We were inspired by both the sustainability track, the mechanical track, and also wanted to incorporate some machine learning into our project. We decided that an automatic trash sorting conveyor belt was the best combination of the 3 for us. The mechanical side involved making sure the motors were in sync, figuring out how a conveyor belt would work, and integrating the machine learning output with our motors. The CS side involved training the machine model, and correctly deciphering the data that it outputted and turning it into usable cues for our motors. We ran into many challenges, like having to reiterate the conveyor belt lenght many times, changing the dowels, syncing the motors, and training the model multiple times.
Log in or sign up for Devpost to join the conversation.