EcoMinds AI is an AI-powered web platform designed to promote sustainable electronic waste management through intelligent automation, cloud technologies, and interactive recycling workflows.
This project was developed during my second-year research internship at MANIT Bhopal (Maulana Azad National Institute of Technology, Bhopal) as an experimental full-stack sustainability-focused system integrating Artificial Intelligence, Firebase, Machine Learning, and responsive web technologies.
The platform aims to simplify e-waste identification, promote responsible disposal behavior, and encourage environmental awareness using AI-driven tools and gamified user engagement.
Electronic waste (e-waste) is one of the fastest-growing environmental challenges worldwide. Improper disposal of electronic devices contributes to:
- soil and water pollution,
- toxic chemical exposure,
- inefficient recycling processes,
- and growing sustainability concerns.
EcoMinds AI was designed to explore how AI and cloud technologies can help users:
- identify recyclable electronic waste,
- understand proper disposal methods,
- interact with recycling workflows,
- and participate in environmentally responsible practices.
- Firebase Authentication based login/signup system
- Secure user session management
- Firestore-integrated user data handling
- TensorFlow.js powered image recognition
- Teachable Machine integration
- MobileNet experimentation
- Real-time e-waste category prediction from uploaded images
- Users can submit e-waste pickup requests
- Pickup data stored in Firebase Firestore
- Recycler-side workflow prototype implemented
- Recycler-side interaction system
- Pickup request acceptance/rejection simulation
- Sustainability workflow demonstration
- Eco-points and reward concepts
- Sustainability engagement mechanisms
- Static leaderboard prototype
- Social challenge workflow concepts
- Admin-oriented monitoring interface
- Pickup overview concepts
- User activity visualization
- Dashboard workflow simulations
- Leaflet.js based map integration
- Recycling center exploration concept
- Sustainability education pages
- Environmental awareness content
- Learning and quiz module concepts
- AI-powered chatbot integration
- Guidance and awareness assistance system
- Originally implemented using external API services
- Mobile-friendly responsive design
- Multi-page web application architecture
- Interactive UI workflows
- HTML5
- CSS3
- JavaScript
- Firebase Hosting
- Firebase Authentication
- Firebase Firestore
- Firebase Realtime Database
- TensorFlow.js
- Google Teachable Machine
- MobileNet
- Leaflet.js
- Firebase Hosting
User Uploads E-Waste Image
↓
TensorFlow.js Processes Image
↓
AI Model Predicts E-Waste Category
↓
User Receives Classification Result
↓
Pickup Scheduling & Recycling Workflow
↓
Reward & Awareness Features
EcoMinds-AI/
│
├── public/
│ ├── css/
│ ├── js/
│ ├── assets/
│ └── *.html
│
├── docs/
│
├── firebase.json
├── firestore.rules
├── firestore.indexes.json
├── database.rules.json
│
├── e_waste_cnn_model.h5
└── train_model_with_mobilenet.py
- AI-integrated sustainability-focused web platform
- Client-side machine learning deployment using TensorFlow.js
- Cloud-integrated authentication and database workflows
- Experimental recycler-management ecosystem
- Environmental awareness and gamification integration
- Responsive multi-page web architecture
The AI classification system demonstrated promising early-stage performance during testing for common e-waste categories such as:
- mobile phones,
- batteries,
- chargers,
- and circuit components.
The responsive UI and Firebase integration enabled smooth interaction across multiple workflows including authentication, scheduling, and reward simulation systems.
Since this project was originally developed as a second-year research internship prototype, some modules may no longer function exactly as originally intended.
Examples:
- The chatbot service may currently be unavailable due to expired API/free-tier limitations.
- Some dashboards and leaderboard systems are prototype/static implementations.
- Multilingual support was planned but not fully implemented.
- The AI model can still be improved using larger and more diverse datasets.
These limitations are part of the experimentation and learning process involved in developing large integrated systems during early-stage engineering exploration.
Potential future enhancements include:
- Improved AI model accuracy
- Larger and more diverse datasets
- Real-time recycler integration
- Pickup tracking system enhancements
- Full multilingual support
- Enhanced gamification
- Carbon footprint analytics
- Advanced admin analytics dashboard
- Mobile application support
Project screenshots are available inside:
docs/
Through this project, I explored:
- Full-stack web development
- AI deployment in browser environments
- Firebase ecosystem integration
- Client-side machine learning
- Responsive UI/UX workflows
- Sustainability-focused software engineering
- Research-oriented system design
- Git/GitHub workflow and deployment practices
Yashti Choudhary
This project represents one of my early large-scale AI-integrated software systems developed during my second-year research internship and reflects my exploration into intelligent sustainability-focused web applications.
This project was developed as part of a second-year research internship and is shared for educational, learning, and portfolio purposes.