How to Optimize Data Management and Access to Enable Predictive Maintenance

By improving efficiency in data access and management, maintenance providers can make it easier to optimize workflows and engage in predictive maintenance.

In a presentation entitled “Less Data, More Action: How to Use Maintenance Insights to Reduce Downtime & Improve Troubleshooting,” at NBAA BACE 2025, a panel of experts discussed how MROs and repair stations can implement predictive maintenance by simplifying data management.

• SVP of Product at Veryon Kris Volrath led the discussion, joined by:
• Aaron Esparza, Director of Maintenance at DV Aviation
• Thomas Hagy, Assistant Director of Maintenance at FlyExclusive

What data management challenges to maintenance organizations face?

Volrath shared, “The operators ultimately deal with the challenges of fragmented systems. How do I pull data from flight operations systems and maintenance systems and inventory, flight planning, and troubleshooting and publication? How do I pull all that data together to help make better decisions?”

Volrath also named two key contributing factors for data overload in aviation maintenance: workforce shortage and supply chain delays.

Discussing a survey conducted by Veryon about data volume in aircraft maintenance, Volrath said, “Almost half the industry said, look, we're struggling with too much data. We're not necessarily sure what to do with it in time.”

Esparza named another challenge, sharing, “It it's mostly just trying to figure out how to action that data for us. That's what the data overload is on our side.”

Hagy supported Esparza, noting, “We had the traditional fragmented systems. We were using Excel documents. We were operating out of a flight system. Maintenance management was in another system, none of which were talking.”

Hagy continued, “From there, somebody had the good idea at one point to come up with our own ecosystem where you're not doing the redundant data entry.

“You're not having the human mistakes and having details missed. And then all that data is talking to each other and displaying it at the same time,” noted Hagy.

How does effective data management inform predictive maintenance?

When Volrath asked about how Hagy and Esparza’s companies are putting predictive maintenance into practice, Hagy answered, “We're just trying to figure out how to integrate it. There's stuff as simple as asking how many landings am I going to get on this tire? Kind of planning that for budgetary reasons.”

He added, “But we're trying to come up with ways and just utilize it to do long-term planning as well, more so high-level predictive maintenance and not just looking at parts and components.”

Esparza noted how his team measures the success of efforts to use more predictive maintenance, responding, “In the utilization of these tools, the way I would measure success is reduction in time spent troubleshooting, getting the problem right the first time, more wrench time for the mechanics.”

Esparza elaborated, “They have to spend less time parsing this data manually and digging for answers, and they can just get a recommendation and get right to the job. It helps us complete jobs faster.”

“It builds better confidence with our customers,” said Esparza, “It helps us increase quoting efficiencies.”

“And I think it provides a good metric for our operations team to be able to have some autonomy and know when an aircraft is going to get released, if it's going to be on time or any kind of delays,” Esparza added.

Why is data access key for effective predictive maintenance?

Volrath asked how Esparza has been thinking about data beyond day-to-day operations, such as how the industry could be helping each other with data management and security.

Esparza answered, “I think component-wise, it’s a major blind spot. There are so many different databases that are seeing who's changing what components, and they're able to say, oh, hey, this break unit is failing 150 cycles before it should be considering its normal wear.”

He continued, “If we could start putting some of that stuff out to the operators, they could plan ahead. If we've got 30 flights in the next three weeks, we may want to try to plan a little bit of downtime in there so we're not stranded somewhere we've got to break back.”

Volrath interjected, “And today, you're looking at your data. Your problem is: I don't necessarily know the other operators, but with a similar same component, are others seeing the same challenges with life on that part versus what the OEMs and the manufacturers said it should be? Your point was how to better learn from a broader ecosystem.”

Volrath also inquired about the importance of gaining access to all necessary data from software providers to ensure optimal efficiency and effectiveness, especially when implementing multiple systems.

Hagy answered, “That's always a big thing for us. Anytime we talk to a new software company or we're entertaining something, we're basically demanding access to their API because we want to continue to integrate products into our system.”

Esparza said, “[For us] It's the same. It's a little bit more difficult with some of these systems to get access to the API. So sometimes we've had conversations with certain providers to where maybe they can provide us a tool to extract the raw JSON data so we can utilize that properly.”

How AI can help without risking safety

Hagy shared an example of applying AI to enhance safety centering on inspections and communication with OEMs.

He explained, “I showed it: 120 times we've done this one inspection. How many discrepancies actually came from it? And I presented that to the OEM that got that particular inspection’s interval longer for us.”

An audience member added perspective, noting, “When you’re talking about uploading aircraft maintenance manuals and using AI to search that, contractually, I can't do that. That's not my data to upload to another company.”

Volrath responded, “Nor can we, by the way…Anybody who's working with aircraft knows a manual is very specific about what you can and can't do in terms of using that. You can't summarize it. You can't change it. You can't create your own version of it.”

Volrath continued, “But you can absolutely search through it to make it easier to find what you’re looking for.”

Volrath also noted how it’s crucial to remember that while AI is designed to help reduce workload and time spent on tasks, it is not meant to fully replace human workers, especially in aviation maintenance.

Volrath said, “AI can cut 50% of my time out, but it's not going to replace the element of: I need to actually look at this, review this. The human blue element is important.”

He continued, “And I think when you get people past that, they’ll understand we still need you involved, we just need you involved in a way that's different.”

Volrath also asked how Esparza and Hagy balance increased productivity from implementing new software and AI-powered solutions with concerns about compliance and safety.

Hagy responded, “Trust me, verify. If it can get you to the reference of whatever it's telling you—like you were talking about utilizing it for the maintenance manual—if it can tell you where to go instantly so you can verify that yourself, it still saves you a ton of time.”

Hagy added, “The thing's not perfect. It's going to make mistakes. You can't let it go completely autonomous at this point.”

“In terms of safety, I've used it a lot to pare down RSN and QCNs, things like that. I send names to it and I pull it out,” explained Esparza.

He noted, “But it also has a basic understanding of the 135 regulations or whatever they are. So, you still have to go through it. You have to be the human factor and you have to edit it on your side.”

Esparza said, “You have to reinforce the idea that you're not replacing the critical thinker on the floor. They still have to take the action. And like [Hagy] said, trust me, verify.”