As the latest generation of business jets enter into service, their operators could reap the benefits of predictive maintenance--as the airlines have done for years.
Unlike preventative maintenance, which is time-based, predictive maintenance uses data generated by various component and system sensors that monitor the condition of the equipment at any given time to predict maintenance needs going forward. With that knowledge in hand, the operator can schedule maintenance well before a failure takes place and, in effect, turn a non-scheduled event into a scheduled maintenance task.
“The mechanism of prediction comes into play with complex and interdependent aircraft systems with a high number of mechanical components and pneumatic connections,” explains Mia Witzig, head of digital solutions at Lufthansa Technik. “A time-based approach works out well for single components without bigger interdependencies when you have sensors reflecting their behavior. Whenever it gets more complex, that's the point where predictive algorithms help, and big data makes the difference.”
Witzig notes that while predictive maintenance depends on the capability of aircraft to deliver the required data, acquiring it depends on recording intervals--not just the number of sensors on the aircraft. Many aircraft, she notes, send standard reports to the ground with “snapshot data” recorded from a specific sensor, in a certain flight phase.
“That’s perfect for monitoring a specific component, but, if you want to recognize trends and derivations on more complex systems, you need a data series covering a complete flight phase or even the entire flight,” Witzig stresses. “With full flight data coming from the quick access recorder [QAR], for example, we can build a digital twin of the component within the digital twin of the aircraft and review the data with prediction models.”
As a case in point, Witzig cites the Airbus A320’s bleed air system, which, she says, has generated “a significantly high number of operational incidents and pilot reports.” The system, she describes, is complex with a huge number of subcomponents, and a high number of possible failure modes--requiring a major troubleshooting effort.
“Our EBASS [Engine Bleed System Suite] predictor monitors all the subsystems and sensors and brings the data back together in a digital model of the bleed system,” says Witzig. “This enables us to identify the root cause--data based--and decrease the no-fault-found [NFF] rate in the MRO shop significantly.” As a result, Lufthansa Technik’s A320 customers saw a 50% decrease in AOGs and a 30% reduction in engine removal rates.
The second part of this article will look at how business aviation is using predictive maintenance.