Alphabet-owned Intrinsic open-sources its robotics platform core under Apache 2.0

Intrinsic Innovation LLC, the intelligent robotics software company owned by Alphabet, said at ROSCon 2026 in Toronto that it is open-sourcing the core components of its industrial robotics platform. The release is called Intrinsic Core and ships under the permissive Apache 2.0 licence, which the company describes as a step towards wider access to intelligent robotics.

Intrinsic was founded in 2021 and folded into Google's operations in February to speed up work on physical AI — systems that use sensors, actuators and control loops to perceive, reason, act and learn in real environments.

What Intrinsic Core includes

Intrinsic Core is a preconfigured, Robot Operating System-compatible software environment designed to run on local hardware. It offers reusable building blocks so developers do not have to rebuild a stack for every new machine. Intrinsic Control is a hardware-agnostic real-time control framework that can adapt a robot's trajectory mid-movement from sensor feedback, so arms and grippers can be swapped without rewriting drivers. Pose Estimation, built on Nvidia's FoundationPose, lets robots detect and handle parts dynamically instead of relying on rigid, expensive fixtures. Motion planning generates collision-free paths automatically, replacing joint-by-joint position programming, and grasp planning lets different grippers adapt to the position and orientation of the objects they pick up. Simulation and calibration services and Intrinsic-ROS drivers for third-party grippers and 3D cameras round out the release.

An open reference design for machine tending

Alongside the framework, Intrinsic published the Open Machine Tending Solution, an open reference design for AI-powered CNC machine tending. It is modular enough to fit hardware from major manufacturers, including Universal Robots A/S and FANUC Corp. The company says the US and Europe are home to tens of thousands of shops fabricating intricate machine parts, yet only 8% use any form of automation, held back by cost and technical barriers. A customisable starting point, Intrinsic argues, lets smaller fabricators automate lower-value tasks first.

Why it matters

Physical AI has been slow to arrive largely because the software underneath it is costly to build and each robot tends to need its own integration work. A shared, open foundation is the bet behind Intrinsic Core: if developers stop rewriting drivers, planning and perception for every project, more of them can spend time on applications. Forbes described the move as Google giving away the "Android of robotics". Intrinsic says it wants a broader robotics ecosystem built on shared open-source tools.