Reuse — AI Chatbot for Smart Recycling

♻️ Reuse — AI Chatbot for Smart Recycling

📌 Overview

Reuse is an AI-powered chatbot that helps users transform everyday waste into reusable objects.
It combines image recognition, machine learning, and conversational AI to provide personalized reuse ideas and sustainability guidance.

Type: Competition Project

Duration: November 2018 - May 2019

Competition: Atos IT Challenge 2019 - Machine Learning for Sustainability

Achievement: Finalist


🎯 Project Goal

Reuse supports the transition toward a circular economy by helping people:

  • Identify objects using AI
  • Receive personalized reuse ideas
  • Learn how to transform items step-by-step
  • Share results with a community

The mission is to make reuse easy, intelligent, and engaging.


🤖 Key Features

Image Recognition

  • Detects objects from user-uploaded photos
  • Implemented using YOLO real-time object detection
  • Predicts bounding boxes and object probabilities

Conversational Chatbot

  • Built with Dialogflow NLP
  • Understands user intent and preferences
  • Suggests reuse ideas interactively

Personalized Recommendation Engine

  • Uses association rule mining
  • Learns from user behavior and feedback
  • Improves suggestion relevance over time

Recycling Guidance

  • If an item cannot be reused
  • The chatbot suggests appropriate recycling options

Community Concept

  • Users share reused items
  • Provide feedback, likes, and suggestions
  • Inspire sustainable habits

🧱 Technical Architecture

Workflow

  1. User sends text or image
  2. Dialogflow extracts intent/entities
  3. Backend Python server processes request
  4. Image recognition and recommendation logic run
  5. Suggestion is returned to the chatbot

Main Components

  • Chatbot Interface (Messenger/Web)
  • Dialogflow NLP
  • Python Backend
  • YOLO Detection Module
  • Recommendation Engine
  • Database and APIs

🛠 Technology Stack

AI / Machine Learning

  • YOLO Object Detection
  • Association Rule Mining
  • Hyperparameter tuning (Grid Search)

Backend

  • Python
  • JSON API communication
  • NGROK for local server exposure

NLP & Integration

  • Google Dialogflow
  • Facebook Messenger API

Environment

  • Linux / Anaconda
  • Python 2.7 (prototype stage)

GitHub Repository

https://github.com/Karim-53/Reuse-Chatbot-Atos-Challenge-2019

Project Demo Video

https://www.youtube.com/watch?v=2Intw22lcmo

Atos Challenge Finalist Video

https://www.youtube.com/watch?v=esOe2S2eJ68


🏆 Competition

  • Event: Atos IT Challenge 2019
  • Achievement: Finalist Team
  • Supervisor: Dr. Thomas Hoberg

👨‍💻 Team

  • Mohamed Karim Belaid
  • Aicha Garci
  • Eya Cherif

🌍 Impact

Reuse contributes to:

  • Reducing landfill waste
  • Increasing resource efficiency
  • Encouraging sustainable habits
  • Promoting circular economy practices
  • Building environmental awareness