Good news everyone: Intrinsic Core is now Open Source!
Open Source has played a huge role in my career. It is hard to overstate how impactful it was at each stage of the way. I first became involved with Open Source projects early in my teenage years, drawn by the open collaboration and joint innovation - and I’ve contributed ever since.
My start was in operating systems and general software. Soon after recognizing robotics as my professional calling, I dove into the Open Robotics ROS ecosystem, starting with ROS Diamondback, and stayed close to it over the past 15+ years. So as I gather with the global community of robotics developers at #ROSCon2026 in Toronto, I couldn’t be more excited that we’re open sourcing core parts of the Intrinsic platform to unlock AI for developers building (industrial) robotics applications.
The release of Intrinsic Core makes it easier and faster to build and run industrial robotics applications: from a state-of-the-art hardware-agnostic real-time control framework to a digital twin that connects simulation with an online world model to ready-to-use AI perception, capabilities, such as 6-DoF pose estimation, hardened online collision-free motion planning, automated camera calibration, grasp planning, pre-configured ROS drivers, and Open Machine Tending Solution (OMTS) as an open reference application. I’m confident this open source release will accelerate moving robots out of the lab and into real world use.
You can learn more about this set of ROS-compatible capabilities on the Intrinsic blog: https://lnkd.in/gNzGK997
and you can get started on GitHub today: https://lnkd.in/gM5c8UU9
Amongst the - subjectively - most impactful parts of Intrinsic Core for developers, I’d like to highlight:
• Intrinsic Control: a high performance, hardware-agnostic, real-time control framework for industrial robots, including sensor-based control
• World Model and Simulation Services: a unified spatial and semantic representation of the environment - tracking initial, belief, and simulated states - facilitating seamless digital-twin synchronization, in simulation and on hardware
• ML Inference: a set of backend-agnostic, reusable building blocks for managing ML inference lifecycle, model asset synchronization, state reconciliation - with native ROS support
• ROS Assets: workflows to containerize ROS nodes and make them seamlessly interoperable with Intrinsic Core
Intrinsic Core also includes an open reference design giving developers an end-to-end starting point for real-world machine tending automation, which makes it even easier to begin today.
While we’re still in an early stage with Intrinsic Core, I can’t wait to support its development with the broader community, keep improving ROS interoperability, and see the impact it’ll have for developers and industry alike!
You can join the Intrinsic developer community at https://lnkd.in/gjSE8J9a
