Solar Sentinel

AI-Powered Space Weather Intelligence Platform for Solar Flare Nowcasting and Forecasting using Aditya-L1 SoLEXS and HEL1OS Observations

Solar Sentinel is an end-to-end AI-powered space weather intelligence platform that transforms Aditya-L1 multi-modal SoLEXS and HEL1OS observations into solar flare nowcasting, forecasting, mission-aware risk assessment, and actionable early warnings.

Overview

Solar flares can disrupt satellite operations, navigation systems, communication networks, and critical infrastructure. Existing approaches primarily focus on scientific observation rather than operational decision support.

Solar Sentinel bridges this gap by using AI to analyze multi-modal solar observations from Aditya-L1 and present the results through an interactive Mission Control Dashboard.

Key Features

• AI-powered solar flare nowcasting and forecasting
• Multi-modal fusion of SoLEXS and HEL1OS payload data
• 5-minute early warning system
• Mission-aware risk assessment
• Explainable AI using feature importance
• Interactive Mission Control Dashboard
• Recent soft and hard X-ray monitoring from processed data
• Decision support, mission logs, alerts, and system health views
• End-to-end AI deployment pipeline

Data Source

The project uses observations from ISRO's Aditya-L1 Mission.

Payloads:

• SoLEXS (Solar Low Energy X-ray Spectrometer): soft X-ray observations
• HEL1OS (High Energy L1 Orbiting X-ray Spectrometer): hard X-ray observations

These datasets are fused into synchronized time-series data for downstream AI analysis.

The GitHub repository includes a small data/features/sample_features.csv file so the backend and dashboard can run as a demo after cloning. Full raw, processed, labeled, and split datasets are generated locally and ignored by Git because they are large.

AI Pipeline

Aditya-L1 Observations
          |
          v
SoLEXS + HEL1OS Data Fusion
          |
          v
Data Ingestion and Synchronization
          |
          v
Feature Engineering
          |
          v
XGBoost AI Prediction Model
          |
          v
Risk Assessment
          |
          v
Decision Support
          |
          v
Mission Control Dashboard

Project Structure

Solar-Sentinel/
|
|-- backend/                 # FastAPI backend
|-- frontend/                # React + Vite frontend
|-- src/
|   |-- ingestion/
|   |-- preprocessing/
|   |-- feature_engineering/
|   |-- models/
|   |-- inference/
|   `-- analysis/
|
|-- data/
|   |-- raw/
|   |-- processed/
|   |-- labels/
|   |-- features/
|   |-- splits/
|   `-- models/
|
|-- notebooks/
|-- config.yaml
|-- requirements.txt
`-- README.md

Only lightweight runtime/demo data and model metadata are intended to be committed. Large generated datasets remain local.

Setup And Run

Install backend dependencies:

python -m venv venv
venv\Scripts\activate
pip install -r requirements.txt

Install frontend dependencies:

cd frontend
npm install

Start the backend from the project root:

uvicorn backend.app:app --reload

Start the frontend:

cd frontend
npm run dev

The frontend expects the backend at http://127.0.0.1:8000 by default. To use a different backend URL, create frontend/.env.local:

VITE_API_BASE_URL=http://127.0.0.1:8001

Optional checks:

python -m unittest discover -s tests
cd frontend
npm run lint
npm run build

Tech Stack

Frontend:

• React 19
• Vite
• React Router
• Axios
• Recharts

Backend:

• FastAPI
• Uvicorn

AI / Machine Learning:

• Python
• XGBoost
• Scikit-learn
• Pandas
• NumPy

Scientific Data Processing:

• Astropy
• FITS file processing

Model Information

Parameter	Value
Model	Tuned XGBoost
Prediction Window	Next 5 Minutes
Input Data	SoLEXS + HEL1OS
Output	Solar Flare Probability
Risk Levels	Very Low to Critical

Dashboard Modules

• Dashboard: mission overview with AI prediction summary and system status
• Data Monitoring: soft and hard X-ray activity visualization from processed data
• AI Prediction: probability gauge, flare prediction, timeline, and metrics
• AI Insights: feature importance and model interpretation
• Alerts: mission alerts, event logs, and recommendations
• System Health: backend status, model information, API status, and diagnostics

Problem Statement

Predicting solar flares with sufficient lead time is essential for protecting:

• Satellite operations
• Navigation systems
• Communication infrastructure
• Power grid operations
• Space missions

Solar Sentinel provides AI-driven early warnings to enable proactive operational decision-making.

Future Scope

• Live Aditya-L1 data integration
• Automated data ingestion pipeline
• Multi-horizon solar flare forecasting
• CME prediction
• Geomagnetic storm prediction
• Cloud deployment
• Mobile and web notifications
• Integration with space weather monitoring systems

Future Deployment And Adoption

• Government and research institution integration
• Space agency and satellite operator collaboration
• Space Weather Intelligence API services
• Enterprise deployment and technical support
• Cloud and on-premise scalable architecture

Application Preview

Dashboard:

Data Monitoring:

AI Prediction:

AI Insights:

Alerts:

System Health:

Vision

Solar Sentinel transforms Aditya-L1 multi-modal SoLEXS and HEL1OS observations into AI-powered space weather intelligence through solar flare nowcasting, forecasting, risk assessment, and intelligent early warnings.