Aviation’s maintenance, repair, and overhaul (MRO) sector has witnessed an incremental and sometimes tentative adoption of technological innovation over the years. Now, advances such as artificial intelligence (AI) present an opportunity to accelerate the industry’s development from a “paper-based” enterprise to one committed to digital transformation, according to General Aviation Manufacturers Association director of airworthiness and regulatory affairs Joe Sambiase.
“AI in my opinion actually represents a pretty significant opportunity for safety improvements,” Sambiase told AIN in a pre-Farnborough Airshow interview. “If you think about the different buckets of safety improvements we've had over the last hundred years, we continue to evolve each one of those, whether it be technology or training or reliability. So we keep adding to those, but every once in a while some new opportunity comes up. And I think AI is one of those.”
Still, the need for modern infrastructure advances remains a hurdle to adopting fully digital processes, he explained. “So obviously a government agency needs funds and resources to go ahead and implement software and hardware and then train its staff on how to use it,” he explained. “We actually did a survey of manufacturers, operators, and maintenance organizations to see how and where they're using digital data versus where they cannot. And what we found was a lot of the inward facing things like internal procedures and manuals that an organization needs to write could be done on a digital platform.”
While MRO providers have access to “all the information in the world,” they can't process it fast enough to do anything useful with it, noted Sambiase. However, AI can grab that data and yield something of value much faster than today’s processes, as long as the information exists in a digital format. “We still have to go in and do a much further deep dive to determine whether or not that information's factual,” explained Sambiase. “But I think at least from that initial research perspective, AI can help us get there and start to pick out trends pretty quickly.”
Digitalization also serves as a prerequisite to support engine health monitoring systems, allowing mechanics to pull data in real time and continue to watch it, he added. Moreover, it applies to the use of so-called digital twins, which replicates all the systems and all the data of the actual engine. “We can go ahead and test and make some changes and do what we need to do on the digital twin before we do it in real life,” said Sambiase. “We can also monitor that digital twin as, again, it's got that digital data behind it to allow us to do an analysis, [such as] risk analysis, performance analysis, reliability analysis.”
Pratt’s Industry 4.0 Journey
Companies adopting digital twins include Pratt & Whitney, whose own Industry 4.0 journey has broadly increased efficiencies both in production and in the MRO environment. In a recent interview with AIN, Gilbert Sim, Pratt & Whitney’s director of aftermarket global operations technology and CORE (customer-oriented results and excellence), explained how Industry 4.0 incorporates digital twins to aid efficiency and yield safety benefits.
“Industry 4.0 is really the portion where we actually get a lot more from automation and robotics, but at the same time, we also insert a human piece into it,” explained Sim. “That's what we call a cobot, which is really the collaboration between human and machinery. And that's where industry 4.0 comes into play. Before we got into Industry 4.0, the majority of our processes were still pretty labor-intensive…but this digital twin…allows us to simulate the layout that we want. It allows us to put in the machines as we want.”
Role of Robotics
For example, Sim added, a shop’s pre-induction process includes taking photos of the engine from different angles. The quality of those photos varies from technician to technician, making it difficult to predict a standard, he noted.
“So what we did was through this automation technology, we actually inserted an inspection Cobot system that will do the necessary survey programming of this robot,” said Sim. “Now, with the press of the button, the robots will go around an engine and collect the photos...The system will actually generate it [for inclusion in] a report.”
Meanwhile, Pratt & Whitney already has seen benefits from its recently established Singapore Technology Accelerator (STA), where more than 30 innovations have emerged to maximize MRO productivity. The company announced a collaboration with the Singapore Economic Development Board to establish the technology accelerator program in 2022. Today, the STA works with more than 20 Singapore companies to develop new technologies in the commercial aviation sector. Its projects have focused on automation, advanced inspections, connected factories, and digital twins, P&W said, noting that the company has already applied the innovations across its four Singapore-based MRO facilities.
“We deliberately set up this technology accelerator in Singapore so that we’re in close proximity to support the Asia-Pacific sites and, at the same time, connect very closely with the headquarters over in Connecticut,” noted Sim.
