Given aviation’s position at the forefront of high-tech engineering, it is surprising how slowly the aircraft maintenance sector has evolved. Clever new inspection techniques have sped up and improved damage detection, and new repair processes have been developed to fix the latest composite materials, but maintenance largely remains a human endeavor, almost as reliant on touch labor as it ever was.
On several fronts, though, technology is threatening to disrupt the paradigm. This is not to say that robots will replace mechanics en masse any time soon; instead, new hardware and software developments have unlocked the door to automation, while also promising to reduce overall MRO costs for aircraft operators and lessors.
“For certain applications in the field of engine and structural overhaul, the use of robotics is appropriate,” says Alexander Simon-Sichart, head of corporate technology innovation and research and development at Lufthansa Technik. “This applies especially to repetitive activities, as well as wherever high precision and reproducibility are required.”
Examples of such tasks include single part repairs and carbon fiber machining. Miniaturization also promises to help the inspection of difficult-to-access components, especially in the engine. One of the most eye-catching developments in this field is Rolls-Royce’s recent announcement of plans to develop 1-cm-long (0.4-in) “swarm” robots to crawl through the guts of engines, feeding back imagery.
Although this would represent a “step change” in inspection capability, Rolls-Royce’s ambitions go significantly further, and include using bug-like robots to remove and replace defective material, says James Kell, on-wing technology specialist at the OEM. “In short, we would like to miniaturize the overhaul facility and deploy it on the end of a snake robot [which deploys the swarm robots].”
However, Kell admits there are “major hurdles” to overcome to realize such ideas, and he estimates that it will be at least five years before the initial generation of tiny robots is operational. Already in use, in contrast, are automated drones for simpler tasks such as external airframe checks.
Airbus demonstrated one at MRO Americas and aims to have the system available in late 2018.
EasyJet and Thomas Cook Airlines have tried out a different autonomous drone that can inspect a full narrowbody exterior in 30 min. and a widebody in 1 hr. Developed by MRO Drone—a joint venture between Blue Bear Systems Research and Output 42—the RAPID (Remote Automated Plane Inspection and Dissemination) system can also inspect specific structures such as after reports of a bird strike. It is an evolution of the RISER drone originally developed by Blue Bear for nuclear reactor work. “A key progression under the [joint venture] was recognizing that RISER, although highly capable as a hazardous environment inspection system, was not suited to the demanding environment of MRO,” says Gavin Goudie, operations director for Blue Bear.
The value of purpose-built robotics for MRO has been demonstrated by Lufthansa Technik, which has industrialized its automated inspection technology, AutoInspect (see Inside MRO cover above), for CFM56 and CF34 combustor components. Development of an automated repair process, AutoRep, for any cracks found in those parts is also near completion, allowing the MRO provider to offer an end-to-end service that can function with minimal human input.
Simon-Sichart says Lufthansa Technik is developing mobile robots and movable, multifunctional ones that use image recognition and artificial intelligence to choose the right work scenario. “We see these kinds of robots ready for use by the end of the decade,” he says.
In addition to inspecting and repairing components, machines can also produce replacement parts via additive manufacturing (AM). The advantages are clear: shorter lead times for spares and lower inventory costs for maintenance providers. In-service feedback also allows for continuous improvement to the durability and performance of AM parts, which can be reengineered at each new printing.
Most large MRO providers are now developing AM capabilities, either in-house or through joint ventures with specialists. Their typical starting point is non-load-bearing cabin components printed in plastic. For now, development of AM for critical structural components is being left to OEMs, but companies such as Lufthansa Technik and Air France Industries KLM Engineering & Maintenance are also pursuing metal printing for rapid tooling. In conjunction with its laser-cladding projects, the latter is also looking into metallic wire deposition to help it rebuild specific areas of certain engine components.
Fancy new robots make for great publicity shots, but sometimes what is under the hood is even more important. However, assessing whether hardware or software advances will have the biggest impact on future MRO practice is a moot point, says Kell at Rolls-Royce, adding: “I would suggest that you cannot benefit from software advances without hardware advances and vice versa.”
What is undoubtedly true is that data analytics, artificial intelligence (AI) and machine learning will have a huge influence on how machines operate and on where the efforts of robots and human workers are directed in the aftermarket. For instance, it is tougher to develop robots for maintenance than it is for manufacturing because the tasks in the former are more varied. “AI and machine learning will help deal with this variability,” says Serge Panabiere, Airbus head of services business development.
The proliferation of sensors on modern aircraft combined with better data routers has led to a sixtyfold increase in the number of data parameters collected from each Airbus flight, says Panabiere. The newest engines, meanwhile, can generate up to one terabyte of data each cycle.
Yet all this information is effectively useless without the means to analyze and act upon it. All aircraft and engine manufacturers now offer data tools as a services product, and although the technology is still in its infancy—with plenty of data still left unanalyzed—the ambitions for it are clear.
“Airbus developed the Skywise open platform, launched in June 2017, precisely to collect the vast amount of data coming from Airbus in-service aircraft, combine them with airline and OEM data and conduct in-depth data analysis to develop applications aiming at anticipating and optimizing maintenance and, more generally, improving airline operations and fleet performance,” says Panabiere.
