Part 5: Finish, Present & Reflect
Activity 1-Refinement and Testing
STEP 1 (Refine): Begin by reviewing your prototype from yesterday. Ask yourself what is working well and what feels confusing, unreliable, or unfinished. If needed, return to createai.microbit.org and retrain your AI model by adding more samples, removing unclear samples, or adjusting the actions your model recognizes. Then open makecode.microbit.org and improve your program by adding clearer feedback, better animations, sounds, or messages so a user understands what the micro:bit is detecting.
STEP 2 (Test): Test your invention with at least three different people. Watch how they interact with it without explaining anything at first. Notice where they hesitate, what confuses them, and what works smoothly. Make final adjustments based on what you observe and what they tell you.
Refinement Checklist:
- Make sure your AI model accurately recognizes all actions by testing each one at least ten times.
- Confirm that the user always knows what the micro:bit is detecting through clear visual or sound feedback.
- Decide what happens when the AI is unsure and make sure your program handles mistakes gracefully.
- Add comments to your code so someone else could understand it.
- Finally, confirm that your invention works without being connected to the computer.
Activity 2-Presentation Preparation
STEP 3 (Prepare): Prepare a short two to three minute presentation that tells the story of your invention. Explain the problem you chose and why it matters. Describe how your AI-powered micro:bit helps solve that problem. Demonstrate your invention in action and explain which movements you trained and why you chose them. End with a reflection on what was challenging and what you would improve if you had more time. Practice your demo several times and have a backup plan in case the AI does not behave perfectly. Focus on explaining your thinking, not just what the invention does.
Activity 3-Showcase and Celebration
STEP 4 (Showcase): Present your invention to the group and take time to explore what others created. Try out each other’s projects and offer positive feedback along with thoughtful suggestions. Celebrate how different ideas came from the same tools and how creative problem solving looks different for everyone.
Final Reflections:
Take a few minutes to reflect in writing or discussion.
- How did AI make your project more powerful? What surprised you most about training a machine learning model?
- If you had another week, what would you create next?
- How could your model improve with more data or better training?
Public Product
Use the Padlet board below to share your project. Create a new post and upload a short video showing your project in action, then add a brief description explaining the problem you solved and how your AI-powered micro:bit works. In the post title or subject line, include the first names of all your team members.
