Robotics and physical AI engineering, wherever you are on the journey.
Noard AI helps teams build robots that work in the real world. From a first prototype to a fleet in the field, I bring hands-on ROS 2, perception and systems engineering, together with the advice and training to make it stick.
Get in touchServices
Support at any stage and in any part of your robotics journey.
- 01
Advice & architecture
Technology choices, system architecture reviews and technical roadmaps. A second opinion before you commit, or a plan when you're not sure where to start.
- 02
Implementation
Hands-on code where you need it: ROS 2 nodes and drivers, perception pipelines, sensor integration, and performant C++ and Python.
- 03
Stack improvements
Finding and fixing what holds your robot back: latency, sensor timing, middleware configuration, zero-copy messaging, builds, CI and reproducible environments.
- 04
Training & workshops
ROS 2 training tailored to your team and your codebase, from the fundamentals to performance and good engineering practice.
- 05
Team growth
Help with hiring robotics engineers, setting up development practices, and mentoring developers who are new to robotics.
- 06
Field testing
On-site integration and testing, test plans, logging and data analysis. Turning what goes wrong in the field into fixes that last.
Approach
Start-ups heading from prototype to product, scale-ups hitting the limits of their first stack, and established companies taking their first steps into robotics each need something different. Engagements are shaped to fit.
- Quick scan
- A short, fixed-scope review of your stack or plans, with a written report and concrete recommendations.
- Project
- A well-defined deliverable, such as a driver, a perception pipeline or a migration to a new ROS 2 release.
- Embedded
- Joining your team for a period, part-time or full-time, to build alongside your engineers.
- Ongoing support
- A standing arrangement for advice, design and code reviews whenever you need them.
About
I’m Sander van Dijk, the engineer behind Noard AI. I have been building robots and AI systems for over 20 years, from hands-on individual projects to leading multiple teams and departments, in sectors including automotive and industrial automation. I know both the code and what it takes to ship it with a team.
I hold an MSc in Artificial Intelligence from the University of Groningen and a PhD in Computer Science from the University of Hertfordshire, where I led the humanoid football robot team Bold Hearts in RoboCup. I maintain the widely used ROS 2 V4L2 camera driver, and write about ROS 2, perception and robot software engineering at sgvd.ai.
Noard is Frisian for north: the aim is to help you find your direction and get there.
Writing
Recent articles on sgvd.ai.
Accurate Sensor Timing in ROS 2
It is crucial to obtain correct and accurate timing in autonomous robotic systems. Complex perception, localisation and planning algorithms rely on combining, or ‘fusing’, data …
Fast Zero-Copy Messaging with ROS 2 Intra-Process Communication
ROS 2 uses so-called “middleware” to perform communication between nodes. Over the years different middleware vendors have vied to become the Chosen Implementation, importantly by …
ROS 2 with Nix
Recently I started getting into the world of Nix. First, to manage my user environment using Home Manager, and now to manage my ROS 2 development environment. On any Linux …