AI-Enabled Parts Tracking
In yet another example of reducing or eliminating reliance on paperwork, U.S.-based Alitheon has devised a way to use AI to identify counterfeit and used parts sold as new. Already in use in applications such as automotive, luxury goods, and precious metals, the version of Alitheon’s system for aerospace has undergone four and a half years of development and, according to company CEO Roei Ganzarski, has drawn significant interest from MRO providers and airlines.
Ganzarski described to AIN how his company’s system for detecting bad parts applies to the aerospace industry.
“So what we do is basically challenge the status quo in aerospace, which is everything that has to do with paperwork,” he explained. “And the paperwork could be actual physical paper or it could be digital paperwork. It doesn't really matter…Paperwork is easy to fake. You just write up a new document, attach it to a piece of paper, and show a serial number or a barcode. What we're doing is saying, forget the paperwork. Let's use, in our case, advanced machine vision or optical AI to make sure that the part is what you think it is.”
Alitheon has developed a platform that can detect the minute differences in a part and determine whether it meets manufacturing tolerances. “We can see with traditional standard cameras…inside those tolerance bands and all of the various minute flaws, features, or aspects of manufacturing,” said Ganzarski. “So we found a way to see those things with standard off-the-shelf cameras and codify them into a mathematical representation. You pull out your phone, take a picture of the part, and it'll tell you that is in fact, part number 12 that the OEM just sent you.”
For Canada’s Bombardier, such AI-enabled advances can aid parts inventory forecasting, for example. Bombardier senior director of aftermarket products and services Elza Brunelle-Yeung noted that the company’s AI journey started about four years ago, just before the onset of Covid-19. Once the pandemic began to result in delivery delays and limited availability of parts, the ability to forecast inventory needs became even more vital, both for third-party MRO providers and Bombardier itself.
“That project was very, very successful and we're using these AI models, which are pulling on real life data to generate an accurate spare parts forecast,” explained Brunelle-Yeung. “We were able to do a pretty good job, I would say, to provide the parts that they needed.”
Next, Bombardier introduced AI-based advances to its “connected aircraft” platform to develop a module to provide maintenance recommendations to customers. Essentially a health monitoring unit for business aviation called Smart Link Plus, the program allows flight and maintenance crews to quickly prioritize and proactively troubleshoot in-flight alerts. At the heart of the program, the GE-provided Smart Link Plus box collects, stores, and transmits aircraft data, allowing operators to quickly track, troubleshoot, and dispatch a mechanic.
The Smart Link Plus box collects that data and then transmits it to the ground, where it gets decoded, decrypted, and sent to Bombardier’s cloud for consolidation of fleetwide data.
Digital ‘Magic’
“That's where the magic happens,” said Brunelle-Yeung. “That's where we're able to provide the next-level insights to our customers.”
Bombardier’s MyMaintenance app gives maintenance directors access to key information coming from the aircraft wherever its location. “It has the capability to send smaller data while the aircraft is flying over high-speed connectivity,” explained Brunelle-Yeung.
When the aircraft lands, it automatically transfers all the big data, typically over cellular networks. “So the service includes worldwide cellular coverage, but it can also be transferred over wi-fi if the person's got wi-fi in their hangar,” she added. “So the small data, the big data, all that gets combined into this app.”
By April of last year, Bombardier had won Transport Canada, EASA, and FAA approvals for Smart Link Plus in almost its entire line of Globals and Challengers.
Brunelle-Yeung noted that the platform has achieved a “very high take rate” among customers acquiring new aircraft, adding that Smart Link Plus box appears in the “vast majority” of Global 7500s and 3500s. She added that Bombardier has seen more and more momentum in demand for retrofits in previously delivered airplanes.
Meanwhile, the company’s work with AI also applies to its e-commerce module, where Bombardier sells spare parts. “In the context of MRO technology, when you are buying a spare part, it's not always very clear what else you need,” noted Brunelle-Yeung. “And it requires quite a bit of research or knowledge around what else you need to buy when you're buying that one part. Well, we've developed a module that's able to provide recommendations to our customers as part of that e-commerce.”
