Machine Learning with the Wio Terminal

The Wio Terminal can be a starting point for learning and implementing simple machine learning systems - and there are growing set of resources that can help with this. These rely on the Wio Terminal to provide the computational power and as an interface for various sensors. The different programming environments provide different interfaces for installing machine learning code.

1. TinyML

Getting started with TinyML:. A copy of the course book can be downloaded at: https://wiki.seeedstudio.com/Wio-Terminal-TinyML/ The structured course is designed to introduce beginners to the basics of embedded Machine Learning with Wio Terminal and Codecraft no-code graphical programming. Codecraft s a free no-code programming environment, similar to MIT Scratch, with code represented by graphical blocks. Program loops are constructed in the conventional way, in contrast to the dataflow model employed in XOD.

A full introduction to machine learning is provided at: https://wiki.seeedstudio.com/Wio-Terminal-TinyML-Kit-Course/, along with course materials and link to the host Github site.

Check out the Codecraft site at: https://ide.tinkergen.com, and documentation for Codecraft - including use of machine learning blocks for programming the Wio Terminal: https://www.yuque.com/tinkergen-help-en/codecraft/tinyml

Codecraft tutorial: https://wiki.seeedstudio.com/Guide_for_Codecraft_using_Arduino/

2. TinyML ‘Course-in-a-box’

In partnership with Seed Studio and Prof Vijay Reddi from Harvard University, this kit offers an exclusive course that enables individuals to obtain Embedded Machine Learning theory and skills through hands-on exercises using Edge Impulse and Wio Terminal & Grove system, aiming to provide educators and TinyML beginners a “course in a box” experience. Access to the course material is via a code provided with the purchased box of hardware. Covers from theory to practice: introduces fundamental theory to actual applications of ultra-low power machine learning at the edge. Supports Codecraft graphical programming, Arduino, Micropython,etc

3. Edge Impulse

Edge Impulse also enables developers to create intelligent device solutions with embedded Machine Learning. Machine Learning at the “edge” enables efficient processing of sensor data on collection, that can then be discarded, saving cost, bandwidth and power. Wio Terminal is now supported by Edge Impulse.

Edge Impulse uses an online Studio/descktop model to create and optimise machine learning models, and to download these onto the connected device - which can include the Wio Terminal. More information at:

Getting started wizard: https://wiki.seeedstudio.com/Getting_started_wizard/

Starting Edge Impulse: https://wiki.seeedstudio.com/Wio-Terminal-TinyML-EI-1/

Hackster tutorial: https://www.hackster.io/news/tinyml-for-all-developers-with-edge-impulse-2cfbbcc14b90

Documentation: https://docs.edgeimpulse.com/docs/