ReBin

An app making recycling more interactive!

Comment

By: William Choi, Nathan Kim, Timothy Park, and Rohan Soma

Inspiration

We saw how few people were really recycling in our area and decided to do something about it. According to the U.S. Environmental Protection Agency, only 9% of plastic materials are actually recycled. By recycling more, less energy will be needed to produce new materials which in turn lowers greenhouse gas emissions and combats climate change. With ReBin, we hope to help our climate by encouraging people to recycle more.

What it does

ReBin has 3 main tabs consisting of a Home Screen, a Scanner, and a Map. The home screen allows users to view articles and websites regarding recycling to help inform users about the concept of recycling. The scanner allows users to scan plastic codes to see whether the item is recyclable or not. The map contains a map of nearby recycling centers and also user-placed points regarding various recycling items such as recycling bins, landfill sites, or clothing donations. User can place these points by taking a picture on their phone using a QuickCapture app which will update the map.

How we built it

We created the website using HTML, CSS, and JavaScript hosted on a Flask server. To scan plastic codes and recognize them, we utilized a TensorFlow CNN model and a plastic codes dataset. We first standardized the images and labeled them and pushed them through a training model which trains 32 images at a time. We then fit the model to make predictions for each recycling code. The map portion was powered by Esri ArcGIS where we created a map and a quick capture app using ArcGIS and then embedded it using HTML.

Challenges we ran into

  • With our very short amount of time and low experience in hackathons, we struggled creating the app and implementing the features we wanted in time.
  • We also struggled with the TensorFlow model to get a accurate response.

Accomplishments that we're proud of

We're proud that we were still able to implement a map and model with our limited time.

What we learned

We learned how to embed ArcGIS and develop a MVP from start to finish.

What's next for ReBin

  • We hope to continue working on the look of our website to make it more aesthetically pleasing
  • Implement Google Maps to our ArcGIS map to guide users to the nearest recycling center
  • Have a social element such as a leaderboard to have ReBin users in a community
  • "Gamify" recycling by adding achievements and counters for every time you recycle
  • Improve recognition by having our model recognize more objects other than recycling codes

Built With

Share this project:

Updates