The study was co-authored with Doron Shavit and Merav Davidovits, graduate students of intelligent systems at Afeka, in collaboration with Israeli Air Force personnel.

ABSTRACT: The article summarizes a study on predictive maintenance. Predictive maintenance enables to monitor the status of a machine or equipment and detect the need for maintenance actions. This approach allows for cost reduction and for early identification of failures and malfunctions.

The published study focuses on the early detection of faults in aircraft steering systems, based on operational data from multiple flights. Researchers developed an algorithm to detect causal links between multiple aircraft sensor data, and to find the important sensors for early fault detection, based on causal link analysis of time series. The study results show that use of alert systems based on the algorithm allows for effective maintenance and may help to prevent disaster.

To read the article:

Temporal causality-based feature selection for fault prediction in rotorcraft flight controls