Digital twins – virtual replicas of aircraft or components – are predicted to transform MRO in a variety of ways.

“Be prepared!” “Failing to prepare is preparing to fail!” There are many common phrases and sayings which involve preparing or being prepared, and they usually have a good dollop of wisdom behind them. The ability to anticipate events in aircraft operations can be the differentiating factor between success and struggle. It often begins with the design of an aircraft, which nowadays is created with maintainability in mind. Here, virtual reality (VR) has been used for many years, ensuring accessibility for maintenance technicians throughout the aircraft and engines.

Gaining deeper insights

The latest advance in this area is the digital twin, which is a facsimile of a real aircraft or engine. Pete Boeskov, director of commercial training, digital and analytics at Boeing Global Services, explains how a digital twin differs from VR designs. “We should differentiate between physical objects and virtual or digital tools,” he begins. “Virtual reality can be an effective tool to interact with under-design digital models/concepts. Whether a designer builds a physical mock-up or prototype, or uses a digital tool like VR, the goal is to gain a deeper understanding of the model or system.

“A digital twin, however, is a digital representation of a physical object. Only after we create a physical item can we then create a digital representation of it,” Boeskov emphasises.

“In the operational space, a digital twin collects all the life experiences and data of the represented object or system. We are working on digital twins of components, systems, aircraft and airports. A digital twin can include information as extensive as what material was used, where it was mined, who machined it, what machine it was created on, and more. This allows analysts to interrogate the twin to deliver more accurate insights,” he adds.

“Digital twins offer the possibility of superior data-driven decision-making and the desire for enhanced operational understanding and optimisation, accurate simulations and future predictions to enable levels of integrated vehicle health management that exceed existing capabilities,” Boeskov remarks.

Airbus takes a similar view, noting that VR is utilised in the industry “to mimic costly and/ or dangerous situations, or to simulate a specific experience”, according to a company spokesperson. “For example, designing assembly stations for part installations in tight spaces, we use VR to ensure that there is enough space for the necessary equipment and that the technician can manoeuvre and perform the required action.

“The digital twin of an aircraft is used to represent an often complex element (aircraft, engine, part, process and so on) in order to test scenarios and the impact of each of these scenarios. This then enables data-driven decision-making,” the spokesperson adds.

Dinakar Deshmukh, vice president of data science at GE Aerospace, notes that the company has been developing and applying digital twins for well over a decade to help build and scale predictive maintenance capabilities. “Using artificial intelligence (AI)/machine learning (ML)- driven digital twin engine models, we’ve been able to expand the number of conditions that can be monitored with greater accuracy and identify needed maintenance measures well ahead of time,” he explains. “These models have enabled us to achieve a 60% earlier lead time identifying preventive maintenance measures, a 45% increase in detection rates, and cut the number of false alerts in half over the past decade.”

Deshmukh continues: “The key difference with a digital twin is that it is a living, learning digital model that reflects the current state or condition of the physical engine it is modelling. It is a virtual replica of its physical counterpart at any given point in time. We do this by continuously updating our digital twin models with the latest data from sensors (virtual and physical) that monitor the condition of critical components and processes with the engine, from fleet data of what other similar engines are experiencing across the fleet, and from the insights of our engineers who design and monitor these engines.”

Observing this technology from an MRO provider vantage point is Dr Jean- Philippe Kremer, senior engineer performance and product owner engine health management at Lufthansa Technik (LHT). “As an MRO, we are not really involved in the design of entire aircraft,” he says. “However, certain fields of our overall business can also entail the application of VR or AR technology.

“Jet engines are not typical consumer goods where it’s all about design and looks – it’s about efficiency, durability, maintainability and so on,” Kremer adds. “A digital twin from a propulsion engineering perspective thus goes way beyond the pure visualisation aspect as it is a lot more sophisticated and can be seen as a multitude of sub-models, each one being focused on/catering to a specific use case.” Multiple MRO disciplines Potential users are now looking at the advantages of digital twins across the different MRO disciplines such as predictive maintenance, optimisation of repair schedules and reducing downtime. Kremer picks up the theme. “Well-designed models based on physical fundamentals help understand cause-and-effect pathways. Downstream fine tuning or improvement by ML models is facilitated if all known influences have been correctly modelled,” he remarks. “These algorithms are then able to find previously unknown correlations and further improve accuracy.

“For engine health management, a digital twin allows quicker detection and diagnosis of small initial faults and failures, before they become issues impacting flight operations. Raw sensor data from the engine or aircraft shows high scatter and does not always directly point to the root cause of the failure, so our digital twin can be seen as a ‘decoding device’, transforming data into information and knowledge,” Kremer continues. “Additionally, they allow us to integrate our own know-how from an MRO perspective so we do not have to rely solely on black-box systems designed by the OEM.”

