Micro:bit Create AI

Using the accelerometer and processor of the micro:bit V2, you can experience and learn about artificial intelligence in the real world through movement and machine learning (ML). micro:bit CreateAI is a free, web-based tool that makes it easy for students to explore AI through movement and machine learning, and take it into the real world with a BBC micro:bit development board (v2). You can training a machine learning model with your own movement data with your web browser. Then you can then create a MakeCode program that uses this machine learning model and runs on micro:bit v2 board.

Adding artificial intelligence (AI) to the BBC micro:bit is made possible with a free online education tool that allows your students to combine AI with coding in Microsoft MakeCode blocks. The plan is that with the micro:bit and CreateAI students can develop an AI-powered piece of tech from scratch in 90 minutes! If you have experience with embedded system and micro:bit, you can do that much faster.

micro:bit CreateAI is a free, web-based tool that lets you program a micro:bit to recognise and respond to your movements, like clapping, waving, dancing or jumping. The movement data used for AI model teaching is recorded from the micro:bit’s accelerometer and transported to PC using Bluetooth. The recorded data is used to train, test and improve your own machine learning model, then use it in a Microsoft MakeCode tool that is used to program the micro:bit board.

Getting started with micro:bit CreateAI
https://www.youtube.com/watch?v=yWl3WxDE6QI

Discover micro:bit CreateAI
https://www.youtube.com/watch?v=kqNy5N3WLtM

Try it at createai.microbit.org
Discover more at microbit.org/ai

BBC micro:bit is a dev board designed for computer science and technology. It runs on a Nordic Cortex-M4 chip with radio + FPU capabilities. The version v2 boasts of a speaker, touch sensor, 25 LEDs, 2 buttons, accelerometer, bluetooth, a mic and a couple of GPIO pins.

The nRF52 application processor is where user programs run. A single, complete application including user code, runtime code and Bluetooth stack is loaded and run directly from on-chip flash memory.

The display is a 5×5 array of LEDs. It is connected to the micro:bit as a 5×5 matrix. Runtime software repeatedly refreshes this matrix at a high speed, such that it is within the user persistence of vision range, and no flicker is detected.

The micro:bit has a combined accelerometer and magnetometer chip that provides 3-axis sensing and magnetic field strength sensing. It also includes some on-board gesture detection (such as fall detection) in hardware, and additional gesture sensing (e.g. logo-up, logo-down, shake) via software algorithms.

Links to more information:
https://microbit.org/get-started/user-guide/microbit-createai/
https://github.com/alan-turing-institute/microbit-rs
https://microbit.org/get-started/features/ai/
https://microbit.org/new-microbit/
https://microbit.org/news/2024-11-20/microbit-CreateAI-launch/
https://microbit.org/news/2024-11-20/teaching-and-learning-about-ai-with-the-microbit/