AI in the Hangar

Hi. Good morning. Good afternoon. Everyone will give just another couple seconds for everyone to get in and get settled here. Alright. Looks like we are at the top of the hour and got some people still following in. We'll go ahead and get started. Welcome. Thanks for joining. My name is James. I'm a sales engineer here at Veryon. I'm joined today with Andrea. She's one of our product managers. And today, what we're gonna be talking about is what AI will change in business aviation maintenance and what it won't. Problem first, product second, we'll walk you through. You know, feel free to drop questions in the q and a panel as we go, and we'll save some time at the end. Hi, Andrea. Yeah. Thanks, James. Yeah, so if you run maintenance for a part 91 or a 135 fleet and you're skeptical of vendor AI claims, then good. You know, we want you to stay skeptical And today, we'll show you what we would want to see before trusting an answer. Cool. Let's start with what keeps DOMs up at night. Yeah. So it isn't fixing the aircraft. It's not seeing what's coming. Most operators learn about unplanned maintenance when an owner's trip cancels and then that cost shows up on an invoice a month later. There's four questions that every operator needs to answer. Your VP of flight operations wants to know which aircraft is next and your accountable executive wants to know what is it costing? But the answers live in maintenance, right? So your director of maintenance needs to know whether you've seen this issue before, your shop foreman, he needs to know whether the fix held up. And every answer gets rebuilt by hand after the fact. Right? So... Yeah. Yep. And before we meet two of those people who have to answer to them, let's hear from you guys. We're gonna go ahead and drop a poll. You'll be able to see that pop up here momentarily if it hasn't already. When unplanned maintenance event comes in, how long does it take your team to build out that work order? Count all of it, checking your maintenance tracking system, looking up the publications, the manuals, checking the OEM portal, and getting all the information entered into your shop system. So we'll take a second there. Let those interest pile up, and we'll keep it moving. Yeah. And while those are coming in, you know, when we ask operators this, most land somewhere between thirty minutes and an hour once they add it all up. So if you are in that range, then you're in good company here. If you're under fifteen minutes, then we'd also love to hear how you do it. So Yeah. Great. Looks like those results are in. So now what we're gonna do is take a look at our scenario here. So picture a part 135 operation, a 20 aircraft in the fleet. We have Mike, who is the director of maintenance. Tom, he's the shop foreman. Keep in mind, these are fictitious. These aren't real customers. These times are averages. Right. So when the VP asks, you know, which aircraft is next, it's Mike who goes and does the digging. He gets a fault report and he needs to know, have we seen this before? And is it on one aircraft or is it fleet wide? Today, let's say that takes around forty five minutes and the answer still covers only his own 20 tail fleet. And then Tom's cost is in the work order. Let's say fifty minutes or even sometimes over an hour entering the same tail number, same part number and the logbook entry again and again, everything that Tom typing into the data, everything that Tom types in, it's data that leadership is waiting on. So none of it comes back to him as an answer. And then now, your numbers are gonna be different as we know. Every operation's different, but the forty five minutes and fifty minutes, they're in line with what we hear in our conversations with customers. And if you answered thirty minutes in the poll or more, you know, numbers are gonna look familiar. Yeah, and those feel familiar, you're not alone. So the front office and the hangar, they're asking the same questions about all those aircraft all from different places. Exactly. And every answer gets rebuilt by hand and after the fact, right? So when maintenance data becomes intelligence instead and the numbers move from our pop or... Sorry, the numbers move, our partner operators saw up to 75% less troubleshooting time for their new technicians and then up to 23% lower down cost time. Great. So now let's transition, and what we're gonna do next is take a look at four questions that we have here, that we hear from directors of maintenance the most. We'll have Andrea kind of review some of them. So what we've seen here with what's real versus what's hype with AI maintenance right now, Andrea, if you wanna take it away. Yeah. So real is AI that surfaces patterns and suggests the next step. It's grounded in records that you can check. The hype is anything that claims to replace a maintenance team's judgment. So if your team can't trace a suggestion back to the source record, then it's not gonna be worth much. So where do teams like Mike and Tom still lose the most time? Yeah, so they're rebuilding those answers by hand. It's the same data that's being entered again and again, and every fleet question gets answered from scratch. And what would it mean to ask one question and get one fleet aware answer? Yeah. So look at Mike, right? He would ask, have we seen this before and on how many tails have we seen it on? And it gets answered with the source records attached and it's in minutes instead of most of an hour. And so last one, probably the hardest, where do these DOMs and why do they still not trust an AI answer and what would have to