RecyScore: An IoT-based Recycling Reward System

Click on the picture to view our demo video

---

Project Overview

• RecyScore is an innovative IoT-based recycling reward system that uses Large Language Model (LLM) to identify and classify recyclable items.
• The system provides real-time feedback and rewards users for proper recycling practices, encouraging sustainable waste management.
• Feel free to scroll down the README or read the slides or 12-minute-presentation.
• Feel free to open our website or with the demo account: vaclis, password: vaclis

Key Features

• Real-time object capturing and uploading using Raspberry Pi camera and official tablet
• AWS IoT Core integration for device status management on web
• AI-powered image analysis using OpenAI's LLM model
• Web-based dashboard for tracking recycling statistics
• Reward system for encouraging proper recycling practices

System Architecture

Team Members

• Song-Ze, Yu (@vaclisinc): Raspberry Pi camera capturing and GUI, AWS ioT core and device shadow, website function writing, slide making and video editing.
• Koying (@koyingtw): LLM Lambda function, AWS DynamoDB, AWS API Gateway, S3 static web hosting.
• Daniel Lin (@trkk28097402): ioT core SDK installing, slide making.

Getting Started

1. Clone the repository
2. Set up your AWS credentials and environment variables (see example_env.txt)
3. Install required dependencies
4. Run the GUI on Raspberry Pi using run.sh (You may need to modfiy the python env directory.)

Project Structure

RecyScore/
├── pi/               # Raspberry Pi related code
├── lambda/           # AWS Lambda functions
├── web/              # Web interface files
├── policy/           # AWS IAM policies
└── run.sh            # Startup script

Technologies Used

• Python
• AWS IoT Core
• AWS Lambda
• AWS DynamoDB
• AWS API Gateway
• OpenAI API
• Raspberry Pi
• HTML/CSS/JavaScript