Plate Predict

Inspiration

In our high school experience, we witnessed first hand the wasting of massive amounts of food during lunch time. This is mostly because students get too much food that they do not want to eat, and end up wasting and throwing it away. Since the school cannot reuse the food that another student has touched, they end up losing lots of money and valuble resources that could be allocated to other places. We wanted to find a solution to this, not only because wasting money is bad, but because wasting food is worse.

What it does

Our app takes in a student ID and a photo of the student's lunch waste taken by a trash monitor. These are then sent to our backend, which assigns an index to how much food the student wasted and, based off past data, predicts how much food the student will need for the next lunch so that the student doesn't end up wasting food. Additionally, it gives statistics for the whole school based off student data so the school doesn't overproduce food and have to throw it out, wasting money.

How we built it

We decided to make a web app using Next.js and Tailwind CSS for the frontend, and Node.js, MongoDB, Flask, and Gemini Requests for the backend. Using AI and machine learning to predict data provied crucial to the development of our project.

Challenges we ran into

We ran into multiple challenges in both the front end and the backend. In the front end, we ran into issues with the CORS policy, SSH certificates for camera access, and learning new TypeScript concepts. For the backend, issues with connection to the database and data transfer. We managed to resolve all these issues to get a final working project.

Accomplishments that we're proud of

We are proud of fully integrating a database and data transfer to and from the database in this project. We are also proud of learning new frameworks and designs for frontend and expanding our knowledge on user input through cameras.

What we learned

We learned that it is important to plan out your project before you actually do it, so that the coding is smooth and effortless. We also learned the importance of collaboration between frontend and backend programmers.

What's next for Plate Predict

To expand on this project, we can get a barcode scanner to more easily recieve student ID, more effectively train machine learning to do better predictions, and get our app verified by school districts for real-world use.