How we designed a robot that sorts waste

When the science club asked us for an Innovation Week project, we were not looking for a robot. We were looking for a problem.

Half an hour watching one bin

Seven in ten containers went to the wrong place. That was our starting number, and it did not come from a study — it came from standing in the canteen with a notebook.

• Cardboard arm — could not hold a single bottle.
• Light wood — held the weight, could not tell what it held.
• Printed frame — three seconds per decision, far too slow for a queue.

The fix was not a pricier sensor. It was a simpler question.

Does the robot need to know exactly what the waste is, or is "metal / not metal" enough? It was enough.

وفي اليوم الأخير صوّت الطلاب على ثلاث أفكار لتتحوّل إلى مشاريع فصلية كاملة.