PyroSense is an advanced wildfire intelligence platform designed to fuse satellite telemetry with machine learning to provide real-time risk scoring for rapid response teams.
PyroSense transforms raw satellite data into actionable risk intelligence. By utilizing real-time detections from NASA's Fire Information for Resource Management System (FIRMS), the engine performs inference to help teams assess and respond to fire threats before they spread.
- Machine Learning: XGBoost model, trained for real-time risk assessment based on weather, fuel, and terrain data.
- Backend: FastAPI (Python) for high-performance API services and model inference.
- Frontend: React.js for a responsive, high-fidelity intelligence dashboard.
- Data Ingestion: Real-time integration with NASA FIRMS API.
- Real-time Hotspot Mapping: Interactive visualization of live satellite detections.
- ML Risk Scoring: Automated risk assessment using an XGBoost pipeline to weigh environmental and geographical variables.
- Micro-Climate Analytics: Real-time environmental metrics (Temperature, Humidity, VPD) to model fuel dryness and fire behavior.
- Incident Logging: Streamlined field incident reporting synced directly with command dashboards.
The platform is built to provide an end-to-end intelligence pipeline:
- Ingestion: Fetching live telemetry from NASA FIRMS.
- Inference: Processing features through the XGBoost engine via FastAPI to generate accurate risk scores.
- Visualization: A React-based interface allowing field analysts to define geographic bounding boxes for specific, actionable intelligence.