Recyclops

An Android app my Bangkit Academy capstone team built that classifies waste with machine learning, shows people where to take it, and rewards them with points for recycling.

Screenshot of Recyclops
Recyclops
Bangkit Academy

The idea

Most people are willing to recycle. What stops them is two small questions: how do I sort this, and where do I take it? Recyclops is an Android app I built with my team as our Bangkit Academy capstone in 2023, to answer both on the spot, and to reward people with points so the habit sticks.

What it does

A user takes a photo of an item and a machine learning model tells them how to sort it, or they scan its barcode with the phone camera to look it up. When the bag is full, the app finds the nearest recycling point and routes them there with Google Maps. Every recycling action earns points.

How I built it

We built it as a native Android app in Kotlin, with Firebase as the backend, and I connected the waste classification model to the app. Recycling points are often in places with a weak signal, so the app works offline: each action is saved on the device first and syncs to Firebase once the connection is back. Nothing a user does gets lost because of a bad connection.

How it went

The app covers the full loop, from sorting an item to dropping it off and getting points for it, and it keeps working where the signal drops. The source code is public on GitHub.

Live links

The public sites for this work.