🌍 EnviroScan AI

EnviroScan AI is a proof-of-concept platform that demonstrates the power of machine learning in environmental science. It provides actionable insights into the sources of air pollution by analyzing real-time air quality data, weather patterns, and geospatial information.

🚀 Live Demo

Check out the deployed app here:
EnviroScan AI Web App

1016.7.mp4

📖 Project Overview

EnviroScan AI aims to go beyond measuring pollution levels by attributing pollution events to specific sources such as:

• Vehicular traffic
• Industrial activity
• Agricultural burning

Target Users:

• Urban Planners & Policymakers
• Environmental Agencies
• Researchers & Students

---

🔬 Methodology

1. Data Pipeline

• Ingestion: Collect air quality metrics (PM₂.₅, NO₂, SO₂), meteorological data, and geospatial features.
• Preprocessing: Clean data, handle missing values, and create time-based features.
• Labeling: Rule-based engine assigns preliminary pollution_source labels for supervised learning.
• Train/Test Split: Data is split into training and testing sets for evaluation.

2. Model Training & Evaluation

• Algorithm: Ensemble Voting Classifier combining:
  • Random Forest Classifier (RF)
  • XGBoost Classifier (XGB)
• Why Ensemble Voting?
  • Combines strengths of multiple models for higher accuracy
  • Reduces overfitting compared to single models
  • Handles non-linear relationships and complex feature interactions
  • Supports multi-class evaluation with classification metrics

Evaluation Metrics:

• Classification report (precision, recall, F1-score for each pollution source class)
• Confusion matrix (visual analysis of predictions)

Model Output:

• Ensemble model saved as outputs/ensemble_model.joblib
• Label encoder saved as config.ENCODER_FILE
• Evaluation report saved as outputs/ensemble_model_report.txt

---

💻 Technology Stack

• Python: Core backend language
• Streamlit: Web app and dashboard
• Scikit-learn: Machine learning models
• Pandas: Data processing and analysis
• Folium & Streamlit-Folium: Interactive geospatial maps
• Plotly & Seaborn: Charts and visualizations
• Joblib: Model serialization
• Python-dotenv: Managing API keys

---

⚙️ How to Run This Project

Windows

# 1. Clone repository
git clone https://github.com/Supriyo760/enviroscan.Ai
cd EnviroScan-Project

# 2. Setup environment
=======
How to Run this Project:- 
Step by Step commands :-
---------------------------------
Windows:-
:: 1. Clone repository
git clone https://github.com/YOUR_USERNAME/YOUR_REPOSITORY_NAME.git
cd YOUR_REPOSITORY_NAME

:: 2. Setup environment
python -m venv venv
venv\Scripts\activate
pip install -r requirements.txt

# 3. Add API key
notepad .env
# Add this line:
OPENWEATHER_API_KEY=your_key_here

# 4. Run data pipeline
python run_pipeline.py

# 5. Start Streamlit app
streamlit run app.py

MacOS / Linux

# 1. Clone repository
git clone https://github.com/Supriyo760/enviroscan.Ai
cd EnviroScan-Project
=======
:: 3. Add API key
notepad .env
:: Add this line: OPENWEATHER_API_KEY=your_key_here

:: 4. Run pipeline
python3 run_pipeline.py

:: 5. Start app
streamlit run app.py
-------------------------------------------------------------------------
 Mac os users:-
# 1. Clone repository
git clone https://github.com/YOUR_USERNAME/YOUR_REPOSITORY_NAME.git
cd YOUR_REPOSITORY_NAME

# 2. Setup environment
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

# 3. Add API key
echo "OPENWEATHER_API_KEY=your_key_here" > .env

<<<<<<< HEAD
# 4. Run data pipeline
python3 run_pipeline.py

# 5. Start Streamlit app
streamlit run app.py

Troubleshooting Dependencies

If some packages fail to install, try installing them individually:

pip install streamlit pandas scikit-learn xgboost
pip install folium streamlit-folium plotly
pip install requests osmnx python-dotenv joblib

---

📂 Repository Structure

EnviroScan-AI-Powered-Team-6/
│
├── app.py                       # Main Streamlit app entry point
├── run_pipeline.py              # Script to run the entire data pipeline
├── EnviroScan.py                # Possibly main project logic / module
├── train_model.py               # Script for training the ML model
├── config.py                    # Configuration file for constants, paths, etc.
├── requirements.txt             # Project dependencies
├── .env                         # Environment variables (API keys)
├── enviroscan.log               # Log file for the project
├── How_to_run_program_file       # Instructions / script guide
├── my_script_code.txt           # Any additional code snippets
├── README.md                    # Project documentation
│
├── pages/                       # Streamlit multi-page app scripts
│   ├── About.py
│   ├── Live_Prediction.py
│   ├── Pollution_Dashboard.py
│   └── image.png                # Image used in About page
│
├── scripts/                     # Scripts for data pipeline and ML preprocessing
│   ├── 01_data_collection.py
│   ├── 02_data_processing.py
│   ├── 03_label_and_split.py
│   ├── 04_train_model.py
│   └── __pycache__/             # Python cache folder
│
├── venv/                        # Python virtual environment
│
├── data/ (optional, if exists)  # Raw / processed datasets
│   └── locations.csv
│
├── outputs/                     # Any output files, models, or results
│
└── cache/                       # Streamlit or project cache files

---

📝 License

This project is open-source. Feel free to use, modify, and share under the MIT License.

---

=======

4. Run pipeline

python3 scripts/run_pipeline.py

5. Start app

streamlit run app.py

others:-

If dependencies fail to install:

Try installing individually

pip install streamlit pandas scikit-learn xgboost pip install folium streamlit-folium plotly pip install requests osmnx python-dotenv joblib