Computer vision
Give your system the ability to see, understand, and act on the world around it.
- Object detection & tracking with YOLO and DETR
- Visual inspection & segmentation with SAM3
- Sensor fusion & robotic perception
Independent robotics & AI studio · UK
Robotics &
embedded
computer vision.
We build robotics, embedded computer vision, and Edge AI systems for your hardware. From perception and sensor fusion to model optimisation and deployment, we connect machine learning to physical systems.
Based in York. Working across the UK & Europe.
BUILT WITH THE
RIGHT TOOLS.
RIGHT TOOLS.
NVIDIA JetsonRaspberry PiPyTorchOpenCVROS 2ONNX
01 / Selected work
Engineering in action.
A closer look at the problems, hardware decisions, and systems behind our work.
VISION / INDUSTRIAL INSPECTION
Computer vision
Every detail matters.
Automated quality inspection for manufacturing, from surface-defect detection to production-line integration.
Explore quality control
ROBOTICS / AUTONOMOUS NAVIGATION
Robotics & perception
A way forward.
Autonomous mobile robot navigation that combines computer vision and LiDAR in changing warehouse environments.
Explore robot navigation
EDGE AI / CONNECTED BUILDINGS
Sensor fusion
Signals into insight.
Occupancy prediction for connected buildings, bringing environmental sensors and local inference together.
Explore sensor fusion
More from the studio
02 / Our expertise
From possibility to practical.
Specialist support where you need it.
A connected approach from model to machine.
Model optimisation
Optimise Edge AI and Tiny ML models for your compute budget, measuring accuracy, latency, and power together.
- Quantisation & model compression
- Vision-language models (VLM) & vision-language-action models (VLA)
- Hardware-specific benchmarking
- Inference pipeline optimisation
Embedded deployment
Deploy Edge AI on your target hardware, with runtime integration and validation in real operating conditions.
- Jetson Orin, Jetson Nano & Jetson Thor
- Raspberry Pi & ARM platforms
- Runtime & system integration
- Field validation & monitoring
Timing improvement
Find latency bottlenecks and improve timing predictability across your robotics system.
- ROS 2 scheduler optimisation
- Real-time DDS tuning
- Timing debugging & tracing
03 / How we work
Clear steps. Working systems.
Start with the constraints. Test the assumptions. Build towards a useful outcome.
Define the problem.
Align on your use case, data, hardware, and what success needs to look like.
↳ Scope & feasibilityProve the approach.
Build a focused prototype and test it against representative data.
↳ Working model & baselineMeet the hardware.
Measure performance on your target device, including real-world edge cases.
↳ Benchmarks & integrationPut it to work.
Integrate the system, document the handover, and agree ongoing support.
↳ Deployment & handover04 / The studio
Curious minds.
Grounded engineering.
YF Studio is an independent robotics and AI engineering studio based in York, UK. We work across embodied intelligence, computer vision, and embedded systems, connecting machine learning to the hardware that brings it to life.
Our areas of focus include perception for autonomous cars and self-driving systems, autonomous mobile vehicle development, and autonomous ground vehicles.
We care about the practical details: the data you actually have, the device you need to run on, and the conditions your system will face. That’s where useful engineering starts.
York, United KingdomWorking across the UK & Europe
Model to deploymentFocused on embedded intelligence
Working with YF Studio
A few useful answers.
Robotics and edge AI engineering, from the first question to a practical next step.
What does YF Studio help engineering teams build?
We develop computer vision, robotics perception, and edge AI systems. Our work includes automated inspection, mobile robot navigation, fleet coordination, and embedded model deployment. Explore the project examples.
What are Edge AI and Tiny ML?
Edge AI runs machine learning on local devices, close to the cameras or sensors that produce the data. Tiny ML (TinyML) focuses on small models for microcontrollers and other devices with limited memory, compute, and power. We help assess model size, accuracy, latency, and hardware constraints before deployment. Explore our on-device security camera project and connected-building sensor fusion project.
What is embodied intelligence?
Embodied intelligence connects perception and decision-making to action in a physical system. For robotics, that means combining sensors, models, and control software so a machine can respond to its surroundings. See the autonomous navigation project.
Can you work with our existing hardware or AI model?
Our services include model optimisation, timing analysis, and integration with existing products. We work with NVIDIA Jetson, Raspberry Pi, ARM platforms, and industrial PCs. Hardware selection and feasibility depend on your data, accuracy, latency, and power requirements.
Which hardware platforms and AI frameworks do you work with?
Hardware: NVIDIA Jetson (Jetson Nano, Jetson Orin including AGX Orin, and Jetson Thor), Raspberry Pi, ESP32, ARM SoCs, industrial and structured-light cameras, LiDAR, and IMUs. Software: ROS 2, Open-RMF, PyTorch, OpenCV, ONNX, TensorRT, TensorFlow Lite, OpenVINO, and MQTT.
Do you work on ROS 2 timing and real-time performance?
Yes. Timing improvement is one of our four service areas. We profile latency bottlenecks, tune ROS 2 executors and scheduling, configure real-time DDS behaviour, and trace end-to-end timing so a robotics system behaves predictably under load. See the timing improvement service.
Where do you work, and how do we start a project?
The studio is based in York and works with teams across the UK and Europe. Send a project enquiry describing the problem, available data, and any hardware or timeline constraints. We’ll discuss technical fit and next steps.
05 / Start a conversation
What are you
working on?
A new idea, a challenging model, or a system ready for its next step. Tell us where you are and where you want to go.
yfstudio.uk@gmail.comA conversation is a good place to start.
We’ll review your enquiry and get in touch to discuss the technical fit and next steps.