sustaino

The initial upload interface of sustaino
Test run of the API's functionality
Revised UI
Our back-end python function where we call ChatGPT
The front-end workings of our image upload
Map of nearby waste disposal centers (relative to our current location of SF)

Inspiration

We are highly concerned about the waste management issue as landfills are filling up fast and the garbage patches in the oceans are only getting bigger. So, we wanted to make waste disposal more sustainable by encouraging more recycling.

What it does

Using the power and precision of OpenAI’s GPT-4o, _ sustaino _ has the photo-analyzing capabilities to identify and classify different waste items based on their disposal category (landfill, recycle, compost, hazardous). Once doing so, our site has a convenient map to point out nearby disposal facilities for your given waste type.

How we built it

On the backend, we mostly used ChatGPT to carry out our main functionalities of identifying and classifying different items. For the front end, we used Flask and React to shape the website's look and add its essential button and image-uploading functionalities.

Challenges we ran into

ChatGPT API keys are extremely finicky, and our project revolved arounddebugging and manipulating other pieces of code in order to get them to work.

Accomplishments that we're proud of

Despite all of the frustrations, in the end, we're quite proud of how the AI response portion turned out in the end. In fact, all the struggle we put into getting the APIs to work only added to how rewarding the finalized project was.

What's next for sustaino

We’d like to equip _ sustaino _ with even more features in the future. A few ideas for this could be to gamify the proper disposal of different items to boost user engagement and motivation. We could add limited-time events where users can compete against their friends, family, neighbors, and other acquaintances.