Airlines, original equipment manufacturers, and maintenance, repair, and overhaul (MRO) organizations are making greater use of drones in inspecting aircraft as they overcome challenges to the deployment of such aircraft in hangars and on airfields.
Drone service providers and their airline and OEM customers say the small, unmanned air vehicles (UAVs) give maintenance shops the ability to perform faster, safer inspections — particularly of aircraft crowns. They also provide traceable and objective inspection results that can help resolve questions about an aircraft’s condition and help settle warranty questions.
“Drones are no longer a thing of the future,” said Jan-Christopher Knufinke, Lean innovation manager at Lufthansa Technik, who oversaw a three-year, multi-agency project to assess the applicability of artificial intelligence (AI)-enabled drones in base maintenance operations. “They will bring benefits to the MRO industry.”
That prospect is the result of years of broad investigation and initial investment in small UAV inspection capabilities.
Starting in 2015, for instance, the U.K.-headquartered, multi-national, low-fare carrier easyJet tapped a variety of service and software providers and academic researchers to help it explore the use of drones equipped with a variety of sensors to inspect its fleet of 200-plus Airbus transports. Early partners included the university collaborative Bristol Robotics Laboratory, drone-services firms Coptercraft and Blue Bear, and software developers Measurement Systems and Output42.
In 2018, Airbus showed off its Advanced Inspection Drone maintenance solution — optimized for checking the upper parts of a fuselage inside hangars — as part of its effort to speed up MRO procedures. Fitted with an integral visual camera, a laser-based obstacle detection sensor, flight planner software, and Airbus’ aircraft inspection software analysis tool, the drone was designed to fly automated paths along an aircraft. Airbus said the system could cut a typical day’s inspection down to three hours by capturing images in 30 minutes and analyzing them in 2.5 hours. The drone was designed to fly autonomously.
That same year, American Airlines partnered with DJI to test that Chinese tech company’s Mavic 2 Enterprise drone in aircraft inspection activities. The tests prompted the airline’s then-vice president of operations and industry affairs, Lorne Cass, to call the drone “a tool of the future that should be in every technician’s toolbox.”
Numerous things stand in that tool’s way. Drones flying in hangars must first map and then avoid overhead structures and elements, things that are rarely concerns for maintenance crews on the floor, work stands, or lifts. Legal requirements must be clarified and met. For instance, drones can’t create hazardous work environments for workers below them. They also must fit legally and operationally into an airport’s facilities and airspace. Singapore currently requires inspection drones on an airport to be tethered to a ground power supply. All that said, the challenges wouldn’t seem to outweigh drones’ benefits.
Advocates see drone-based capabilities as offering advantages in a number of areas. These include zone (or general) visual inspections, which can be long, costly, and subjective processes that pose human-factors and other hurdles for inspectors in accurately detecting damage and locating it on a fuselage. They can lead to prolonged email exchanges with OEMs and last-minute job card updates that cause late deliveries or warranty claims from operators and lessors. An automated drone inspection as aircraft enters the hangar may provide detailed, objective reports and precise, repeatable locations of defects in relation to aircraft structure, proponents say. Such precise detail also would support objective comparison of an aircraft status over time.
Another area is dent detection and measurement that follows events like ground support equipment collisions, bird strikes, hail and damage during production. Detecting and measuring surface dents and buckles is time-consuming. It typically is done by hand, and measurements can be unreliable. Automated 3D technology tools like 8tree’s dentCHECK (which that company says is approved by all major aerospace OEMs) offer mechanics/engineers and inspectors faster means of completing this laborious task. They also improve the traceability and reliability of the work. (8tree says its dentCHECK tool has slashed dent-mapping/reporting times by 90 percent, while delivering greater measurement accuracy and supporting answers that are compliant with structural repair manual requirements.)
When lightning strikes an aircraft, the bird must be grounded until inspectors find its entry and exit points. This can take several hours and lead to operational losses. Automated drone inspection of lightning strike impacts with detection and image analysis by software could take less than two hours, proponents say.
Drone inspections also could speed placard checks and improve the objectivity of paint quality evaluation as well as the efficiency of paint claims (which could help avoid adverse fuel burn from dirty airflow).
A drone inspection’s capture of digital images and precise damage location data also will help support reliable, objective records of visual inspections for comparison and speedier analysis of an aircraft’s condition over time, advocates of drone systems say. This would include assessments of how damage has evolved over time.
Many outfits are pursuing those and other benefits from greater MRO drone use.
Korean Air said it is conducting operational trials of a system can cut visual inspection times by 60 percent with “swarm drone” technology it has developed. The technology uses four small drones programmed to photograph four separate zones on a jet. The airline held a demonstration of the swarm technology last December at its Incheon International Airport headquarters west of Seoul. It expects to launch use of the swarm drones this year.
The airline developed the 12.1-pound (5.5-kilogram), 3.28-foot (1 meter) wide and high drones in house. By deploying four at once, Korean Air said, it expects to cut the usual visual inspection time of about 10 hours down to about four hours, a 60 percent decrease. This would help to improve on-time flight operations and “and dramatically increase operational stability,” the airline said.
Korean Air said it has developed an operations program that allows the four drones to be programmed to take photos of pre-planned areas. If one fails to operate, the system is configured to automatically complete the mission using the remaining drones.
