New lessons, new games, new tools for teachers, and news of the robot.
Scatter guesses, move them with the robot, weigh them against a sensor reading, resample. The guide shows the cloud of guesses closing in on the robot, failing with too few particles, and recovering after the robot is picked up and moved.
Read the particle filter guideHow a robot tracks where it is from its own wheel and gyro readings, why that estimate drifts, and how calibration and landmark fixes pull it back. Every demo in the guide is a program you can change and run.
Read the dead reckoning guideState transition diagrams and tables, Mealy and Moore machines, a traffic light, and a robot that searches for a marker, drives up to it and comes home. Change the numbers and see what a missing transition does.
Read the state machine guideThe only difference in the code is a queue or a stack, and the guide shows what that changes: breadth-first search finds the shortest path and depth-first may not. Live robot demos, Python for both, time complexity and when to use each.
Read the BFS and DFS guideA worked example traced in a table, then the same algorithm running live as a robot floods a mat to find the cheapest route round two walls. The guide also covers how Dijkstra relates to A*, breadth-first search and negative weights.
Read the Dijkstra guideThe guide explains g, h and f = g + h, the priority queue and why the heuristic must never overestimate, then shows A* searching a mat with two walls and compares it with Dijkstra. At the end a robot drives the route it found.
Read the A* guideFind the line, measure the error, steer to correct it. The guide runs bang-bang, proportional and PD line followers on a live robot, shows why the simple one zig-zags, and what to do when the robot loses the line.
Read the line following guideA Kalman filter blends a prediction with a noisy measurement, trusting each as much as it deserves. The guide shows it on a robot measuring its distance to a wall, with demos of what the Kalman gain, Q and R change: a smooth estimate, one that lags and one that jumps about.
Read the Kalman filter guidePID control is how a robot turns to face a direction and a drone holds its height. The new guide explains what Kp, Ki and Kd each change and how to tune them, on a robot you can tune in the page: change a number, press Run, and watch it undershoot, overshoot and settle.
Read the PID guideEvery task in the course is now a challenge you can open straight away: 118 for GCSE, 149 for A level and more for robot club and university. Press Run and it tells you if you have done it.
Try the GCSE challengesThe lessons now have guides: one page for each idea students and engineers search for, such as PID control, Kalman filters and A*. Each explains the idea in plain words and has live demos on the robot that you can change and run in the page.
They are written for A level students and anyone starting in robotics, and each links to the lessons that teach it properly.
See the guidesThe competitions have been rebuilt from the ground up. Every game is sorted into easy, medium and hard, and into single robot, free-for-all and team games, so a class can pick one that fits what it has just learned. None of them are about pushing another robot off the mat: the best program wins.
See the gamesTwo teams, a line down the middle and balls on both sides. Robots pick up a ball with the gripper and kick it at the other team without crossing the line. Get hit and you are out; catch a ball and the kicker is out instead.
Play DodgeballPrint race now has noisy sensors and one question at a time, so a robot that averages its readings well beats one that answers fast and guesses. The simulator loads games and tasks faster, and replays play smoothly with a Play and Stop button.
Try Print raceThe course now runs all the way through A level Computer Science, for OCR H446, AQA 7517 and Eduqas A500QS. Every lesson has a lesson plan, and the specification points each lesson teaches are on the coverage map.
See the coverage mapTeachers can now make their own tasks and competitions in the teacher app: put robots, balls and zones on the mat, then write the rules as simple IF and THEN steps. The simulator checks students' programs against them, just like the built-in tasks.
Open the teacher appAny lesson task can be embedded in a school website, a VLE page or a blog, with your own starting code. The embed builder writes the snippet for you.
Build an embedA new module on the GCSE exam itself: how the papers work, trace tables, pseudocode, programming and long-answer questions, and where marks are most often lost.
Start the lessons