Situation: company cafeteria. An innocent worker is thinking about something else, and throws everything on her tray in the trash on the way out. On her way up the stairs, she remembers — she was supposed to compost her spoon.
Our project aims to avoid instances like this by providing a quick, easy, and no-focus-needed way to sort discarded items and lead the user to the correct bin. Also some of us would rather die than write a web app so there's that.
What it does
An item the user wants to discard is placed in front of a camera. The MyRio then identifies the item in real-time using machine vision. Once the item is identified, the name of the item appears on an LCD screen. Finally, colored LEDs corresponding to the bins light up based on whether the item should go in the compost, recycling, or trash.
How we built it
So. Much. Labview.
Challenges we ran into
Machine vision sucks. Surprise field trips back to E school. We had to carry all our hardware from Rice to Newcomb and back again, and that stuff's not light.
Accomplishments that we're proud of
Machine vision works finally! Teamwork and integration of discrete components went surprisingly great.
What we learned
Don't plug power supplies into things without checking the output first.
What's next for Sustain-A-Trash
Integrating it with outside devices to improve the user experience. AKA stop using single LEDs as indicators because we can do better.