Food Boost

Homepage!
Our maps allow people to locate donors

Inspiration

Throughout our lives, we have constantly noticed gargantuan amounts of food being thrown out. Not much thought was given to it, until we came across a New York Times article that explained how 1.3 billion tons of food are wasted worldwide each year! This horrifying fact plagued us, how could restaurants be plaguing the environment this much? We knew something had to be done about this issue...

What it does

Food Boost is the solution to this atrocious issue; a platform where users and donors alike can reciprocate a shared love of the environment, and aiding the rise to a wastage-free future! It allows users to donate food and old-clothes, and allows people in need to receive these items, saving on costs and emissions.

How we built it

Food Boost was build using a myriad of tools. The website was mostly done using JavaScript, while using libraries like leaflet.js for the map, tensorflow.js for the AI image recognition, and much more. The chatbot for any queries was built using Twilio & GPT-3. The mobile app was developed using the Kivy library in Python.

Challenges we ran into

The AI image recognition had lots of data-points (over 20,000!), which was hard to develop, given the inability to share files, and the long training and uploading times. Also, given the difference in time-zones (our team is located in India!), we were attending interesting hackathon workshops during the night time, and had to pull all-nighters! The project we chose to work on was also extremely large and forced us to go through intense amounts of research, testing, and learning to be able to effectively create a working program that easily accomplishes our goals.

Accomplishments that we're proud of

Given these restrictions, we're proud of the sheer amount of functionality our website & mobile app have, and are proud to have developed such a platform for environmental aid. Making a feasible chatbot was also a massive undertaking, but also incredibly rewarding when we managed to make it work.

What we learned

As they say, every experience in your life is being orchestrated to teach you something. Personally, we believe we learnt a lot from this hackathon, especially in the machine learning and front-end design departments (also how to center a div!). We also had to explore a lot of previously unknown territory to be able to accomplish to tasks we needed to, including how to make a mobile app from scratch!

What's next for Food Boost

We wish to improve the mobile app to have more functionality (such as the ability to upload images from the camera), as the mobile user-base is growing more prevalent than the one on PCs. Also, we hoped to improve the QR code system, given the small vulnerability in the way it works (using URL parameters, while using cryptographic proofs, and distinguishing accounts between donors & users would have been safer). Also, more options for uploads would be hugely beneficial, as food and clothes are not the only items wasted (eg: paper, and especially plastic). We would also like to create incentives for big name companies and restaurants to support, promote, and use our app to overall help our cause.