SMARTBIN - Smart Waste Management System

🗑️ An intelligent waste management solution using IoT sensors, machine learning, and route optimization.

Features

• Real-time Monitoring: Track fill levels across all smart bins
• Waste Classification: AI-powered image and audio classification
• Route Optimization: Efficient collection routes using OR-Tools
• Anomaly Detection: Identify illegal dumping and sensor malfunctions
• CO2 Tracking: Monitor environmental impact and sustainability metrics

Live Demo

Quick Start

# Clone the repository
git clone https://github.com/yourusername/smartbin.git
cd smartbin

# Create virtual environment
python -m venv venv
source venv/bin/activate  # Linux/Mac
# or: venv\Scripts\activate  # Windows

# Install dependencies
pip install -r requirements.txt

# Run the dashboard
streamlit run app.py

Project Structure

SMARTBIN/
├── app.py                    # Streamlit dashboard
├── requirements.txt          # Python dependencies
├── .streamlit/
│   └── config.toml          # Streamlit configuration
├── simulate_bin_events.py   # IoT event simulator
├── waste_classifier.py      # MobileNetV2 image classifier
├── audio_pipeline.py        # Mel-spectrogram extraction
├── audio_classifier.py      # CNN audio classifier
├── multimodal_fusion.py     # Image+audio fusion model
├── fill_level_predictor.py  # LSTM fill prediction
├── anomaly_detector.py      # Anomaly detection system
├── route_optimizer.py       # Collection route optimization
├── data/
│   ├── smart_bin_events.csv
│   ├── detected_anomalies.json
│   └── route_data.json
└── models/
    └── [trained models]

Components

1. IoT Simulator

Generates realistic smart bin events with fill levels, waste types, and sensor data.

2. Waste Classification

• Image: MobileNetV2 transfer learning (5 classes)
• Audio: CNN with Mel-spectrograms
• Fusion: Late fusion combining both modalities

3. Fill Level Prediction

LSTM model predicting bin overflow 12 hours ahead based on historical patterns.

4. Anomaly Detection

Rule-based and statistical detection of:

• Illegal dumping (fill jumps >15%)
• Sensor malfunction
• Audio-image mismatch

5. Route Optimization

• Greedy nearest-neighbor algorithm
• OR-Tools VRP solver
• Interactive map visualization

Deployment

Streamlit Cloud

1. Push code to GitHub
2. Go to share.streamlit.io
3. Connect your repository
4. Set main file: app.py
5. Deploy!

Docker

FROM python:3.11-slim
WORKDIR /app
COPY requirements.txt .
RUN pip install -r requirements.txt
COPY . .
EXPOSE 8501
CMD ["streamlit", "run", "app.py"]

API Reference

Anomaly Detector

from anomaly_detector import AnomalyDetector

detector = AnomalyDetector()
anomalies = detector.detect_anomalies({
    "bin_id": "BIN_001",
    "fill_level": 95,
    "timestamp": "2026-01-05T10:30:00"
})

Route Optimizer

from route_optimizer import RouteOptimizer, Bin

optimizer = RouteOptimizer()
bins = [Bin("BIN_001", 12.97, 77.59, 85)]
route = optimizer.optimize_route(bins)
print(f"Distance: {route.total_distance_km} km")

Environment Variables

Variable	Description	Default
DEPOT_LAT	Depot latitude	12.9716
DEPOT_LON	Depot longitude	77.5946
FILL_THRESHOLD	Collection threshold	60

Contributing

1. Fork the repository
2. Create a feature branch
3. Commit changes
4. Push to branch
5. Open a Pull Request

License

MIT License - see LICENSE for details.

Acknowledgments

• TrashNet dataset for waste images
• OR-Tools by Google for optimization
• Streamlit for the dashboard framework