Pratt & Whitney's 'Big Data' Projects Advancing Analytics Efforts in Aftermarket

Using data analytics to monitor engine performance is nothing new at Pratt & Whitney, company executives told reporters at a recent company event, but a shift in the aftermarket business and advances in technology mean new opportunities to deliver customer value through Big Data.

Speaking on a panel discussion that opened the annual Pratt & Whitney Media Day in East Hartford, Matthew Bromberg, president, Aftermarket, and Larry Volz, vice president and chief information officer, explained why now is the right time for the company to invest in its approach to predictive analytics in engine maintenance.

"A given Pratt & Whitney engine will experience an inflight event once in 100 years. That's an amazing statistic. And a delay and cancellation associated with that engine? Once a year. Now, there are a lot of engines out there, a lot of aircraft, so it does happen. But the engines are incredibly reliable and incredibly safe," Bromberg said. "But we want to do better. We want to capture every parameter, from every engine, every second."

[Audio] Matthew Bromberg: We want to capture every parameter, from every engine, every second.

Lynn Fraga, an analytics manager in Engine Services, described a project focused on unscheduled engine maintenance events, one of 14 Big Data projects now underway at Pratt & Whitney. That project, which targets PW4000 engines in service on widebody aircraft, is the first step toward developing a predictive model for anticipating and preventing in-flight shutdowns and engine-related delays and cancellations.

"What we're trying to do is expand our forward-looking horizon to make those types of engine maintenance events highly predictive and to start to move our discussion from a reactive to a proactive maintenance discussion," Fraga said. "We've developed a tool that proactively looks at these types of events and gives us an early warning and early detection on any potential fleet trends."

[Audio] Lynn Fraga: Data only becomes useful when we're able to translate that to a value proposition.

Pratt & Whitney currently captures about 100 parameters at multiple snapshots through a given flight, Bromberg explained, but that number will grow when the next generation of commercial engines enters service. The Geared Turbofan engine will collect 5,000 parameters continuously throughout a flight, generating massive amounts of data.

At maturity, the Pratt & Whitney Geared Turbofan engine fleet will be generating more than two petabytes of data annually. That is the equivalent of a new American Library of Congress every year.

Pratt & Whitney formalized an agreement last year with IBM to work toward better visualization of the data streams generated by its commercial engines. IBM, which has the world's deepest portfolio of Big Data and analytics technology, will help Pratt & Whitney broaden its current performance monitoring capabilities. The result will be better value for customers, as no two operators are the same.

[Audio] Larry Volz: Working with Big Data is not new at Pratt & Whitney.

"There are different climates, different temperatures, different air pollution, and those impact engine performance and maintenance," Bromberg said. "With the ability to monitor every engine, every parameter, we can actually go one step further. We can customize the engine maintenance for that operator, for that city pair, and for the way their pilots are trained. That's good for us and it's good for them. The reliability will improve and the cost of ownership will improve."

Advancements in technology and investments in enterprise-wide systems and high-performance computing enable this expanded approach to analytics, Volz explained. The technology has reached a point where the business vision can become reality. The IBM agreement will serve as an "accelerator" as Pratt & Whitney makes strides in analytics.

"We're able to handle terabytes of data and speeds upward of gigabits per second and beyond. That's very, very fast processing of data," Volz said. "We understand how to process large amounts of data, but it's been pretty much focused in the engineering space. Now we can look at expanding that to other areas of the business."