Intermediate track
AI for Robotics
See how AI gives robots eyes and judgement: computer vision for perception, and mapping and path planning for navigation.
Who this track is for
This intermediate track is for learners who already know the basic parts of a robot and the idea of a trained model, and now want to see how the two meet. It suits students working on a final-year or hobby robotics project, developers moving from pure software into robotics, and anyone preparing to try ROS or a simulator.
What you should know first
We recommend finishing Robotics Fundamentals and AI Fundamentals first. You should understand what sensors and actuators do, what a feedback loop is, and roughly how a neural network is trained. Some comfort reading simple diagrams and tables is useful, but no advanced mathematics is assumed.
What you will learn
- Describe how a robot turns camera images into useful information
- Explain the difference between localisation, mapping and planning
- Compare common path planning approaches such as A* and sampling-based planners
Lessons
- 1Computer Vision for Robots: From Pixels to UnderstandingHow robots use cameras and AI to detect objects, estimate depth and understand scenes, including classification, detection, segmentation and common pitfalls.8 min read · Intermediate
- 2Path Planning and SLAM: How Robots NavigateUnderstand how mobile robots build maps, work out where they are (SLAM), and plan collision-free routes using algorithms like A*, Dijkstra and RRT.9 min read · Intermediate
How to study this track
- Read the computer vision lesson and list the vision task, from classification to depth estimation, that suits a robot you would like to build.
- Note how each perception output, such as a bounding box or a depth value, becomes an input for the robot navigation system rather than an end in itself.
- Read the path planning and SLAM lesson and sketch a small occupancy grid of your room, marking free, occupied and unknown cells, then trace an A* style route by hand.
- Try a simulator such as Gazebo with ROS if you can. Seeing a robot localise and plan on screen makes the ideas concrete without any hardware risk.
- Keep a short notebook of what failed and why. In robotics, understanding that lighting ruined a detection or that wheel slip corrupted odometry teaches more than a run where everything worked.
Frequently asked questions
What will I learn in AI for Robotics?
You will learn how robots use computer vision to understand images, and how localisation, mapping, SLAM and path planning let them navigate.
What should I know before this track?
We recommend finishing Robotics Fundamentals and AI Fundamentals first, so sensors, feedback loops and trained models are already familiar.
Can I practise these topics without a real robot?
Yes. Simulators such as Gazebo, used with ROS, let you test navigation and perception ideas on a computer without hardware.
Explore other tracks
