EcoRewards

A waste management and recycling incentive program (demo based in Kenya) that encourages buyers and merchants to contribute towards a sustainable future through recycling.

Comment

About the Project

Inspiration

A recent survey showed that 66% of Americans won't recycle if it is not convenient for them. And this is true in most parts of the world, including here in Nairobi, Kenya where I live. Hence EcoRewards came to mind and the question was "How can we incentivize recycling in such a way that benefits customers, merchants and the planet?"

What it does

EcoRewards has various purposes:

  1. Creating awareness on recycling - Through a personalized AI assistant, or as we call it, EcoBuddy. This coupled with blogs, Image Classification, AR, VR and video will ensure that our users remain actively educated on such issues and will constantly encourage them to play a role in taking care of our planet.

  2. Incentives and Donations - EcoRewards makes recycling worth it by offering users a variety of rewards through partner retail stores and merchants. The more a user recycles, the more rewards they get. The more merchants join EcoRewards, the more customers they attract and eco-conscious they become. And you can imagine the positive effect that this will have in the long-run. Better yet, points garnered from recycling can be used to donate to charities in need.

How I built it

  • Mobile App: Flutter with BLOC
  • Database: Firebase
  • ML(Object detection & classification): TensorFlow
  • Generative AI: Open AI's GPT
  • Developed for: Android(tested on Android 13), no iOS app just yet(I don't have a Macbook lol)

Challenges I ran into

  • Designing a good AR experience: There was quite some technical debt when it came to integrating Augmented Reality experience. I ended up using the Life Tree concept which I plan on refining in the near future

  • Fine-tuning the AI: Finding good data to train my Object Detection model was quite the task. The model is still in its early stages so some predictions are off. The current model is trained on 12,000 images from different sources but more can definitely be done.

  • Integrating Square's APIs: Since I'm currently in Africa, I can't verify my online merchant account. And the Sandbox account doesn't offer Online Checkout(for processing donations).

Accomplishments that we're proud of

  • Created a user-friendly application: I am proud to have developed an application that's not only easy to navigate, but also enjoyable to use, making recycling a more appealing activity for our users.

  • Built an interactive AI assistant: EcoBuddy was pretty fun to build and play around with.

What I learned

Square has a very robust API that can be applied in several fields. I'll definitely explore it further in the near future.

What's next for EcoRewards

  • Further Integration: I hope to expand integration with other eco-friendly companies to broaden the rewards catalog and increase user impact

  • Advanced Object Detection: I hope to also enhance the machine learning model for more accurate recognition of a wider variety of recyclables, improving the user experience

  • Social Features: I hope to develop a social aspect for users to share their achievements, compete, and inspire each other towards more eco-friendly environment

Built With

Share this project:

Updates