Water Quality Monitoring API

A Flask-based REST API for real-time water quality monitoring, alerting, and prediction. Integrates MongoDB for data storage, Firebase Cloud Messaging and Slack for notifications, and a machine learning model for water safety prediction.

---

Features

• User Registration & Login: Register and authenticate users with FCM tokens for push notifications.
• Sensor Data Ingestion: Accepts water quality sensor data (pH, turbidity, temperature, TDS).
• Data Storage: Persists all data in MongoDB.
• Real-Time Alerts: Sends alerts via Firebase and Slack if water quality is unsafe.
• Machine Learning Prediction: Predicts water safety using a Random Forest classifier.
• RESTful API: Provides endpoints for data submission, retrieval, and prediction.

---

Environment Setup

1. Clone the repository and install dependencies:

   pip install -r requirements.txt

2. Configure environment variables in a .env file:

   MONGO_URI=your_mongodb_uri
   FIREBASE_PROJECT_ID=...
   FIREBASE_PRIVATE_KEY_ID=...
   FIREBASE_PRIVATE_KEY=...
   FIREBASE_CLIENT_EMAIL=...
   FIREBASE_CLIENT_ID=...
   FIREBASE_AUTH_URI=...
   FIREBASE_TOKEN_URI=...
   FIREBASE_AUTH_PROVIDER_CERT_URL=...
   FIREBASE_CLIENT_CERT_URL=...
   FIREBASE_UNIVERSE_DOMAIN=...
   SLACK_WEBHOOK_URL=your_slack_webhook_url
   PORT=5000
   

3. Run the application:

   python app.py

---

API Endpoints

GET /

• Description: Health check endpoint.
• Response: "Welcome to the Water Quality Monitoring API!"

---

POST /api/register

• Description: Register a new user.

• Request Body:

  {
    "username": "user1",
    "password": "pass",
    "fcmtoken": "firebase_token"
  }

• Response: User info or error.

---

POST /api/login

• Description: Authenticate a user.

• Request Body:

  {
    "username": "user1",
    "password": "pass"
  }

• Response: User info or error.

---

POST /api/sensor_data

• Description: Submit new sensor data.

• Request Body:

  {
    "pH": 7.2,
    "turbidity": 2.5,
    "temp": 25.0,
    "tds": 500
  }

• Response: Status, prediction, confidence, and alerts.

---

GET /api/sensor_data

• Description: Retrieve the 50 most recent sensor data records.
• Response: List of sensor data.

---

GET /api/predict

• Description: Get prediction and alerts for the latest sensor data.
• Response: Water quality status, confidence, and alerts.

---

Machine Learning Model

• Algorithm: Random Forest Classifier (sklearn)
• Features: pH, turbidity, temperature, TDS
• Label: Unsafe (1) if any parameter is out of WHO range, else Safe (0)
• Training: Retrained every 10 new records (if at least 10 records exist)
• Prediction: Returns "Safe" or "Unsafe" with confidence score

---

Notifications

• Firebase: Sends push notifications to the first registered user's FCM token.
• Slack: Sends alerts to a configured Slack channel via webhook.
• Trigger: Alerts are sent if any parameter is out of the safe range.

---

WHO Water Quality Standards

Parameter	Safe Range
pH	6.5 - 8.5
Turbidity	≤ 5.0 NTU
Temperature	≤ 30.0 °C
TDS	≤ 1000.0 ppm

---

Code Structure

• Flask App Initialization: Sets up Flask, logging, and loads environment variables.
• MongoDB Connection: Connects to MongoDB (cloud/local) and initializes collections.
• Firebase Initialization: Loads credentials and initializes Firebase Admin SDK.
• Notification Functions: Functions to send alerts via Firebase and Slack.
• Water Quality Check: Checks sensor data against WHO standards and triggers notifications.
• Model Training: Trains a Random Forest model on historical data.
• API Endpoints: Implements all REST endpoints for registration, login, data submission, retrieval, and prediction.
• Shutdown Hook: Closes MongoDB connection gracefully on shutdown.

---

Shutdown & Cleanup

• The application registers a shutdown hook to close the MongoDB connection when the app exits.

---

License

This project is licensed under the MIT License.

---

Note:

• Ensure all environment variables are set correctly for MongoDB, Firebase, and Slack integration.
• The API is designed for demonstration and prototyping; for production, enhance security (e.g., password hashing, authentication, validation).