Sydney AI Design Lab Weekend Class | Years 7–10
Could an AI recycling assistant still recognise a drink can after someone crushes it? Students spend eight weeks building an AI prototype for a real-world problem, testing it on unfamiliar examples and deciding when a person needs to step in.
Book nowWhat your child will do
Define the problem
Choose who the system will help and what decision it needs to make. For a recycling assistant, students might start with three items commonly found at school and check the school’s bin rules.
Build a dataset
Photograph and label examples, decide how much variety to include and set aside unfamiliar items for the final test.
Train and investigate
Train an image model and record its predictions. Does it recognise a can on a different table, or has it learned to rely on the background? Students change one thing at a time to find out.
Build and defend a prototype
Connect the model to a simple interface, decide when it should ask the user to check an item, and demonstrate where the system works reliably and where it still needs improvement.
What they’ll make
Each term, students build and demonstrate a working prototype. In a recycling project, for example, their system might use a photo to suggest which bin an item belongs in. By the end of term, students can show families a prediction that went wrong, explain what they changed and use their test results to show whether it helped.