Home - CankayaUniversity/ceng-407-408-2024-2025-ReSort GitHub Wiki

Welcome to the ceng-407-408-2024-2025-ReSort wiki!

ReSort: Intelligent Waste Sorting and Recycling System

ReSort is an innovative recycling solution leveraging IoT and machine learning to automate waste sorting processes. By combining sensors, cameras, and cutting-edge algorithms, ReSort enhances recycling efficiency, reduces environmental impact, and provides real-time feedback to users.

📄 Documents

♻️ About ReSort

ReSort aims to address inefficiencies in waste management by automating the classification and sorting of recyclable materials. The system serves households, municipalities, and industrial users, providing:

  • Accurate Waste Classification: Categorizing waste into types like plastic, metal, and organic with over 90% accuracy.
  • Real-Time Feedback: Delivering insights on recycling practices and environmental impact.
  • Scalable Design: Adaptable for various settings, from small households to large municipalities.

Our mission is to revolutionize recycling by increasing efficiency, reducing human error, and promoting sustainable practices.

🔑 Key Features

  • IoT Integration: Sensors detect weight, material type, and other properties of waste.
  • Machine Learning Models: Utilizes YOLO algorithm for high-accuracy waste classification.
  • User-Friendly Interfaces:
    • Real-time feedback via dashboards.
    • Recommendations for improving recycling habits.
  • Environmental Analytics: Tracks metrics like carbon emission reduction and resource conservation.
  • Agile Development: Built using the Agile Scrum methodology for iterative improvements.

🏩 System Architecture

ReSort comprises five main modules:

  1. User Interface (UI):

    • Intuitive dashboards for users to interact with the system.
    • Panels for progress tracking, feedback, and waste analysis.

  2. Waste Environment Design:

    • Physical and virtual spaces for waste sorting.
    • 3D representations of waste and IoT components.

  3. Data and Feedback:

    • Real-time waste classification results.
    • Environmental impact analytics for user engagement.

  4. Sensors and Machine Learning:

    • IoT-enabled sensors for material detection.
    • YOLO model for image-based classification.

  5. System Analytics:

    • Aggregates and analyzes user data.
    • Generates detailed reports on recycling trends.

🔧 Technologies Used

  • Programming Languages: Python, React, Node.js, C++
  • Hardware: Raspberry Pi, IoT sensors, cameras
  • Machine Learning Models: YOLO
  • 3D Modeling: Shapr3D
  • Database: PostgreSQL
  • API: WebSocket
  • Tools: LabelImg, Label Studio, Roboflow, Ultralytics HUB, Google Colab

👥 Team Members

  • Ahmet Eren BOSTAN - 202011018
  • Emre Can ERKUL - 202011007
  • Fatih Cumhur ÖĞÜTÇÜ - 202011004
  • Fırat Can AĞA - 202011080
  • Ömer ELMAS - 202011209

🧑‍🏫 Advisor

  • Dr. Abdül Kadir GÖRÜR