Intrinsic, the industrial robotics software company owned by Google, is open-sourcing foundational components of its robotics platform under the name Intrinsic Core. The release covers ROS-compatible software for robot control, motion and grasp planning, machine vision, simulation and hardware integration, distributed under an Apache 2.0 license.
The centerpiece is Intrinsic Control, a hardware-agnostic real-time control framework that adjusts a robot's behavior mid-trajectory based on sensor feedback. Developers can swap arms, grippers and sensors without rewriting hardware drivers, according to Intrinsic. The platform also includes automated motion planning for collision-free paths, grasp planning that adapts to a part's position and orientation, and pose-estimation technology built on NVIDIA FoundationPose. That last piece lets a robot locate and manipulate 3D parts dynamically instead of relying on rigid fixtures. For manufacturers building automated inspection or material-handling cells, fixture-free pose estimation could trim tooling costs and changeover time on lines that run mixed part geometries.
Intrinsic Core also bundles simulation services based on Gazebo, automated camera calibration and preconfigured ROS-compatible drivers for supported robots, grippers and 3D cameras. Intrinsic says the goal is to give developers reusable building blocks instead of requiring them to build basic robot infrastructure from scratch for every application, shortening the path from prototype to production.
The company already tested pieces of the platform in its AI for Industry Challenge, where 5,000 developers worked on automating the insertion of flexible cables into ports, a task Intrinsic notes is harder to automate reliably than handling rigid components because cables bend and move unpredictably. Two teams reached a 100 percent success rate in final testing on a physical robot, using several of the same software building blocks now included in Intrinsic Core.
Alongside the open-source release, Intrinsic is publishing an Open Machine Tending Solution, a reference application for AI-enabled CNC machine tending that can be customized for equipment including FANUC and Universal Robots systems. Intrinsic Core is available now through the company's GitHub repository.
