MRO Embraces Artificial Intelligence, But Slowly

SINGAPORE—The incorporation of artificial intelligence (AI) into MRO operations is underway, but adaption of the technology is in its early days, industry executives say—and aviation’s role as an industry where safety is paramount means maintenance shops will move forward cautiously.

That was the upshot from two panels at Aviation Week’s MRO Asia-Pacific conference and an associated, not-for-attribution roundtable of industry leaders from across the region.

“I think it’s important to start small and have some early proof points,” David Harper, GE Aerospace’s fleet support director for customer service, said during one of the panel discussions. “Set expectations realistically and pay attention to some of the less glamorous sides of this. There’s rather dull work that has to be done before you get to the fancy stuff.”

Executives say AI already is being used in functions such as documentation, planning and helping optimize scarce repair slots. HAECO has started using AI to analyze maintenance work packages and now it is moving into the scheduling piece to “orchestrate our people, material and tooling in the hangar in the most effective way,” says Alex Chen, group general manager of digital.

“We have done a lot of pilot cases” with AI and are on the cusp of building something “more core” for the business, says David So, SVP for corporate planning and continuous improvement at SIA Engineering, the MRO arm of Singapore Airlines.

The idea of machines replacing humans on a large scale remains far-fetched, at least for now. “We are trying to introduce AI as a form of assistive technology,” says Kenneth Low, CTO, SVP and head of innovation and sustainability at ST Engineering's commercial aerospace business.

In an audience poll at one of the panels, a 42% plurality of respondents believe predictive maintenance will be the area in MRO where AI will have the biggest impact in the next 2-3 years. While engine OEMs in particular have provided predictive maintenance services for years, AI is improving it further. GE Aerospace, one of the biggest AI patent holders, uses it “in the monitoring diagnostic space to see earlier signs of trouble and give instructions to airlines to avoid disruptions and unplanned maintenance,” Harper says.  

A second poll shows the industry has woken up to AI, but slowly. Just 4% of respondents said they have no plans to incorporate the technology into their operations. Yet only 34% say they already are doing so, with a solid majority (62%) saying they are interested in AI but just starting to explore its potential. One challenge is getting workers to embrace learning how to use AI. Some are enthusiastic, others are wary.

Executives agree that any efforts to regulate the use of AI in aviation could short circuit its adaptation. “I hope we don’t start regulating AI,” Low said. “The moment you start to do that for our industry, everybody is going to just wait and see.”