SmartFridge

Scan Reciept Screen
Main Fridge Screen
All Current Food Screen
Soon To Expire Food Screen
Recipe Generation Screen
Project in MIT App Inventor

PROBLEM STATEMENT

1/3 of all food in the world is ultimately wasted.

According to the US Department of Agriculture, food waste causes problems such as... greenhouse gas emissions; resource waste; deforestation; biodiversity loss

The average household wastes an average of... 6.2 cups of food per week; 322 cups per year; 360 medium-sized take-out containers a year;

This problem is severely damaging NATURE.

Inspiration

Ever find moldy and expired food in the back of your fridge, forcing you to throw it away? I agree, we all have. Our project SmartFridge's goal is to eliminate this problem that EVERYONE faces in order to reduce food waste.

What it does

SmartFridge allows you to see what you currently have in your fridge, when these items will expire, and what recipes you can make using these soon-to-be expired foods. To add food to your fridge, simply use the app to take a picture of your shopping reciept. The app is "smart" and uses AI technologies to filter out food items and non-food items. To see what foods will expire soon, drag the slider to indicate your desired time-frame and see in chronological order which foods will expire. To consume this food before expiration, allow the app to generate multiple recipes of your choice.

How we built it

We used MIT App Inventor, an app developing software, to create SmartFridge. We used many AI technologies to assist us, including chatGPT API and an Image to Text API. Additionally, we used TinyDB, a local storage system offered by MIT App Inventor.

Challenges we ran into

The logic behind our code is quite complex. We had multi-layer loops of lists and dictionaries, including custom sorting algorithms to organize the foods. The unique syntax and formatting of MIT APP Inventor added on to this challenge. Additionally, we faced many challenges while trying to implement the image to text API. It is particularily challenging to be able to use OCR on a crumpled and disoriented reciept. It took us hours of researching to find and tune a working model that would meet our goals. Also, to actually use and test our app, we had many technical issues in the first couple hours of the hackathon. It turns out that the UCSD guest WiFi blocks this connectivity, causing us to have to figure out a workaround.

Accomplishments that we're proud of

I believe that this is the best project we have ever made at a hackathon (to date). SmartFridge is fully functional and complete, allowing us time to work more on the UI design and the presentation. This is also the most complex project we have coded at a hackathon, as we had complex code logic while seamlessly integrating AI techniques.

What we learned

This hackathon was truly a great learning experience. This hackathon taught us the importance of starting early to work out any kinks along the way. Additionally, it improved our skills of researching tools and reading documentation.

What's next for SmartFridge

SmartFridge is an app that can be impactful in the real world. With almost every family owning a fridge, this app can be integrated to almost EVERYONE's daily lives. More importantly, because of the vast impact of this app, there will be noticable change in nature and the environment. Based off of user feedback, we can also continue to develop an extension of SmartFridge---SmartPantry. This would have the same concept but for foods in your pantry. Afterall, everyone hates it when they forget about their yummy snacks, just to discover that they expired last week. Next steps in the near future including improving the smoothness of the app, releasing to users for user testing and feedback and editing the app based on this feedback.