The inspiration for Recycle Rush came from observing how litter negatively impacts the appearance of public places. We noticed that much of the litter consisted of bottles, cans, and paper products—items that could easily be recycled. Realizing that recycling these products would significantly improve the cleanliness of public areas, we decided to create a gamified recycling app. By making recycling fun and competitive, Recycle Rush encourages good recycling habits and helps keep public spaces cleaner.
Recycle Rush is an innovative app designed to promote conscientious recycling by allowing users to track their recycling efforts. Utilizing Google Firebase, the app keeps a detailed account of each user's progress through a personalized account system. Users can easily see how much they have recycled, with their achievements displayed on a dynamic leaderboard that highlights top recyclers. This engaging feature encourages friendly competition within various groups, such as friend circles, classrooms, or community organizations, by running recycling challenges. We have also implemented a machine learning recognition system where the user must photograph the objects they are recycling as proof, which helps promote integrity within our app. By making recycling a fun and competitive activity, Recycle Rush inspires users to recycle more and adopt sustainable habits.
Recycle Rush was developed using Swift and Swift Storyboards for the front-end interface, ensuring a smooth and intuitive user experience. For the back-end server, we utilized Google Firebase, which provides robust and scalable infrastructure to manage user accounts, track recycling progress, and maintain the leaderboard system. This combination of technologies ensures that Recycle Rush is both reliable and user-friendly, making it easy for users to engage in and promote conscientious recycling. The AI litter detection feature enhances its functionality, employing CoreML for iOS, Roboflow for image annotation, and Liner for model training. Users can simply upload an image via a new entry button, allowing the system to swiftly identify litter presence and update accordingly, reinforcing the app's commitment to environmental sustainability.
During the development of Recycle Rush, we encountered several challenges. Firstly, there was a significant learning curve when working with Swift compared to other languages like Python, which required additional time and effort to master. Additionally, managing constraints to ensure the app works seamlessly across various iPhone sizes proved to be a complex task. We also faced issues with cell sizes not adapting correctly when creating the leaderboard, which required meticulous adjustments to ensure a consistent and user-friendly interface. Furthermore, we initially struggled with app crashes due to excessive RAM usage caused by the AI model. To address this, we implemented code to control RAM usage, preventing crashes and ensuring smooth performance. Despite these challenges, we successfully overcame them to deliver a robust and engaging recycling app.
We are proud of several key accomplishments in developing Recycle Rush. Firstly, we successfully integrated a Firebase back-end server, which provides robust and scalable infrastructure for our app. We also ensured that the app works seamlessly on multiple phone sizes, making it more accessible to a wider audience. Additionally, we implemented an AI model that correctly detects litter, significantly enhancing the app's functionality and user engagement. Lastly, we are particularly proud of being able to create a fully functional app within just a few days, showcasing our ability to quickly and efficiently bring our ideas to life.
Throughout the development of Recycle Rush, we gained valuable experience in several key areas. We became proficient in using Xcode for iOS app development and learned how to integrate Firebase servers into our Xcode apps to manage user data and track progress. Additionally, we mastered implementing adaptive UI through code, allowing individual cells to adjust based on screen size and the information displayed. We also discovered the importance of iterative design; our app's initial versions lacked visual appeal and functionality, but through continuous iterations, we refined and improved it to create the polished and effective version we have today. Moreover, we learned to use CoreML and incorporate it into Swift, enabling our app to recognize recyclable items through machine learning.
Looking ahead, we plan to introduce the ability for users to compete in smaller, private groups, as the current version operates on a worldwide server. This feature will enhance the app's appeal for friend circles, classrooms, and community groups looking for a more personalized experience. Additionally, we aim to publish Recycle Rush on the App Store, making it accessible to a broader audience. While there are cost barriers associated with this, potential prize money could help us overcome these financial challenges and bring Recycle Rush to users everywhere.