EcoSort

The Quest Menu
The social page
The item classification page

What is EcoSort?

EcoSort is a web-based platform that classifies waste as trash, recycling, or compost. The built-in social media feature incentivizes users to increase their sustainability, letting them gain points for using the app and completing quests, enabling them to compete against their friends and on the global leaderboard. It also allows users to post their EcoSort journey through a feed system, sharing their most recent scans with followers.

Purpose

EcoSort addresses two key environmental issues at our high school: proper waste disposal and littering. We saw this putting strain on custodians along with the environment. We needed to create something that could motivate students to clean up effectively, so we created a waste classification tool combined with a social media platform. By using the popularity of social media to drive change, EcoSort is an extremely marketable product. EcoSort started for our school but has further applications both locally and globally.

Implementation

For implementation, we needed to process images from users and recognize them. The text from image classification is further categorized into trash, recycling, compost, ambiguous, illegal, and undefined. Instead of creating the classification algorithm from scratch, which would not be realistic due to time constraints, we used IMAGGA API which could provide us with a vast data set. We considered many options for implementation, such as utilizing graph theory and strongly connected components. We ended up taking a simple solution that kept the program from being overengineered: taking the words retrieved from the API’s classification of the user input, comparing those words to a pre-categorized database we created, and using that for the final output. We utilized Google’s Firebase service to secure user data while providing scalability to EcoSort. Firebase allowed us to authenticate users, store photos to post to feed, implement a follow feature, and let users complete quests. We saved photos as URLs and associated a URL to each account to quickly access each user's last photo for the feed.

User Interface

We designed our user interface with these essential attributes: aesthetics, marketability, and ease of use. Optimized for mobile use, the bottom navigation bar allows users to easily navigate the website. Our branding invokes nature so that users always have sustainability in mind when using our app. We followed a simple design to make it easy for users to follow the app while adhering to popular minimalist principles. Creating an easy-to-use site was essential for users to enjoy using EcoSort, ensuring sustainability isn't a chore.

Built With

Submitted to

HackPNW Summer

Created by

I developed the first idea for EcoSort and contributed by adding to our categories and helping with front end design. I did the write-up and I also made the video (in the last 10 minutes)!

Akash Krishna
I created the back end and implemented the item classification function.

Rio Shintani
I worked on both the frontend and the backend. I created the whole UI and the backend web server.

James Wilson
hey! im James. i love to code
I started up the Firebase project and contributed to its implementation in starting up the points system in the backend. I also learned 2 lines of html to make the error popup green :)

Jun Malit