Day 3: Machine Learning at Mission Control

STEP 1: Now that you’ve had some experience with coding, we’re going to shift into the world of predictive AI and machine learning by exploring a tool called Teachable Machine. This simple, web-based platform allows you to train your own AI model to recognize different categories, such as images, sounds, or poses. In real-world applications, tools like this can help monitor components of a spacecraft or track an astronaut’s health and well-being—without needing constant human input or physical presence. To get started, go to https://teachablemachine.withgoogle.com/train and watch the videos provided to learn how to train your first model. Once you understand the basics, try building a simple model yourself to see how AI can be used to make predictions and automate decisions, just like it’s done in mission control.

STEP 2: Now that you’ve learned how to use Teachable Machine, your next challenge is to create and export a model that includes at least four image-based categories used to monitor astronaut well-being during spaceflight. The goal is to help mission control assess the emotional or physical state of astronauts based on visual input, even when direct communication isn’t possible. You will train your model using the following four categories: Focused, Tired, Stressed, and Unresponsive. Using your webcam, collect clear and consistent examples of your face or body language that represent each state. Try to be expressive and distinct with each category so your model can learn the differences effectively. Once your model is trained, export it and test it to see how well it can classify each state. This project gives you a hands-on look at how predictive AI can play a critical role in supporting astronaut safety, mental health, and overall mission success.  Share link to your public model  using the QR code below. In the “Subject” area, include your first name and/or your team members’ names. Click here to see all models. [Click here for instructions on how to share a Scratch lin