Integrated STEM: Programming Edge/IoT Systems with AI

Hold your micro:bit up to the camera! While TAs verify everyone's connection, open MakeCode and flash the default "Hello" program to your micro:bit. If it works, you're golden. If not, join the Help Room breakout.

In the shared doc: Name one real-world sensor or data-collection scenario that connects to a topic you teach. (Temperature in a greenhouse = biology. Accelerometer on a roller coaster = physics. Light sensor for accessibility = CS.) Star your favorite from someone else.

What image comes to mind? (a) A dark screen with green text, (b) A kid in a robotics club, (c) Something I don't do, (d) Physical things happening. Vote!

In small groups: What's the #1 barrier to using physical computing in your classroom? (Cost? Time? Your own comfort? Curriculum fit?) Brainstorm one way the micro:bit + LLM combo addresses that barrier. Share out to whole group.

Replicate what you just saw: in MakeCode, read the temperature sensor and show it on the LED display. Got it working? Add the light sensor too. Feel free to ask an LLM for help — paste "How do I read the light level on a micro:bit in MakeCode?" and see what it says. Flash it to your micro:bit and test.

In your group, build a program that combines 2+ sensors and triggers an action. Ideas: LED alert when temperature is high AND light is low. Shake detector + compass for direction. Sound level + temperature for "classroom comfort index." Use an LLM to help write or debug your code. Document in the shared doc: what you asked the LLM, what it got right, what you fixed. Everyone programs their own micro:bit.

Run these tests on your micro:bit: (1) Cover the light sensor — does the value go to 0? (2) Put your thumb on the CPU — watch the temp rise. (3) Compare the temp reading to any thermometer you have. Apply the Check the Machine protocol: Task → Before (what I expected) → After (what I got) → Takeaway. The micro:bit temp sensor reads CPU heat, not ambient air — that's a real engineering calibration problem! Verification isn't just for AI — it's for ALL data.

Open the NGSS-aligned IoT lesson template. Customize it for YOUR classroom: (1) Pick your sensors. (2) Define the data collection task. (3) Write the student-facing Check the Machine prompt. (4) Note the NGSS standard it hits. Pair-share when done: "What's the first IoT lesson you're going to try?"

Browse your resources: MakeCode project files, MicroPython starters, sensor reference guide, NGSS alignment crosswalk. The micro:bit is yours — take it to your classroom.