CFM International has expanded its engine health monitoring capabilities with machine learning, to provide operators with more insights. The system now being used for CFM LEAP-1A and LEAP-1B engines models data from multiple engine sensors at takeoff, climb, and cruise via probabilistic diagnostic and prognostic machine learning tools. The company says these tools then provide targeted alerts based on known engine operating signatures. The operational models being generated are the most accurate the company has seen to date and are resulting in earlier detection of potential issues.
The system has been used extensively for more than a decade on CFM56 and other engines and, thus, benefits from the experience of a wide portfolio of products, a quality dataset, and deep domain expertise. The CFM LEAP engine is helping advance the state-of-the art in predictive maintenance.
“These state-of-the-art analytics are providing LEAP operators with the data they need to make informed, insightful decisions about fleet management,” said Agathe Venard, head of Fleet Data Engineering for CFM parent company Safran Aircraft Engines. “As a result, they gain a level of fleet stability they can rely on in their day-to-day operations.”
“With the combination of this health monitoring system and the expertise of our global CFM fleet support team, we have achieved 60% earlier lead time in identifying preventative maintenance recommendations, a 45% increase in detection rates and cut the number of false alerts in half over the past decade,” said David Harper, fleet support director for CFM parent company GE Aerospace.