Atmos – Hybrid Air Quality Monitoring

Overview

Atmos is a Python + FastAPI project that simulates and optimizes air quality sensor placement across the city. The system combines real-time WAQI data, traffic-weighted zone analysis, and optimization algorithms to propose an efficient sensor network that is cheaper and more effective than the existing CAAQMS stations.

Key features:

• Fetch live PM2.5 and AQI readings for Bengaluru from WAQI API.
• Generate fallback realistic data if API fails.
• Hybrid model optimization:
  • Bounding Phase Method
  • Golden Ratio Search
• Proportional sensor allocation based on pollution and traffic.
• Interactive web interface (FastAPI + Jinja2 templates):
  • Home map showing sensor coverage and PM2.5 per zone.
  • Compare current government network vs optimized hybrid model.
  • Query AQI per neighborhood.
  • Statistics and zone-wise coverage.
  • Educational info about PM2.5 and air quality safety.

---

Project Structure

hybrid/
├── backend/
│   ├── __init__.py
│   ├── main.py            # FastAPI server
│   ├── optimize.py        # Hybrid sensor optimization
│   └── waqi_service.py    # WAQI API integration & fallback
│
├── frontend/
│   ├── templates/         # HTML templates
│   └── static/            # CSS/JS assets
│
├── data/                  # Generated hybrid_model.json
├── venv1/                 # Virtual environment
└── README.md

---

Requirements

• Python 3.13+
• FastAPI
• Uvicorn
• Requests
• Numpy

Install dependencies:

source venv1/bin/activate
pip install fastapi uvicorn requests numpy jinja2

---

Setup

1. Clone the repository to your machine:

   git clone <repo-url>
   cd hybrid

2. Activate virtual environment:

   source venv1/bin/activate

3. Set your WAQI API key:

   export WAQI_API_KEY="YOUR_API_KEY_HERE"

   Get your WAQI token at: https://aqicn.org/data-platform/token/

4. Ensure backend/__init__.py exists (empty is fine).

---

Run the Server

python -m uvicorn backend.main:app --reload

• Open your browser at http://127.0.0.1:8000 to access the app.

---

Testing Imports (Optional)

To verify that imports work correctly:

python -c "from backend.optimize import optimize_hybrid_model; print('optimize works!')"
python -c "from backend.waqi_service import fetch_live_aqi; print('WAQI works!')"

---

API Endpoints

Endpoint	Method	Description
/api/model	GET	Returns current hybrid sensor allocation JSON
/api/waqi/live	GET	Fetches live AQI data from WAQI
/api/zone/{zone_name}	GET	Returns data for a specific zone
/api/optimize	POST	Trigger a new hybrid optimization run
/api/stats	GET	Returns statistics (total sensors, average AQI, worst/best zone)

---

Frontend Pages

• / → Home page with interactive map of sensors and PM2.5 per zone.
• /compare → Compare hybrid model vs existing 14 CAAQMS stations.
• /coverage → Visual representation of sensor coverage.
• /query → Input zone name to get live AQI.
• /info → PM2.5 explanation, health advice, and safety measures.
• /contrast → Compare the worst and best polluted zones.

---

Notes

• Hybrid model uses mid-grade sensors (~₹30k each) to cover more areas efficiently under a budget of ₹25 lakh.
• Fallback data is generated when WAQI API is unavailable.
• The project is designed as a demonstration/prototype; it is not deployed to production.

---

Author

MD – Engineering student