For GE Aerospace’s Deshmukh, the big advantage of digital twins is the earlier lead times they provide for understanding the maintenance needs of critical parts. “Digital twins give us the ability to forecast maintenance well before the need has even materialised, which then allows us to plan ahead with our repair schedules. That, in turn, helps reduce the downtime for how long an engine needs to be in a shop,” he explains. “It can even inform adjustments to our longer-term material forecast to the new-make supply chain if our physics-based models indicate the component will not be repairable at the upcoming maintenance event.”

Boeing’s Boeskov hones in on digital twins in the world of predictive maintenance. “Here, digital twins are used to increase maintenance efficiency and reduce aircraft downtime by analysing the data to anticipate maintenance needs, while also ensuring the right components are available when and where they are needed,” he elaborates. “For example, general processor modules (GPM) are prone to overheating for a variety of reasons. This results in early failures for parts of the GPM and scheduled interruptions for the aircraft while the GPM is replaced.”

Boeskov continues: “We can track the configuration of GPMs across the fleet, in addition to tracking which parts have been installed and for how long. We can also measure the temperature on the internal boards and aggregate them across the life of the part. This allows us to develop an understanding of which parts are at risk of failure, plan for replacements of those parts and work with the supply chain to get the parts ordered and on-hand.”

According to Airbus, MRO activities can be mapped in a digital twin so that the company can always have a clear view of what’s happening in the maintenance chain – mapping all the chain and the link between each action or actor – and even simulate scenarios and find the most optimal slot allocation for an urgent task, with respect to its required skills, equipment and parts. “This can enable the company to see the impact of a problem throughout the whole chain and also to easily define and implement possible solutions,” the spokesperson notes.

Making business sense

Utilising digital twins has to make sense for a business. There needs to be a positive effect for the customers of MRO organisations – the airlines – in areas such as aircraft availability, cost savings and enhanced safety.

For an MRO, according to Airbus, the creation and optimum use of a digital twin should make it possible to improve processes and therefore save time and money through more effective decision-making and better prevention of problems. “This has a direct impact on aircraft availability and therefore on the airline’s profits,” the spokesperson emphasises.

The LHT view of the business case from Kremer centres on digital twins being all about understanding a system and its sensitivities. “This improved insight can then be made available to the customer – in the case of engine health management, through timely fault detection and customised maintenance recommendations, possibly even pointing out several courses of action, allowing the airline/customer to integrate them into their maintenance schedule at their convenience,” he explains.

Kremer adds: “Punctual interventions help maintain or improve engine performance and thus aircraft availability, while information improving the scheduled engine removal process leads to cost savings by preventing impact on operations and expensive secondary engine damages. Our goal is to support our customers in their journey to an improved operational excellence.”

According to Boeskov, when describing the value of a digital twin, the twin does not necessarily deliver value or savings. “They come from the insights that are derived from a digital twin that truly make an impact in the highlighted areas of increased aircraft availability, cost savings and enhanced safety,” he says.

Boeskov adds: “When we can find ways to increase the efficiency of MROs, that leads to reduced costs for operators, and increased profits for the MRO. When we detect a component failure early and recommend replacement, that reduces surprises and enables an early aircraft delivery from a maintenance check, enabling a faster return to revenue service and improved aircraft reliability.

“Digital twins deliver value when their insights are implemented well, saving operators cost, reducing repair times and putting aircraft back into service faster. Additionally, they enhance safety by reducing the risk factors that are involved in inspections and repairs,” Boeskov emphasises.

GE’s Deshmukh is another who believes that digital twins help deliver better outcomes. “For us, that’s principally with aircraft availability and fleet management, by helping us see and identify issues earlier, and by expanding the number of conditions that we monitor 24/7,” he comments.

“With MRO shops, it’s not unusual for them to experience an escalation in workscopes for engines coming for service. The insights from our AI/MLenabled digital twins are helping us predict part needs ahead of time so that we can be ready with the right materials and personnel to make any needed repairs,” Deshmukh explains. “We can also avoid any undue delays and find alternative, more cost-effective solutions for replacement parts for our customers in the form of used serviceable material (USM).”

With many undoubted attributes, one would expect enthusiastic adoption by the MRO industry. “On an engineering level, the enthusiasm is definitely there, but turning use cases into business cases remains a challenge,” says LHT’s Kremer.

Boeing’s Boeskov points out that seeing the value of digital twins requires a big picture view. “MROs don’t exist in isolation, but are part of a larger ecosystem including operators, the supply chain and OEMs. The ability and willingness to share data across the lifecycle is foundational to maintaining a digital twin,” he concludes. And that means being prepared!