Introduction to Artificial Intelligence with MakeCode and micro:bit

Purpose & objective

The purpose of this workshop is to support educators with a simple, technically accurate, and creative way to apply nearly any data collection activity to machine learning with students.

Outcomes:

1) Participants will know how to collect data using digital sensors and be able to use that data train a machine learning model to recognize conditions. (For example, participants will train a personalized 'smart hat' to recognize when they are nodding 'yes' or shaking their heads 'no'.)

2) Participants will know how to integrate these skills cross curricularly and be able to teach lessons found at microbit.org and makecode.microbit.org. (For example, in a lesson on informational texts, students may self-select a topic to research and generate a solution. Students may use activities found at https://microbit.org/teach/do-your-bit/#micro:bit-projects to support their design process.)

Tools we will use:
1) makecode.microbit.org
2) ml-machine.org
3) microbit.org
4) micro:bit devices

Evidence of success:
1) Participants will each collect data and train at least two different models in the workshop. Participants will create one prototype project using a machine learning model they trained.
2) Participants will post to a digital board at least one way they plan to integrate something they've learned with students.

Outline

Note on accessibility: These learning resources can be adapted with Microsoft Immersive Reader (free accessibility tool integrated into tutorials). Each of these resources are translated into between six and thirty-two languages.

Example activities:

1) Tutorials https://microbit.org/projects/make-it-code-it/
2) Tutorials https://makecode.microbit.org/
3) Lesson plans https://microbit.org/lessons/
4) Design challenges https://microbit.org/projects/do-your-bit/
5) Ml-machine.org

We will use exemplars (by permission) from educators and students around the world.

Example: https://microbit.org/news/2021-09-27/students-invent-asthma-prevention-idea/

5 min - Introductions of presenters and attendees. Share why you came today.
15 min - Participants create a step-counter (or some other data logging device) - (hands on, everyone makes something)
15 min - Participants extend activity by creating machine learning model using accelerometer. (Participants learn how accuracy improves with their model as compared to first example.)
5 min - Participants see how teachers in UK are using this activity (plus four other related ML activities) with students to contribute to Office of National Statistics database.
5 min - Participants review and explore accessibility features of Make Code.
7 min - Participants work independently or with someone at table to brainstorm one idea they want to try with their students and post to Padlet.
3 min - Participants share out their ideas

All remaining time: hear from expert practitioners through video clips, photos, and lesson plans, participants share their own experiences for discussion and feedback. Q and A.

Supporting research

aka.ms/MakeCodeResearch

aka.ms/PhysicalComp

https://microbit.org/impact/research/