Huge time series datasets are constantly being recorded from sensors in numerous (sub)systems on the aircraft. In this talk we show how deep learning can predict normal behavior of particular sensors and discuss its usage in an automated semi-supervised anomaly detection system.
Key Takeaways
- Deep learning for contextual predictions of time series sensor data
- Generation of synthetic sensor data
- Industrial viability of automated anomaly detection systems
————————————————————————————————————————————————————
Bio
Dr. Sarah Andreas – Data Scientist | Airbus
Dr. Sarah Andreas is a Data Scientist in the Airbus Digital Transformation Office. Her work focuses on applying AI to time series use cases, in particular for anomaly detection.
Sarah has more than five years of hands-on Data Science and Machine learning expertise across various industries. She also holds a PhD in theoretical physics from the University of Hamburg and did research in the field of theoretical particle physics.
Virtual Program
Dr. Sarah Andreas – Data Scientist | Airbus