Equipped with a high-performance camera, each drone can identify objects down to 0.004 inch (0.1 millimeter) in size, the carrier said, allowing for detection of micro defects that cannot be seen from above with the naked eye. Korean Air shares inspection data through the cloud, enabling employees to easily check inspection results anywhere and any time.
The drone system includes a collision avoidance function and geo-fencing to keep each UAV at safe distances from surrounding facilities and prevent break-aways from the mission area, the airline said. Korean Air has revised its procedures to require the presence of safety personnel in addition to pilots and mechanics while the drones are flying. Korean Air said it “will work to improve safety and convenience for workers, stabilize operations, and increase the accuracy of inspections” through continuous trials before officially launching the inspection drones next year.
ST Engineering in Singapore also has developed an in-house drone capability, the DroScan tethered system. The company said the Civil Aviation Authority of Singapore (CAAS) has authorized it to use DroScan within the country’s aerodromes in performing general visual inspections of aircraft during maintenance, provided it is tethered to a power source during operations. DroScan is designed to operate on onboard batteries, the company said. Its systems also include collision avoidance, geo-fencing, and smart analytics capabilities.
The drone is paired with a ground control station to automate inspections by flying along pre-planned routes to capture high-definition images. Items picked up by DroScan’s defect-detection algorithm are displayed on the ground control station and an operator can zoom in to certain areas or enhance the images to examine suspected defects. The operator also can manually annotate defects before inputting them into the database.
ST Engineering tested DroScan (which it said offers up to 40 percent time-savings over traditional visual inspection methods) with Air New Zealand and other airline customers. The company said it is in discussions with some customers to refine DroScan’s MRO workflow and conduct operational trials for their fleets.
Several airlines, third-party MROs, aircraft OEMs, and military operators have enlisted the drone services provider Donecle to improve the efficiency and accuracy of aircraft inspections. Founded in 2015, Toulouse, France-headquartered Donecle has become a world leader in the drone-based automated aircraft visual inspection business. Last April, it expanded its robotic inspections portfolio when it acquired the French autonomous mapping company Dronétix. That company specializes in automatically capturing data and doing 3D reconstruction of small assets (such as aircraft engines or landing gear). Dronétix customers include Safran, which regularly uses a drone for aircraft engine inspections at its Villaroche facility outside Paris.
“Combining the assets and know-how of both companies will strengthen our offer and boost the development of future capabilities,” Donecle CEO Matthieu Claybrough, “Our goal remains unchanged: to offer our customers cost-saving solutions while improving traceability and safety.” He said Dronétix systems will be upgraded with Donecle’s leading imaging technology and cloud connectivity, and the image datasets and AI technologies of both companies will be merged to further increase their performance. “This is an important milestone in Donecle’s growth.”
Dronétix founder and ex-CEO Franck Levy, said, “Donecle is the perfect match to further develop and commercialize our technology. We look forward to seeing the industry further adopt robotic inspection.”
Donecle’s drone inspection tools are being used by such airlines and MROs as AAR, Austrian Airlines, KLM Royal Dutch Airlines, and LATAM Airlines Group in South America. Last year, Airbus qualified Donecle’s technology for use in lightning inspections on A320-family aircraft. In April, LOTAMS (the MRO offshoot of LOT Polish Airlines) signed a multi-year deal for Donecle to help it optimize visual inspections. Donecle will support exhaustive inspections at LOTAMS of Boeing 737NGs and 737MAXs, as well as Embraer 170s, 190s, and 195s. That MRO also will participate in the development and validation of drone capabilities on Boeing 767s and Boeing 787s as Donecle rolls out new wide-body and outdoor inspection capabilities.
On the military front, Donecle said that it completed an automatic, drone-based dent inspection of an entire Dassault Rafale fighter within one hour as part of an 18-month collaborative project with that manufacturer, the French Defence Innovation Agency, and 8tree (which makes 3D optical surface inspection tools) to improve aircraft maintenance and operational condition. 8tree has been collaborating with Donecle and other leading UAV companies on automated flying versions of its 3D tools. The Donecle drone, which used 8tree’s dentCHECK dent-mapping technology, achieved an accuracy of 0.004 inch (0.1 millimeter) depth and 0.079 inch (2 millimeter) width in the Rafale inspection, Donecle said.
The collaborative project had three main objectives, the company said. Demonstrate that the drone can inspect an entire aircraft body quickly and easily, providing time savings. Show that the system’s 3D structured light scanner provides consistent results in all conditions. Prove that the 3D scanner and associated software can identify and measure damage such as dents, impacts, and misalignments.
As part of the project, Donecle said, it has improved its drone automatic navigation with advanced stability algorithms and novel management of the onboard 3D sensor to address a workspace with dynamic lighting conditions and various surface colors and types. It also used a 3D digital twin of the Rafale to map dents, using global reconstruction algorithms “to perform an automated diagnosis leading to instantly actionable results.” All acquired information (such as scans, localization information, and dent measurement results) were saved in a digital database, which “will help tracking the structure evolution in time to improve the aircraft availability and safety,” the company said.
Late last year, Donecle signed a customer agreement with Regional Jet Center, the Amsterdam-based Embraer MRO specialist, to deploy its automated drone inspections on Embraer aircraft for the first time. “We have been looking into drone inspections for the past couple of years as we believe they have the potential to accelerate inspections and improve overall traceability,” said Regional Jet Center Managing Director Michiel van der Eijk. “We were impressed by Donecle’s drone system capabilities. We are looking forwards to this partnership.”