The problem
Engine maintenance, repair and overhaul (MRO) shops were under pressure. GE Aerospace described a significant ramp in demand for engine services, driven by growing air travel and by supply chain constraints that delayed deliveries of new planes and engines (GE Aerospace). Every hour an engine spends in the shop is an hour it is not on the wing earning revenue for an airline.
High pressure turbine blades are among the parts inspected most closely, because they produce much of the engine’s thrust. In a standard borescope inspection, a technician threads a camera into the engine, captures images of each blade and judges which images show a problem. The work is slow and depends on the individual inspector, so consistency is hard to guarantee across shops and shifts. GE Aerospace framed the goal as faster, more accurate inspections that reduce turnaround times and maximize time on wing.
What they did
GE Aerospace built the AI-enabled Blade Inspection Tool (BIT) for stage 1 and 2 high pressure turbine blades on the GEnx engine. Trained technicians use the tool to capture images of the blades, and the AI then guides them on which images to review (GE Aerospace). The company compares it to AI that helps doctors review X-rays: the technician still makes the call, and the tool narrows and standardizes what they look at. GE Aerospace adds that consistent images are a key input for building predictive maintenance models later.
In October 2024, at Aviation Week’s MRO Europe conference, the company announced it was finalizing plans to deploy a similar tool for the CFM LEAP narrowbody engine and rapidly developing one for the GE9X so it would be ready when that engine enters service (GE Aerospace). The LEAP deployment started in February 2025, with more than a dozen GE Aerospace MRO facilities and customers who service the engine receiving the tool (GE Aerospace). CFM International, which makes the LEAP, is a 50-50 joint company between GE Aerospace and Safran Aircraft Engines.
The tool sits inside a larger program. GE Aerospace says it has been integrating AI across its operations for more than a decade, has more than doubled its investment in AI technology over that time, and is one of the top AI patent holders in aviation. It also announced in 2024 that it would invest more than $1 billion in its MRO shops over five years. The company’s AI-use guidelines emphasize human oversight, data integrity and transparency. The tool was built in-house; the sources name no outside AI vendor, and trade press coverage repeats the company’s account (Military & Aerospace Electronics).
Results
According to GE Aerospace, the GEnx tool cut the inspection process from 3 to 1.5 hours while improving quality and accuracy. “With the deployment of our AI-enabled BIT for the GEnx engine, we have seen inspection process times cut in half,” said Nicole Jenkins, the company’s chief MRO engineer (GE Aerospace). The company says the improvement in accuracy is measured against standard borescope inspections.
The figures are self-reported. GE Aerospace has not published the sample size, the accuracy measure or how many inspections the comparison covers. For the LEAP and GE9X tools, the company states an expectation rather than a result: they are expected to increase the accuracy of part inspections in half the time compared with standard borescope inspections. The February 2025 release reports that the LEAP rollout was under way, not what it has achieved.
The strongest evidence of value is the decision itself. After about three years of GEnx use, the company chose to extend the same approach to its two most advanced commercial engine platforms. “AI, used the right way, is proving to be a powerful tool in our work,” Jenkins said (GE Aerospace).
What didn’t go to plan
Public sources report no setbacks, reversals or safety findings tied to the tool. The open risks are about evidence and scope. The only measured result comes from the company itself, with no independent evaluation, and no LEAP results have been published. Inspection time is one step in a shop visit, so halving it does not halve an engine’s turnaround, and GE Aerospace has not said how much the tool changed total turnaround time. A tool that steers which images a technician reviews also creates a new failure mode: if the AI deprioritizes an image showing damage, a human may never look at it. GE Aerospace’s stress on human oversight addresses this in principle, but the sources do not describe how missed defects are audited.
Lessons
- Aim AI at the bottleneck step, not the whole process. GE Aerospace targeted one inspection step on one blade set and reports halving its time.
- Keep the expert in the decision. The tool guides which images to review; trained technicians still judge the blades.
- Prove it on one platform before scaling. The GEnx tool ran for about three years before the company extended it to LEAP and GE9X.
- Treat consistent data as a second payoff. The company says standardized images feed the predictive maintenance models it builds next.
- Report the measure behind the claim. Without a published method, “half the time” is hard for customers or regulators to verify.
Adapt it in your organization
Is this for you?
- You run a high-volume visual inspection that technicians perform the same way many times a week.
- Inspection time sits on the critical path of an asset’s return to service.
- You can capture consistent images and keep labeled records of past defects.
- Your regulators or customers accept AI-assisted inspection with a human sign-off.
- You can time inspections and audit missed defects before and after the change.
A 90-day version
Pilot: add AI-guided image selection to one inspection type on one asset line, with half the technicians using it and half working as usual, and a senior inspector re-checking a sample of both groups’ results. Owner: head of maintenance engineering. Success measure: inspection time down at least 30%, with no rise in defects missed on the re-checked sample.
Metric to track
Time per inspection, paired with the rate of defects found on re-inspection. GE Aerospace reports a fall from 3 to 1.5 hours on the GEnx; a smaller cut, measured honestly, is a realistic first target. For a case where self-reported time savings needed testing against observed work, see the UK Government case.
Failure modes to pre-empt
- Counting inspection minutes saved as engine turnaround saved, when the inspection is only one step.
- Letting the AI’s image selection hide a defect no human then reviews.
- Rolling out to new engine types on the strength of the first one’s results, before measuring them.
- Publishing a time saving without the method, which weakens it with customers and regulators.
Questions to take to your leadership team
- Which inspection on our critical path depends most on individual judgment, and how would we measure its accuracy today?
- How would we detect a defect the AI steered a technician away from?
- What share of our turnaround time does that inspection represent?
Executive FAQ
How much time does GE Aerospace’s AI blade inspection tool save?
GE Aerospace says it cut GEnx blade inspections from 3 to 1.5 hours while improving accuracy, a self-reported figure.
Does the AI replace the inspector?
No: trained technicians capture the images and make the call, while the AI guides which images to review.
About the curator & AI disclosure
Curated by Luciano Ayres
Luciano Ayres is Director of Artificial Intelligence at ArcTouch, with more than 20 years of experience in software and technology leadership and specialized studies in deep learning. Connect with him on LinkedIn.
AI-generated briefing. This case was produced through an automated AI research and writing workflow curated by Luciano Ayres. Sources are linked for independent verification; summaries and analysis may contain errors and do not replace professional due diligence.