As robotics developers from around the world gathered this week in Singapore for ROSCon 2025, our Singapore-based and global Intrinsic team joined the open source robotics community to support the infrastructure and people who continue to make the robotics ecosystem more open and collaborative. In addition to creating a more seamless experience between ROS, Gazebo, and Open-RMF, at ROSCon we announced better integration between Intrinsic Flowstate and ROS-based applications through a native Flowstate-ROS Bridge.

ROS (the Robot Operating System) remains a foundational framework for robotics from personal projects and proofs of concept, to deep-sea submersibles and commercial applications. We see our investments in open source tools and deepening ROS integrations with the Intrinsic platform as essential to the growth of the robotics ecosystem. 

‍

At ROSCon, we were excited to share recent integrations we've made with the Intrinsic platform, with the goal of making the best of ROS and Intrinsic easier to build and deploy with.

‍

• The Intrinsic platform now supports camera integration via ROS: Intrinsic Flowstate, our digital-twin developer environment, can now support any camera with ROS drivers. This integration allows for standardized data acquisition and interaction, which enables features like viewing camera feeds from the front end and using ROS-integrated cameras for pose estimation training. Both OEMs (Original Equipment Manufacturers) and the broader community provide ROS drivers for cameras, and this functionality is currently facilitated by APIs in our ROS SDK, which are available to Flowstate users.

‍

• Our ROS driver integration also includes gripper support: We have now added ad-hoc support for ROS-based grippers - for instance, this solution includes support for a Schunk pinch gripper driver in Flowstate which we demonstrated in our booth at ROSCon. With ROS driver gripper support, you can now integrate grippers via ROS — bringing the same functionality and customizability you're used to in ROS, within the Intrinsic platform.

‍

• A streamlined Flowstate-ROS bridge enhances data sharing and request handling: An “executive bridge” enables new functionality that the previous one-directional ROS gateway did not have, allowing developers to send requests from ROS to Flowstate for the first time. This makes it possible to control the execution of a Flowstate solution over  ROS and enables such solutions  to be integrated into existing ROS-based automations. Available through the Flowstate service catalog and with the ability to extend functionality using APIs in our ROS SDK, this integration is an improved, rearchitected, and bidirectional version of the ROS gateway. A “world bridge” replicates the functionality of the original ROS gateway by publishing dynamic update information about the digital twin from Flowstate to ROS — with data showing up in RViz (a three-dimensional robotic visualization program using ROS). The Flowstate-ROS Bridge is fully open-source, allowing anyone to easily extend it and bridge any other Flowstate service into the ROS ecosystem.

‍

In addition to our contributions to the broader open source robotics ecosystem through ROS, Gazebo, Open-RMF, and Infrastructure at ROSCon we announced the “AI for Industry” developer challenge to support the next-generation developer ecosystem for intelligent robotics and highlighted two new community-driven initiatives:

‍

• Proposing and leading a new Special Interest Group for Physical AI: Under the OSRA Technical Governance Committee, and alongside other members including NVIDIA, this group is working to ensure that the ROS community is able to connect with new innovations around bringing AI to physical robotics. While these new developments will be powerful for ROS users in the future, the standards and abstractions already developed in ROS will help guide the introduction of physical AI.

‍

• Joining the Alliance for OpenUSD: As we explore how to better connect the OpenUSD standard with the ROS ecosystem and Intrinsic platform, we see opportunities to collaborate further through this mechanism, in concert with partners like NVIDIA. Keep an eye out for more updates, as we look forward to sharing more.

‍

If you’re excited to accelerate the introduction of AI to industry and solve some of robotics biggest problems, please sign up to participate in the challenge. You can register as an individual or as part of a team – with a USD $180,000 prize pool for the top 5 teams. And if you were also at ROSCon this year, we hope you were able to stop by our booth and see Intrinsic Intelligence in action. If you have ideas or wish to get involved in contributing to open source robotics, we’d love to work together!