As he indicates, the information to drive better maintenance planning extends beyond sensor inputs to other factors such as operational practice, weather and technical data based on the experience and knowledge of MRO providers and OEMs. “Combining operations data with weather and sensor data gives us an increasing understanding of fault analysis and prediction of reliability,” confirms Simon-Sichart, adding: “These insights improve the fleet reliability of our customers and reduce unplanned maintenance.”
Accordingly, each player in the aftermarket—airlines, MRO providers and OEMs—generates its own valuable data, and the lengths to which each goes to protect, monetize and co-opt such information will be a major feature of the aftermarket going forward. Under the Skywise platform, for example, Airbus offers free access to anonymized operational data to any airline that submits its own. Thus, participating airlines benefit from a useful benchmarking tool, while Airbus receives the data it needs to refine its paid-for predictive maintenance product.
Independent software providers offer alternatives and supplements to OEM platforms for maintenance planning. “Intelligent predictive maintenance planning allows MRO managers to create an accurate forecast of required maintenance activity on aircraft and aircraft systems, based on their predicted flying patterns over a defined period,” says Gary Vickers, CEO of one such provider, Aerogility.
He adds that “intelligent” platforms such as his company’s provide the ability to include unscheduled maintenance and repair activity based on a probabilistic modeling of the failure rates of key systems.
Connectivity and the cloud-based resources of the internet have made it possible to smarten up even the simplest tools, which join the Internet of Things (IoT) once you add a Wi-Fi connection. Improvements to issues such as latency and lag are still needed for these networks to function seamlessly, but the potential is obvious.
The Airbus Hangar of the Future project envisages the use of Skywise data with IoT equipment such as collaborative robots, drones, scanners, cameras and nondestructive sensors to optimize maintenance planning and task execution.
Combining machine learning with advanced robotics will lead to some startling new tools for maintenance providers. Add a human, too, and even greater benefits should result. “AI will help us to build new user interfaces by voice and image recognition and to find defects in structures and components resulting in more efficiency and reduced maintenance costs,” says Simon-Sichart.
Lufthansa Technik is enhancing “classic” industrial robots such as automated arms with additional sensors and wireless communications. Not only does this make them more capable at existing functions, such as applying component coatings, it also allows them to be used in novel ways, such as human-robot collaboration. One example might be a robotic arm that uses voice and image recognition to pass tools to an engineer on request. A less intelligent—but equally useful—development in other parts of industry are power-assisted suits that ease the lifting of heavy loads by workers.
Other technology-based assistants already in limited use are virtual reality applications to train engineers and augmented reality systems to improve situational awareness, declutter the workspace and improve performance. Smart glasses tested by GE Aviation, for example, alert mechanics about the correct amount of torque to apply when tightening engine bolts with a Wi-Fi-connected torque wrench. Other applications might allow MRO workers to call up technical documentation or instruction videos without breaking from their work to consult paper manuals or laptops.
The Path Ahead
Although some emerging maintenance technologies are already in operational use, others are little more than concepts or aspirations. Some build on or combine well-refined technologies and subsystems; others require further breakthroughs to fully realize their potential.
“The major subtechnology we are investing in for in situ robotics is deployment technology like snake robotics. These techniques are key enablers to allow us to make use of other processes that we are developing, such as miniaturizing coating deposition equipment,” says Kell at Rolls-Royce.
At Airbus, Panabiere highlights AI, augmented reality and collaborative robotics as three key enabling technologies going forward. He also points to four realizable goals for technology in the next few years: faster troubleshooting, thanks to cognitive assistance; automated aircraft inspection to reduce aircraft downtime; better maintenance task planning and preparation; and optimized spare-parts stock management and delivery.
Despite these advances, aviation remains a conservative industry—one in which the benefits of new systems rest not just on the technology but also the will and regulatory acceptance to implement it.
“Working with the innovations leads and commercial departments has shown that there is a huge appetite for drone-based inspection, but there is still some reticence in adopting such technology in what is traditionally a very risk-aware, highly regulated and cautious industry,” says Goudie.
He adds that MRO Drone had “no illusions from the outset that this would be a challenging environment to push new ways of doing things, when the traditional methods are so well understood and have been developed over such a long period of time.”
In some cases, new ways of working will require new regulation. The RAPID drone’s sensors mean that it can work inside and outside the hangar, for example, but the idea of drones operating on the ramp and near runways can be disconcerting and is, indeed, expressly forbidden by many aviation authorities. However, Goudie reports that several airlines in the U.S. have expressed interest in outdoor inspections to take advantage of the favorable weather enjoyed in certain states.
“We actively engage with the regulators and stakeholders in the territories [where] RAPID is garnering interest, and find that our position as a U.K.-based unmanned systems company facilitates practical discussions on how to achieve outdoor airside operations, where it may not have been considered in the past,” he says.
Some have questioned whether the final hurdle for advanced technology will be pushback from unions that fear for workers’ jobs. In some industries this is a valid concern, but virtually no maintenance managers believe they will lose their reliance on touch labor in the foreseeable future. Despite some huge leaps, our robotics and AI capabilities are still too immature to put humans out of a job in a sector as complex as aircraft MRO.
Related Gallery: See how robotics, blockchain and other innovative technologies are being used in the MRO industry: AviationWeek.com/MROEmergingTech