Save Your Food!

image of interface
second image, showing more features, with the print output in the terminal for recipes

Inspiration

My friend (like many people) wastes a lot of food.

What it does

Notifies your food in the fridge is going to expire and goes a step further and suggests multiple recipes to use that expiring food (with optimization put in place to preferentially suggest recipes that use up the most urgently expiring food as possible - reducing food waste). Part of the idea is it also offers something knew to try too.

How we built it

Using VScode, and implementing api (spoonacular, with api key) into a kivy virtual environment to simulate an app interface.

Challenges we ran into

Attempted to compile the program into a format for android phones (with the goal of launching it on an app store) using buildozer with WSL (Ubuntu) but was unsuccessful and ran into issues regarding time pressure, and version and package incompatibilities on the Linux virtual machine we were using. Tested Implementing Open Source LLMs from Hugging Face and using the open AI API. We ran into challenges with version compatibility and modules not working the way they were meant to. In the Case of the Open AI API we were unable read the key from the .env file and therefore we were unable to implement it securely. For the Hugging face LLMs we were unable to find one that responded flexibly enough to prompts.

Accomplishments that we're proud of

Successfully developed a prototype app interface using kivy which works as intended to reduce food waste.

What we learned

To use kivy to an extent, understood trying to integrate apis, learned how to use hugging face and spoonacular. Learned how to use Linux and Linux virtual machines to develop android apps and familiarised ourselves with the app development process. Gained a better understanding of classes in python.

What's next for Save Your Food!

We hope to further develop the app to work in android phones and also include the steps to make the listed recipes, as well as all the total ingredients they need. Refine the app interface too. Add in frozen food toggle (frozen vs unfrozen). Add in non-perishables category. Add in classes of food (i.e. carbs, vegetables, etc.) and refine recipes to preferentially suggest balanced and healthy meals.