Image Analytics
For GE Aerospace, AI has enabled engineers to process data contained in advanced image analytics, according to company chief manufacturing engineer for MRO Nicole Tibbets. Most of GE’s portfolio from an R&D perspective centers on standard image capture, with which AI and image analytics can determine how much longer an operator can fly the airplane. “And then all of that data from our analytics-based maintenance team and those physics-based models goes back to design,” explained Tibbets. “It's really a full lifecycle because they can inform design in terms of where and how those components are failing. So when they introduce the next generation of stage one blade, stage one nozzle or combustor, we get improved durability.”
GE also has incorporated some virtual reality technology to accelerate and standardize training, allowing trainees to visualize a particular function at the same time they perform a manual task. “The other thing it's really good for is line maintenance activities,” noted Tibbets. “So if you have multiple crews of line maintenance activities that are happening, distributed across the world, how do you ensure that all the line maintenance technicians have the same training or if they see a defect on the engine, that they understand how to repair that in a very timely fashion? So it’s probably less about AI, but more about how do you create a new training regime that allows people to very quickly come up the learning curve in a highly regulated industry.”
Another advance that GE has embraced—additive manufacture—allows for tip restoration on compressor airfoils, for example. As a compressor ages, it loses tip clearance, which results in a loss of engine efficiency. “Additive is a really interesting space for repair because it does allow you to do some interesting geometries,” explained Tibbets. “And then obviously it gives you some unique material properties in terms of how you can build backup surfaces that have lost dimensional tolerance.”
Furthermore, additive technology replaces the need for a welder to work on each individual tip, allowing for batch production. “We have fixtures and trays and large powder metal beds that allow us to do up to 16 airfoils simultaneously in a much smaller footprint,” she noted. “So it allows us to achieve our capacity ramp with less labor.”
StandardAero Sets AR Standard
Already actively applying automation innovations and concepts such as “big data” to its everyday activities, Scottsdale, Arizona-based MRO provider StandardAero has spent almost a year preparing to employ augmented reality (AR) in its maintenance activities.
StandardAero director of operations for engines Jamie Gardner, who leads the company’s CFM Leap industrialization process in San Antonio, Texas, identified the applications and benefits of the AR glasses. “One of the things that we needed to do was to make sure that we could teach our students and our trainees, if you will, what it looks like to do a test properly,” Gardner explained. “Many of [the trainees] are very visual learners, depending on where they're at in their learning curve. So this technology allows us to not only video what we have in real time, but actually project it in house on TV screens…But then the technology also has the ability for you and I to call that up with an app. So I'm able to look what the team is doing off of my cell phone as they're actually going through training, if I so choose.”
The technology, which StandardAero plans to eventually apply to all the powerplants it supports, can give customers an unobstructed view of a boroscope inspection, for example, allowing for a better understanding of the conditions inside the engine, explained Gardner. “You know how interpretation comes up and how things develop between what's being read in the manual and what the technician sees via a scope,” he said. “You're actually able to break down barriers and start speaking to the actual issue at hand.”
Supplied by Indian information technology company Tech Mahindra, the glasses also facilitate infrared inspections and thermal imaging, allowing users to spot bleed air leaks and hot spots in the engine within the test environment. “Then there's also another side of this that we'll be implementing, and that will be as we're going through key critical steps along the production line, be it in the high-pressure turbine, low-pressure turbine, core assembly, maybe some of the inner workings of a gearbox,” added Lloyd Barker, senior vice president of Leap industrialization. “We will actually record and project that across our lines where people can see it in real-time as it's being worked.”
Barker added that StandardAero expected to induct its first Leap powerplants for Continued Time Engine Maintenance (CTEM) events (i.e. quick-time shop visits) in late March, including work scopes for high-pressure turbine (HPT) shroud replacements. He explained that the company planned to introduce functional and performance engine testing for the Leap in mid-2024 and achieve Leap Performance Restoration Shop Visit (PRSV) capability by the end of the year.