Skip to content
Iftekhar Rafi
Back to projects
Machine LearningExtended multi-phase developmentOpen-source contribution

Aircraft Health Monitoring System with Machine Learning

AircraftHealthML is a machine learning system for real-time aircraft health monitoring using NASA aviation datasets.

Preview of Aircraft Health Monitoring System with Machine Learning

What it was

AircraftHealthML focused on early fault detection and predictive maintenance using NASA ADAPT and NGAFID aviation datasets.

The project framed ML as part of a larger system: data ingestion, preprocessing, model inference, and operator-facing visualization all had to work together.

Technical approach

  • Combined a transformer-based neural network for temporal pattern recognition with an isolation forest for anomaly detection.
  • Used Django and Django REST Framework for the backend, React for the frontend, and Plotly.js for interactive dashboards.
  • Built the data pipeline with Pandas and feature engineering steps tailored to aviation telemetry and maintenance signals.

Outcome

This project shows that model work is most useful when paired with data infrastructure and operator-facing tooling.