Bincredible

Splash Screen
Home Screen
Image Capture Screen
Statistics Screen

Inspiration

Around campus and beyond, there is a need for better waste management and sustainability. We felt that this issue of improper disposal of waste could easily be avoided if people were more educated about the types of waste and how to identify them. Thus, we decided to make Bincredible ('bin' as in 'waste bins') in hopes of making a positive environmental impact.

What it does

Bincredible enables users to take an image of their waste and check if belongs to the recycling, trash, or compost bins.

How we built it

Our front end was an iOS app launched with Expo Go, developed in JavaScript and React Native. Our Computer Vision Model used for image classification was created with YOLO11 and Llama-3.2 AI. Our frontend-backend integration was handled by Flask API. And finally, our app wireframes were designed in Figma.

Challenges we ran into

We ran into many challenges along the way. One difficulty was running computer vision models on the Intel PC. It was a brand environment that we were unused to and we even accidentally lost connection to it, resulting in hours of lost work. Another difficulty was integrating the frontend with the backend using Flask. This was quite challenging because we were all relatively unfamiliar with app development. Also, our statistics page was not fully functional because of time-restraints.

Accomplishments that we're proud of

Our greatest accomplishment was getting our computer vision model and LLM to work in conjunction, which was very time-consuming with the YOLO11 training we had to do. Another accomplishment we were proud of was creating a functional UI with a smooth page flow.

What we learned

Firstly, we learned various languages/tools we had previously little experience with in a short time. Secondly, because it was the first hackathon for most of our team, we learned how to work efficiently and collaboratively in a fast-paced environment.

What's next for Bincredible

We learned a lot of new skills from this hackathon, but there is a lot to improve on. Our computer model, app UX, and statistics feature can all be improved. We hope to make Bincredible a practice tool to improve waste disposal practices.