AI-Powered Space Weather Prediction & Impact Assessment System Forecasting solar flares and geomagnetic storms to protect India's critical infrastructure
🚀 Live Demo • 📖 Documentation • 🔧 Installation • 🤝 Contributing
AetherGuard-INDIA is a machine learning-powered platform that predicts space weather events — solar flares and geomagnetic storms — and assesses their potential impact on satellites, power grids, communication networks, and other infrastructure. The system combines historical space weather data with a trained prediction model to deliver early warnings through a simple, accessible web interface.
Give India's infrastructure operators, researchers, and the public an early-warning system for space weather disruptions, so that satellite operators, grid operators, and aviation/communication services can prepare before an event hits.
- Machine learning model trained on historical space weather data (
space_weather_model.joblib) - Forecasts the likelihood and severity of solar flare / geomagnetic storm activity
- Continuously retrainable pipeline via
train_model.py
- Translates raw predictions into practical impact categories (
impact.py) — e.g. effects on satellites, power grids, GPS, and radio communications - Helps non-technical users understand what a prediction means rather than just a raw probability
- Lightweight Flask web app (
app.py) for running predictions and viewing results - Simple interface designed for quick checks without needing to run scripts manually
- Prediction history logged per session/user (
history_Space.csv,history_default_username.csv,prediction_history.csv) - Enables trend review and comparison of past forecasts against outcomes
- Basic user accounts (
users.txt) for tracking who ran which predictions - Subscriber list (
subscribers.txt) to support alert/notification distribution
Space Weather Data → Feature Preparation → Trained ML Model (predict.py) →
Risk/Impact Scoring (impact.py) → Web Dashboard (app.py) → History Log
- 🧠 Prediction Engine:
predict.py— loadsspace_weather_model.jobliband generates forecasts - 📉 Impact Engine:
impact.py— converts predictions into infrastructure impact assessments - 🏋️ Training Pipeline:
train_model.py— trains/retrains the space weather model - 🌐 Web Interface:
app.py— Flask-based dashboard for interacting with the system - 💾 Data Store: CSV-based history logs for predictions and user activity
- Python 3.8+
- pip package manager
- Git for version control
git clone https://github.com/hiralM001/Aetherlite-project.git
cd Aetherlite-projectpip install -r requirements.txtpython app.pyOpen your browser and navigate to:
- Main Dashboard: https://aetherlite-project-ehaceuxwfwyfhbdd7txgqh.streamlit.app/
- Open the dashboard in your browser after starting
app.py - Run a prediction using the current/latest space weather inputs
- Review results, including the predicted event likelihood and impact assessment
- Check history to compare against past predictions
# Run a prediction directly
python predict.py
# Assess impact of a given prediction
python impact.py
# Retrain the model on updated data
python train_model.pyAetherlite-project/
├── 📂 .devcontainer/ # Dev container configuration
├── 🌐 app.py # Flask web application (main entry point)
├── 🗄️ app_backup.py # Backup copy of the web application
├── 🧠 predict.py # Prediction script using the trained model
├── 📉 impact.py # Converts predictions into impact assessments
├── 🏋️ train_model.py # Model training / retraining script
├── 🤖 space_weather_model.joblib # Trained ML model artifact
├── 📊 history_Space.csv # Space weather prediction history
├── 📊 history_default_username.csv# Per-user prediction history
├── 📊 prediction_history.csv # General prediction log
├── 👤 users.txt # Registered users
├── 📧 subscribers.txt # Alert subscriber list
├── 📋 requirements.txt # Python dependencies
└── 📖 README.md # This file
- Anticipate solar activity that may affect satellite electronics or communications
- Early warning for geomagnetically induced currents that can stress transformers
- Study space weather trends using logged historical prediction data
- Experiment with retraining the model on new data
- Check current space weather risk through a simple web dashboard
# Clone repository
git clone https://github.com/hiralM001/Aetherlite-project.git
cd Aetherlite-project
# Create virtual environment
python -m venv venv
source venv/bin/activate # Linux/Mac
# venv\Scripts\activate # Windows
# Install dependencies
pip install -r requirements.txt
# Run the app
python app.pypython train_model.pyThis regenerates space_weather_model.joblib using the latest training data.
Contributions are welcome! Here's how to get involved:
- 🧠 Model Improvements: Improve prediction accuracy or add new features
- 📉 Impact Logic: Refine how predictions map to real-world impact categories
- 🎨 UI/UX: Improve the Flask dashboard
- 📚 Documentation: Help others understand and use the system
- Fork the repository
- Create a feature branch:
git checkout -b feature/amazing-feature - Commit changes:
git commit -m 'Add amazing feature' - Push to branch:
git push origin feature/amazing-feature - Submit a Pull Request with a detailed description
No license has been specified for this project yet. Consider adding one (e.g. MIT) so others know how they can use and contribute to it.
- 🐛 Bug Reports: GitHub Issues
- 💬 Discussions: GitHub Discussions
🌌 Forecasting the Skies to Protect What's on the Ground