Inspiration

We aim to revolutionize waste disposal and promote sustainability. Improper waste disposal contributes to pollution, harming ecosystems and threatening human health. Waste that could be recycled or composted ends up in landfills, wasting valuable resources. Many individuals lack the knowledge and motivation to sort their waste properly. We hope to raise awareness about proper waste disposal.

What it does

Users can take photos of their waste, and the app analyzes the image using machine learning to identify the type of material. Based on the analysis, the app classifies the waste as trash, recyclable, or compostable, providing clear instructions for disposal.

How we built it

We gathered a large dataset of images of different types of waste from various sources, including Kragle. We carefully labeled each image, categorizing it as trash, recyclable, or compostable, to train the model. We utilized Create ML to build and train our machine learning model, using the labeled data to teach it to recognize different types of waste. We rigorously tested the trained model on a separate set of images to ensure its accuracy and effectiveness.

Challenges we ran into

It was difficult and time-consuming to integrate our Custom Create ML model into our code on XCode. We also had to train and change our dataset multiple times for accurate results.

Accomplishments that we're proud of

We are really glad that we could create an ML model by ourselves.

What we learned

We learned a lot about Xcode, Swift, Github, and machine learning.

What's next for TrashTracker

We plan to incorporate additional features, such as waste tracking and analysis, to provide users with a comprehensive understanding of their waste management practices. To expand our reach and impact, we aim to partner with local governments, recycling companies, and environmental organizations. We envision a world where TrashTracker is available in multiple languages and regions, promoting sustainability on a global scale.

Built With

Share this project:

Updates