be true for them to act on one? Yeah, so I mean, nobody trusts a black box, right? So to act, a DOM needs to see where the answer came from, know the model was tested against verified customers and keep the final call with their own maintenance team in which those were gonna come back to each of those. Great. Next, I wanna go back to the fleet awareness answer. So we said asking one question and getting a fleet aware answer would change things. What does that actually look like for these users? Yeah. So it looks like Veryon AIRE. Veryon AIRE is our AI and data intelligence engine. It's built specifically for aviation. It takes your operational data, your maintenance history, and what we know across the fleet, and it turns it into intelligence that you can act on. So it's always working in the background and it finds patterns and flags issues as they emerge before they turn into downtime. So the goal is simple, It's less downtime, more aircraft available to fly. So then here's what that looks like for Mike. He's in the wreck for the aircraft with the fault and he asks Ayer, have we seen this before and across how many tails? He's gonna get the answer right there, right on the spot without an hour of digging for that answer. And then how does Veryon AIRE know enough to answer those questions? Yeah, so that's because everything sits on one data thread. Every discrepancy, work order, part, manual revision, it's all connected within one record of the operation, and that's our fleet wide data intelligence. So that same intelligence tells a technician what to do next and also shows leadership where availability and costs are heading. So Tom on the shop floor, Mike in the maintenance control, the VP of flight operations, they're all working from the same record. And then that's the same assistant everywhere. It's in maintenance tracking, it's on the shop floor, it's in the manuals, it's in defect analysis and knows which aircraft and which task you're working on. Great. And for our customers, how do they use AIRE? What do they have to do to get it? Yeah. So, I mean, if you're already on one Veryon product, then you don't need to do anything. You just sign in and Eir is gonna be there. Eir is gonna suggest and then your team is gonna do the reviewing and the acting on that. So that's really the key with Veryon AIRE. We're turning aviation maintenance into aviation intelligence. Great. And so let's talk about next saving time here. So let's put a number on it for our DOM, Mike, when it comes to, you know, his standard day operation coming through papers, looking at multiple sources of information, you know, fragmented, you know, systems where he's going through and hear, hear, hear. Does AIRE save him time? Yeah. So, like we said before, that could look at roughly forty five minutes of digging to confirm one fault isn't fleet wide. And then again, that answer is gonna only cover his own 20 tails. But after one question to AIRE, it's gonna match the history across the tails and it's gonna be across the worldwide fleet that's gonna help give them guidance. And then it's gonna link straight to the source record and that's all only gonna be within five minutes or less. He asks that question, he gets the answer back right away. And then like we said, to be clear, these numbers are representative. The forty five minutes matches what customers tell us that that check takes today. And the five minutes is our estimate of that same check with AIRE. So, you know, your own time, it's gonna depend on the fleet and your processes, but think back to that number that you gave in the poll earlier and whatever that was, it's gonna be directionally the same. Yeah. And it's forty minutes back in... On the hangar floor too every time Absolutely. Running these processes. Yep. So within the records here, we have one record, four outcomes. These slides up for everybody. Let's talk a little bit more about what a director of maintenance can actually get from that one answer throughout his fleet. Yeah. So so there's four things. So the first is gonna be faster decisions. One question gets a fleet aware answer instead of digging an hour to try to find that same answer. The second thing is less downtime. So chronics are gonna get caught earlier before an owner's trip gets canceled. Third is gonna be no guesswork. So every answer is gonna trace back to that source record so that your team can check and verify it. And then fourth is gonna be intelligence beyond your own tails, which is the de-identified outcomes that come across the fleet. So AIRE does the surfacing and the suggesting and then your people make the call and that's aviation intelligence at your fingertips. Great. So next, mentioned the de-identified data. You see here the slide evidence not guesswork. It leaves the final question, why should we trust the answers that we're getting? Yeah, so, you know, most AI in maintenance pattern matches on raw data with no ground truth behind it, and anyone can make a model more aggressive, right? But then every Veryon AIRE recommendation is built on three layers instead. So that first layer is the data, That's our fleet wide data intelligence that sits beneath AIRE, which is more than a 100m maintenance events. That's the fleet experience behind every answer and it's de-identified. So nothing identifies your operation. The second layer keeps it honest. It's the verified data set and holds more than ten years of expert verified outcomes and every change to how we detect chronics is tested against that before it ships. And then the third layer is your team. So as we've