WastEd

(Sorry about the AI summary, we are a bit short on time)

Inspiration We wanted to tackle the growing challenge of food waste in schools by merging IoT technology with AI. Our goal was to create a system that not only monitors waste in real time but also offers actionable insights, encouraging more sustainable practices.

What it does WastEd attaches Raspberry Pis to trash bins, capturing images every 30 seconds via an endpoint (e.g., /latest-image/). Through a user-friendly interface, schools can register, log in, and add bins. Each bin is displayed as a card showing a live image, a food waste score, and key statistics. Additionally, AI analyzes bin snapshots taken every ten minutes, providing detailed counts (like food trays, unfinished burgers, milk cartons, and more) with corresponding emojis and color-coded scores.

How we built it Hardware Integration: Raspberry Pis are installed on trash bins to capture images regularly.

Backend Development: We used FastAPI to build a robust backend that serves the bin images and handles requests.

User Interface: The admin dashboard allows schools to register, log in, and manage bins effortlessly.

AI Integration: An OpenAI API call runs once every ten minutes for each bin, generating detailed food waste statistics.

Data Visualization: Snapshot statistics, including images, counts, and overall food scores, are presented in a compact and easy-to-read format.

Challenges we ran into Real-Time Image Capture: Ensuring that the Raspberry Pis consistently provide up-to-date images.

Data Synchronization: Maintaining smooth communication between the hardware and server, especially under varying network conditions.

Efficient AI Calls: Balancing the frequency of AI-generated insights to avoid overloading the system while ensuring accurate monitoring.

User Interface Design: Creating a dashboard that is both visually appealing and capable of handling a large volume of snapshot data in a compact layout.

Accomplishments that we're proud of Successfully integrating IoT devices with a fast, responsive FastAPI backend.

Developing an intuitive interface that empowers schools to actively monitor and address food waste.

Seamlessly incorporating AI to extract meaningful insights from periodic bin snapshots.

Ensuring consistent color coding and emoji associations for easy, at-a-glance interpretation of food waste severity.

What we learned The importance of reliable hardware-software integration in IoT projects.

Strategies for optimizing API calls to balance performance with resource limitations.

How to design user interfaces that effectively manage and display large datasets in real time.

Valuable lessons on using technology to drive sustainability and environmental responsibility.

What's next for WastEd Expansion: Scaling the system to include more schools and additional types of waste bins.

Enhanced AI: Refining our AI algorithms for even more precise waste categorization and predictive insights.

Improved Analytics: Developing deeper analytical tools and dashboards to help schools make informed decisions.

Community Integration: Exploring partnerships with local waste management services to broaden the impact of our sustainability efforts.

Built With

• fastapi
• next.js
• sqlite