Incorporating prognostics recommendations into maintenance scheduling may require a shift in how work is planned. GE Aviation
Incorporating prognostics recommendations into maintenance scheduling may require a shift in how work is planned.

Next Steps In Harnessing Big Data For MRO

Mastering how to use predictive maintenance effectively now will help MROs keep up with fast-changing aircraft data collection and analysis.

Engine makers have been using sensor data to monitor and manage the maintenance of engines for a long time. The technology is rapidly expanding to other components, and as coverage expands, analytical sophistication is likely to improve as well. This revolution is just beginning, and it will bring many changes both on the ground and in the air.

Everyone talks about performing predictive maintenance now, but what—beyond the fancy algorithms—is actually necessary to do it effectively? First, airline maintenance and other departments must make major changes in the way they do business on the ground. Cavok Vice President David Marcontell ticks off some key steps:

True prognostics require granular data about all maintenance and maintenance-related activities at every step, from onboard sensors through the supply chain. Prognostics will use machine learning and algorithms to identify trends in data captured in many places about the five Ws: who, what, when, where, and why?

Effortless Data Capture 

Companies cannot burden technicians with data entry. The infrastructure must be there to capture data without additional work, by exploiting tools such as RFID tags and barcodes whenever possible. This also means technicians will need mobile devices to enter data electronically, without time-wasting trips to a kiosk or computer.

Mobile devices also must give maintenance engineers instant access to the results of predictive maintenance, the work orders, maintenance manuals and much else, right where they work, so they can execute the new approach, both on the line and in hangars.

Sustainable Trust in Results 

Mechanics must trust the results of predictive maintenance tools, and the accuracy of these tools must in turn justify this trust. Marcontell says the new generation of digital-savvy mechanics has learned to trust diagnostics, not just their own experience. But they will sometimes have to switch out $100,000 parts that appear to be functioning normally. If many of those removals turn out to be unnecessary, confidence can quickly be lost, and the whole effort might collapse.

Prognostics System That Closes the Loop 

As all maintenance activities must be tracked, so must all core steps: notification of discrepancies from technicians, cabin crew or pilots; tasks performed; work orders; materials; completion and the ultimate results. That way the system can reveal to managers whether prognostics are really working and how to improve the process.

Maintenance Planners Who Work Smarter 

Planners must be able to exploit data to determine when an unscheduled removal is wise and to improve the work-scope of scheduled maintenance. They must add to scheduled maintenance and deferred items a third element prognostic recommendations—and do it all on the fly while ensuring that the right skills and other resources are available. This will be a big and difficult change for many planning offices.

Supply Chain Managers Who Support Prognostics 

These managers must ensure the right parts are in the right place at the right time by moving, borrowing or other means to support planners’ decisions and keep aircraft flying rather than waiting for parts. The supply chain must be as sensitive and responsive to predictive maintenance recommendations as technicians and engineers are. 

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.