capabilities is empowering companies like Noble Machines to create increasingly capable and versatile robots.

Stay current

See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.

Avatar of Rusty Flint

Rusty Flint

Rusty is a quantum science nerd. He's been into academic science all his life, but spent his formative years doing less academic things. Now he turns his attention to write about his passion, the quantum realm. He loves all things Quantum Physics especially. Rusty likes the more esoteric side of Quantum Computing and the Quantum world. Everything from Quantum Entanglement to Quantum Physics. Rusty thinks that we are in the 1950s quantum equivalent of the classical computing world. While other quantum journalists focus on IBM's latest chip or which startup just raised $50 million, Rusty's over here writing 3,000-word deep dives on whether quantum entanglement might explain why you sometimes think about someone right before they text you. (Spoiler: it doesn't, but the exploration is fascinating)

Latest Posts by Rusty Flint: