ATLANTA – Most corners of the aerospace and defense sector grasp the promise of so-called big data analytics by now – and likely no place better than the MRO domain.
But increasingly attention is turning to how to realize the benefits of big data, and more importantly, how to overcome profound barriers to implementation such as culture, ambition and misunderstanding, according to a panel of analytic executives from manufacturers and operators at MRO Americas 2019.
“One of the things that slows us down is we try to do everything,” said Eva Azoulay, vice president for engine services at United Technologies Corp.’s engine-maker Pratt & Whitney.
“We need to be modest,” echoed Ghislaine Doukhan, executive vice president at Safran Analytics. “The choice will be made by airlines on a case by case example.”
Azoulay said the hardest part for companies like Pratt in pursuing big data offerings is figuring out the business case first. “I’m always amazed as an engine manufacturer that we can make a business case and invest billions of dollars based on assumptions in the market and go,” she said. “When it comes to investing in this area, where people are less knowledgeable, it’s a harder business case to make.”
The next leading challenge is recruiting and retaining a critical mass of technology workers who can create the product and underlying system necessary. That is particularly challenging for companies that have reputations and workforces that resemble classic manufacturers, not modern tech companies.
Doukhan said product creation is a huge hurdle. “That’s not magic, it takes time,” she said. About 80% of the initial effort is spent on cleaning data feeds into a system.
Valerie Manning, senior vice president of customer support at Airbus, said it can take six weeks for new airline customers of her OEM’s analytics offering to start getting insight from the data they provide as part of the system. Part of the issue is getting all of the necessary data. “One of the things is the data has to be clean, and we have to get it, but it is not a binary system between OEM and airline. It has to be global,” Manning said.
Several panelists said human factors, such as how credible and approachable big data is for users, remains a major hurdle. From corporate officers buying the product to shop-floor users who will be told to take action, or not, based on data but which differs from their practices, the information must be believable and compelling.
“Culture eats strategy for lunch,” said Dawen Nozdryn-Plotnicki, director of analytics solutions at Boeing Global Services. “Evidence-based decision-making is a culture and that has to be top down, [and] it has to be met by enthusiasm by the bottom up, one layer at a time. That kind of culture change is not easy in any industry,” she said.
What is more, Azoulay said it is especially challenging in an industry where safety is core. “You’re assuming that nothing we’re going to do will have a negative impact on safety when you’re removing an engine. But when you’re removing an engine or a component, this is multiple millions of dollars, this is an operational impact. And if you’re having a lot of false positives [then] you’re having the wrong impact on the organization,” she said. “I don’t have five years to test out the algorithm. We have to think how do we validate the capabilities quickly.”
To that effect, Doukhan reminded the conference audience that the business world is full of potential competitors, many from the tech sector that could suddenly enter and disrupt the A&D and MRO sector. It is in industry’s best interest to work together to make big data work.