WasteSmart

Inspiration

The inspiration behind WasteSmart came from both the growing challenge of properly sorting waste and personal experiences of incorrectly disposing of trash. We noticed how often recyclables ended up in the wrong bins, contributing to pollution and wasted resources. This made it harder for materials to be repurposed and negatively impacted the environment. Motivated by this, we wanted to create a solution that simplifies waste sorting and encourages more effective recycling.

What it does

WasteSmart is an AI-based waste classification tool that uses images to identify waste items and categorize them correctly. With just a photo, WasteSmart helps users sort their waste into recycling, compost, or trash categories, making it easier to dispose of materials in the right way.

How we built it

We built WasteSmart using Teachable Machine for image classification. First, we trained a model using different images of waste items such as plastic bottles, food scraps, and paper. After exporting the model from Teachable Machine, we integrated it into our website. The website allows users to upload photos of waste items, which are then classified by the model and displayed with recommendations for proper disposal.

Challenges we ran into

One of the main challenges we faced was ensuring the accuracy of the model. Training the image classifier to recognize a wide range of waste items with varying levels of detail took a lot of time and effort. We also ran into technical challenges while integrating the model into the website and fine-tuning the user interface for a smooth experience. Additionally, handling edge cases where the model couldn't confidently classify an item proved tricky.

Accomplishments that we're proud of

We are proud of creating an accessible solution that helps users easily sort waste with just a photo. Our AI-based classifier works well for a variety of items and can assist in reducing contamination in recycling efforts. We are also pleased with how quickly we were able to build the project and deploy it for a functional demo.

What we learned

Through this project, we learned how to integrate machine learning models into web applications and how to improve model accuracy with a well-curated dataset. We also gained experience in designing a user-friendly interface and understanding the challenges of implementing AI solutions in real-world applications. Lastly, we learned the importance of testing and iterating on models for continuous improvement.

What's next for WasteSmart

We plan to make WasteSmart more accessible by turning it into a mobile app, allowing users to easily sort waste on the go. Additionally, we aim to expand our dataset for better accuracy in classification and provide sustainability tips to help users reduce waste and make more eco-friendly decisions.

Built With

• css
• html
• javascript
• teachablemachine