WHY THIS MATTERS
CONFIRMED: Intrinsic, Alphabet's robotics unit, released Intrinsic Core on September 22 under an Apache 2.0 license, with real time control, part perception, motion and grasp planning, the same building blocks the company uses in real factories.In this article
CONFIRMED: Google opens up the engine behind its industrial robotics
Intrinsic, Alphabet's robotics company, unveiled Intrinsic Core on September 22 during ROSCon 2026 in Toronto. The package bundles ROS compatible capabilities under a permissive Apache 2.0 license and is publicly available on GitHub, with a pre configured environment that runs on local hardware.
What is inside the box
Intrinsic Core includes a real time control framework that works across hardware vendors and adjusts a robot's path mid move using sensor feedback. It also brings pose estimation built on Nvidia's FoundationPose, letting a robot locate a part without relying on rigid fixtures, plus collision free motion planning, grasp planning that adapts to how an object sits, and simulation, calibration and ROS driver services.
Not a demo, real production code
What sets this release apart from another lab robotics kit is where the code comes from: these are the same capabilities and services Intrinsic already runs in real manufacturing deployments. Instead of forcing every team to write control and infrastructure from scratch, the company is packaging ready made blocks of robot behavior to be combined.
A ready recipe for repetitive tasks
Alongside Intrinsic Core, the company released the Open Machine Tending Solution, an open reference design for CNC machine tending. It runs on Intrinsic Core and the Open Robotics Suite and can be adapted to hardware from manufacturers such as Fanuc and Universal Robots, two familiar names on the factory floor.
Why this matters beyond the production line
For a small or midsize manufacturer in Brazil, the immediate value is probably not building a robot from scratch, but lowering the entry barrier for local integrators and suppliers who already want to automate repetitive tasks, such as loading a machine or sorting parts. Open source code and a permissive license cut licensing cost, but they do not remove the need for hardware, integration work and skilled labor to get a robot running safely.
MaxAssistant's read
Open sourcing the core that runs your own production robots is a platform bet, not charity: the more integrators build on Intrinsic Core, the more the ecosystem orbits Google's standard. Worth watching whether rival robot makers adopt the package or treat the move as a competitor's push to become the default software layer for industrial robotics.
Sources
Intrinsic, Introducing Intrinsic Core, an open source approach to Physical AI: https://www.intrinsic.ai/blog/posts/introducing-intrinsic-core | The Robot Report, Intrinsic open sources key parts of its platform for easier development: https://www.therobotreport.com/intrinsic-open-sources-key-parts-platform-easier-development/ | SiliconANGLE, Google's robotics unit Intrinsic open-sources its foundational infrastructure for intelligent robots: https://siliconangle.com/2026/09/22/googles-robotics-unit-intrinsic-open-sources-its-foundational-infrastructure-for-intelligent-robots/