
GE Aerospace says preliminary studies have shown AI-enabled visual inspections have more accurate, consistent results.
NISKAYUNA, New York—Newer, more complex aircraft engine designs are driving significant advances in performance and fuel efficiency, but they also require new, improved maintenance techniques.
As GE Aerospace prepares for the GE9X’s service entry next year and tests new technologies through its RISE program, the engine manufacturer is also tackling how advanced technologies can accelerate maintenance and technician training.
During a press event last week at the GE Aerospace Research Center (GRC), the company shared examples of how it is utilizing technologies such as artificial intelligence (AI), robotics, digital modeling, and augmented and virtual reality to improve aftermarket support for in-service engines.
According to GE, visual inspections account for the largest share of work performed in its global network of more than 18 MRO facilities. The company’s engineering division, which performs more than 4,000 shop visits annually, has partnered with the GRC to understand how to use technology like AI to capture consistent high-resolution inspection images, stitch them together and digitally capture information to help drive maintenance decisions and performance recommendations.
For example, GE has developed an AI-assisted engine inspection system that uses a robot to place a borescope in a standard position within the engine alongside a turning tool that rotates blades in front of it to collect a standard set of images. The system pairs this with image analytics, machine learning and 13 individual AI applications to select the correct image, check for damage such as cracks or missing coating, and compare it to GE’s requirements for whether a part can continue service or needs to be removed for maintenance.
GE says the inspection findings and digital records can be fed upstream to its engineering division, which uses the information to consider potential design upgrades and provide recommendations to operators on how to operate assets optimally. These findings can also be applied to GE’s next-generation products, such as the GE90 or future RISE designs.
The inspection section of the GE90’s engine shop manual is 19 pages long, and it takes an individual trained inspector between 8.5-12 hr. to inspect a single part, the company says. However, robots can implement between 25-40 different tools—such as cameras or mirrors—to recreate human visual acuity of a part and capture 100% of the inspection process in less than 8 hr.
Augmentation And Collaboration
Automation also helps eliminate inherent human and environmental shortcomings. Human performance in visual inspections can vary depending on a wide variety of factors, such as different lighting conditions or inspector fatigue.
GE says it has conducted broad studies related to human versus machine performance, such as for fluorescent penetrant inspection. The results showed that a camera system paired with machine learning had more than 85% accuracy compared with around 70% accuracy from a human relying on subjective visual findings. GE says the technology also has an advantage because standardized image collection provides the same image quality from the same representative location on a part every time.
Embracing automated image capturing and analysis does not mean humans are out of the loop, GE emphasizes. Rather, they help cut down on data-aggregation time and isolate issues that require an expert’s review. They also can be shared to bring more maintenance experts together.
GE created a connected inspection tool that enables inspectors to work with a remote expert to review images and make informed decisions based on what both parties see simultaneously. The technology was created in collaboration with the U.S. Navy to enhance aircraft readiness, and now GE wants to roll it out for all inspections across its MRO network.
Advanced Repair Techniques

At its Singapore service shop, GE has developed a geometric adaptive machining technology to improve the accuracy of CFM56 high-pressure compressor blade repairs. GE says these parts have a specific three-dimensional geometric shape they must maintain to achieve optimal aerodynamic efficiency, but the blades can become distorted or shortened during operation.
The repair, which involves welding new material onto the tip of the blade and machining it to blend with the rest of the part, must account for these shape changes. Manual finishing of the tip repair process requires a skilled operator, most of whom typically take around 18 months to achieve competency at the repair, according to GE. Post-pandemic workforce shortages led GE to look at automation options for the repair process.
The technology works by capturing blade geometry during inspection. This data is fed into GE’s adaptive machine, where software uses an advanced 3D morphing technique to create a customized repair program for each individual blade to compensate for any distortion or variation that has occurred during operation. GE industrialized the technology late last year in Singapore and now it is looking to scale it to high-pressure compressor airfoils.
Another engine with complex parts that will require advanced repair and inspection techniques is the GE9X. Designed with what the engine manufacturer says are its biggest, most swept blades, the GE9X also features ceramic matrix composites that will require new inspection technologies.
For instance, GE used computed tomography to view the layup of GE9X fibers and coatings in relation to one another in a very high-resolution 3D space. The company says FAR Part 145 does not have computed tomography as a validated inspection method across any of the next-generation engine OEMs, so to support service entry, it is working to understand how to clean and inspect components effectively for service return. This may mean qualifying a new inspection technology or finding an alternative equivalent already certified for the aftermarket.
Network Innovations

GRC is central to another repair development effort that involves GE experts at both its recently opened Services Technology Acceleration Center (STAC) as well as its Singapore engine overhaul facility. The company is modifying its patented Activated Diffusion Healing (ADH) process for braze and patch repairs—essentially filling cracks—in nickel-based super alloys.
The repairs are key for keeping critical low-and high-pressure turbine parts in service. The existing process is effective, but limiting in the scope of repairs it can handle, GE says. This causes many parts to be scrapped. Chemicals in the process can also be hazardous to workers.
GRC and the Singapore shop, which performs most of GE’s braze and patch repairs, is developing a hydrogen furnace to conduct the repairs. A furnace is used for both cleaning the parts and joining them with the filler material, which bonds to the base metal as both are heated. The first unit has been commissioned at the STAC and is in testing.
The furnace will be paired with a new type of ADH repair with improved material properties, GE says. The combination will see a new repair process that can handle larger flaws, which will lower parts-scrap rates. It also will be safer for workers, GE says. GE is confident the new repair will be ready for everyday use in 2026.
Meanwhile, GE has invested over $1 billion in training, tooling and capacity to support the GE9X’s service entry in 2026. For instance, its Customer Training Education Center in Cincinnati is working to add augmented or virtual reality training to complement classroom training.
Mohamed Ali, GE Aerospace’s SVP and chief technology and operations officer, says the company recently revamped many of its technician training programs to identify and plug in gaps related to the use of new technologies such as AI.
“Sometimes it is a limiting factor in getting people in and having them fully productive,” he told Aviation Week Network. “We want them [to] first get that training, because we believe it’s an investment in the future that’s going to pan out more, even though you might have to wait a little bit and spend some time and some money in breaking the box.”
GE has invested around $2.7 billion in research and development for RISE, including hiring around 900 engineers last year. Ali says RISE is the largest demonstration program in GE Aerospace’s history. The program is targeting a 20% or greater fuel burn improvement through propulsive efficiencies, including an open fan architecture, a compact core, and the integration of alternative fuels and hybrid electric systems.