Team Name: Girls Who Code at Broward
Project Title:Trash to Treasure (T³)
URL of Project: https://replit.com/@kalinessa/Trash-To-Treasure-T3
Video: https://youtu.be/5cbSDr9zY5A
Write-Up: As the climate crisis worsens, communities in South Florida are especially impacted. Our team focused on an issue connected to climate change: waste. Having learned about reducing, reusing, and recycling since elementary school, we were surprised that our communities of Davie and Pembroke Pines recently began sending all recyclables to be incinerated. A prominent reason for the shift from recycling to incineration was contamination, which refers to accidentally mixing non-recyclable items with recyclables, interfering with the recycling process. Many people cannot differentiate between certain trash and recyclables, failing to realize the harm of mixing up the two. By educating peers about recycling and composting, and providing tools to identify non-recyclables, we can help reduce contamination. Recycling not only saves energy and natural resources, but also turns what would have been trash into treasure, which is how we came up with our project name “Trash to Treasure.”
Our project is an informational website with an embedded artificial intelligence (AI) model to raise awareness about recycling and composting, offering tools to make the process more accessible and convenient. We created the AI model using Google’s Teachable Machine to detect whether certain objects can be recycled or not. We trained the model by taking hundreds of pictures of objects commonly found at our school and community. For example, we included photos of Dunkin’ Donuts bags and cups in our training data because our school has a Dunkin’ Donuts. Our project can be used by anyone, but our intended audience is people in our community. Through Trash to Treasure, we hope to raise awareness about the importance of recycling and reducing food waste, to keep our school and community clean. We believe in the power of education to inspire positive change. On our website, we provided a “Recycling 101” page with guidelines about recycling, a “Composting 101” page about composting and reducing food waste, and a “Resources” page with more information and action steps to take. While our AI model is currently only effective for a set amount of objects, we plan to expand our project, making the model able to detect compostable items, such as food scraps, and specific details. For example, wooden pencils can be recycled, but only if the eraser and metal elements are removed. We plan to add more information to the model so that it detects these small elements and shares actionable advice.
The team had different levels of knowledge in coding websites, machine learning, and reducing waste when we started this project. Because of this, we all contributed in various ways. For example, one team member had a background in working with climate justice organizations. From her explanation about recycling, the rest of the team could understand the motivation and potential impact of our project. Another member had experience using Teachable Machine. When training our model, we learned about how to take pictures that would produce the best result. When working on the website, some of us had intermediate experience with HTML, but half of us were beginners. Working on the project pushed us into learning HTML, as well as CSS and Javascript, so we could all collaborate and help a teammate who couldn’t figure out a bug. We also learned about Broward County’s specific recycling guidelines, as well as general guidelines, to implement into our model and website. We became aware of not just the global impacts of sustainability, but also the recycling efforts and consequences in South Florida. Lastly, we learned about teamwork and accountability. By communicating consistently and respecting each other’s ideas and input, we were able to create a cohesive final project.
The biggest challenge for our project was embedding the AI model into the website. We initially planned to create just an AI model, but we then learned that creating a website would not only showcase the model better but also challenge ourselves to see what we can create. Once we decided to build an AI model and website to address recycling and reducing waste, the rest was putting our idea on paper—or, in our case, code. We established a timeline and divided tasks: Every member took pictures of objects to train the AI model, and two started coding while the others researched and wrote about the issue. As we made progress, we became more excited and motivated. The most rewarding part of this project was seeing our ideas come to life, from creating our own logo, to feeding hundreds of images to the AI model, to coding the website.