Intrinsic opens up its physical AI core

Recap

Source: Robohub, report of Sept. 25, 2026

According to the report, Intrinsic, an AI robotics group at Google, has announced that it is making parts of its platform open source. The company says this new release, Intrinsic Core, is a set of ROS-compatible capabilities designed for building sophisticated robotic applications.

The open-source package includes several key modules. Intrinsic Control is described as a hardware-agnostic, real-time control framework that delivers sensor-based control, allowing robots to dynamically adapt their behavior mid-trajectory based on environmental changes. It is designed to deliver efficient, collision-free paths and allows users to swap out different robot arms, grippers, and sensors without needing to rewrite drivers.

Other capabilities in the package include pose estimation, which provides accurate six degrees of freedom pose estimation of 3D parts, and motion planning, which auto-generates efficient paths while avoiding collisions. The system also features grasp planning, designed to enable various grippers to automatically and dynamically adapt to the location and orientation of a part to increase the probability of a successful grasp. Additionally, the package includes simulation services for testing and troubleshooting, camera calibration tools for automated alignment, and pre-configured, ROS-compatible drivers for supported robots, grippers, and 3D cameras.

Alongside this package, Intrinsic has also released the Open Machine Tending Solution (OMTS). This is a ready-made reference design for a real-world use case, intended to give developers a starting point to customize their own solution without needing to build from scratch. The company says Intrinsic Core can be downloaded on GitHub.

Context

Intrinsic has long been known for its work on physical AI and the integration of machine learning into robotic control. By releasing Intrinsic Core, the company is moving away from a purely proprietary model toward a more open ecosystem. This approach aligns with broader industry trends where major tech companies are increasingly contributing to open-source robotics frameworks to accelerate development and standardize best practices. The inclusion of ROS (Robot Operating System) compatibility is a significant step, as ROS is the de facto standard for many robotics research and development projects.

The release of OMTS as a reference design is particularly noteworthy. It provides a practical, real-world application example that developers can use as a foundation for their own projects. This can significantly lower the barrier to entry for companies and researchers looking to implement advanced robotic solutions, as they can start from a proven design rather than building everything from the ground up.

Robot's take

The open-sourcing of Intrinsic Core is a significant move that could have a substantial impact on the robotics community. By providing a robust, real-time control framework and other advanced capabilities, Intrinsic is making it easier for developers to build sophisticated robotic applications. The hardware-agnostic nature of the control framework is particularly appealing, as it allows for greater flexibility and interoperability between different robotic components.

However, the success of this initiative will depend on how well the open-source community adopts and contributes to the project. The availability of a reference design like OMTS is a strong positive, as it provides a practical starting point for developers. It remains to be seen how widely this toolkit will be used and whether it will become a standard in the industry. The next few months will be crucial in determining the impact of this release on the robotics landscape.