Big Data MRO.jpg GE Aviation

Analytics 1.0, 2.0 and 3.0

There could be much more to do with Big Data.

The issue of Big Data in MRO is a contentious one but nevertheless airlines are recognizing its potential and the benefits it can bring.

Mike Harris, manager of engineering technology & innovation at Jetstar, sees applying Big Data analytics to improve maintenance as happening in three stages. He calls them version 1.0 for predictive maintenance, version 2.0 for machine learning and version 3.0 for digital twinning.

Version 1.0 is happening now at many airlines. Version 2.0 is beginning at Jetstar and others as engineers begin to look at machine learning for what Harris calls “knowledge capture.”

Harris points out that the looming shortage of aircraft mechanics will occur as senior mechanics with their accumulated expertise retire. “A lot of deep understanding of aircraft systems will disappear. And these days, if young people cannot find something with Google, they struggle.”

But machine learning can go through the records of a fleet and all its repairs by those 55-year-old veteran mechanics. “You can feed both repair manuals and maintenance history into it,” Harris notes. “We hope it can cut down turnaround times.”

The technique requires natural language processing to read and parse manuals and optical character recognition to read handwritten notes of mechanics and pilots. Jetstar is working with a firm, Lexx Technologies, on the approach. Harris says it may eventually not only speed repairs, but predict cancellations and delays, helping operation planners to deal with these incidents.

Eventually, Harris believes aviation will be ready for version 3.0, digital twinning. This would mean every major aircraft component has its digital twin, so engineers can better predict the performance of each component individually. “It’s not just statistics on many assets,” Harris stresses. “When you take delivery of a physical component, it will come with a digital twin that is intelligent. Then you will accumulate the history of the asset, for example, an engine’s bird strike or a landing gear’s hard landing.”

Harris says digital twinning is already being used for special assets in some industries. “There’s lots of work going on, and it proves you can get astonishing predictive power.” But the computing power needed to apply twinning to all major components on aircraft is immense.

The Jetstar manager is also enthusiastic about the potential on blockchain, or what he calls distributed ledgers, to improve safety and maintenance. But he wants a collaborative, disciplined effort to set it up. “The risk is a hodge-podge attempt.”

He believes airlines must decide what data will be kept, who manages the ledger and how it is managed, how the ledger will work with all the software now used by airlines, which transactions trigger an entry in the ledger and who has access to it.

The aim is ultimately to “democratize” the data. One gain would be safety, if all airlines around the world could see all the safety-related incidents that occur to a particular part under particular circumstances, not just their own incidents. Another benefit would be smart contracts that execute automatically, and a third would be the ability to spot rogue or counterfeit parts. Harris believes leasing companies would also be big beneficiaries, as lessors seek digital records but cannot impose digital record-keeping on their less-sophisticated airline customers.

TAGS: Big Data
Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.