mentioned, AIRE is assistive. It surfaces patterns and suggests the next steps and your maintenance desk reviews it before anyone acts. And that's the, you know, then the airworthiness decision, like we've said, that stays with you. So how does this make us different from other platforms? Yeah. You know, most stop at visibility. They'll show you the problem. But what we do is we close the loop from a fault all the way to the documented fix. That's right. And as you guys can see here on the screen with over a 100,000,000 maintenance events within our fleet wide data and intelligence, in addition to the ten plus years of those expert verified outcomes. So the evidence is, you know, right there within our data. And, you know, speaking of data, the evidence is one part. But what else does a director of maintenance... Andrea, what does a director of maintenance need before acting on an AI answer? Yeah. So there's three questions that every director of maintenance asks, and AIRE has an answer for each. So first, where does the answer come from? With Veryon AIRE, every recommendation draws from more than fifty years of aviation experience and it's grounded in real world evidence, not just guesswork. The second is what happens to my data? We don't use your identifiable data, it's your records, those stay yours. And then third, who makes the call? So we touched on this a minute ago, but it's your team, they're reviewing the recommendation, they give feedback on it. And then behind the scenes, we're continuously reviewing the outputs, running the evals and updating our models so that the answers get better over time. So the big takeaway here is right there at the top. It's your data, it's your call. Yep, exactly. And that's why Varian leads with trust in our AI. And from that example earlier, asks the question, he reviews the answer, and he only has to act once. Great. And speaking about, you know, acting once, that brings us to NBAA BACE next week in Las Vegas. What's waiting for us at the booth? Yeah. Faster fixes, less downtime, and the answers that you can trust. So, you know, Booth 645, we'll show you what we haven't covered today. We're really excited about that. You know, you'll get a view into your entire operation, the right manual inside the work order and how to fix the pattern instead of the symptom and much, much more. And we want people to come in to stop by to, you know, see what we have. You want people to come with something specific. What can they expect? Yeah, bring your hardest problem. You know, the chronic defect that you've fixed more than once that keeps coming back, the fault that takes an hour to confirm, the question that leadership keeps, that your leadership keeps asking you. And any questions on trust or data or de-identification, we'll be here to, you know, walk through all of that with you. Right, yep. So yeah, you know, like she mentioned, Booth 645. Let's see how much time we can save you. You can talk to your fleet, ask the fleet, act act once correctly. You know, so with that, we're gonna go ahead and drop one final poll for everybody. What is it that you'd like to see next? Are you gonna be at BACE next week? Let us know. We can schedule a demo. If you're not gonna be at base, you know, feel free to, you know, reach out. We can set up some time, you know, virtually to meet. Or if you know, just more general information, you know, about Veryon and our business aviation. You know, just these three here, but you know, you can reach out, you know, for any of our products, you know, as well. Yeah. And, you know, any of these problems, that's what we wanna work through with you. So, again, please bring those to our booth at BACE. Again, Booth 645. Looks like we are getting lot of votes in. Looks like there will be several of you there next week. Great. Yeah. Will have a hour on Tuesday on the first day from two to five if you wanna stop by and, you know, see the system and grab yourself a nice refreshment as well. Stay in the air conditioning. I'm sure Vegas will be hot. Alright. And then just, again, wanna thank everybody for joining us today. Before we go, we do have a couple questions that I saw in the q and a here. So wanted to get your thoughts on them, Andrea. With publications, are there plans to integrate that into WorkCenter? Yes. So we are working on that and really excited to be able to deliver that as we've been... We've brought WorkCenter into our ecosystem, so, formerly EBIS MRO. But, yeah, what it's gonna look like is, you know, if you are a WorkCenter user, you're in work orders and then work order line items, our system is gonna know, you know, what aircraft you're working on. And then at the same time, if you are a publication subscriber, so you subscribe to the library for that same aircraft that you're working on. It's gonna link those together. So you're not gonna have to navigate away manually. You'll be right in work center, right in that line item, the aircraft you're working on. You're gonna have the hyperlink right there to all those pubs. So you're gonna have the manuals, the IPCs, anything you're subscribed to. And then just at a click of a button, you'll just be able to navigate right to those documents. Yeah. That's great. You know, already with WorkCenter having the, you know, ability of labor kits, for example, bring everything that you're gonna work on and then now one less thing to research for your manuals all right within there. Yeah. Absolutely. I guess staying on with WorkCenter and integrations, can you talk a little bit about if there'll be the work center to tracking integration? Yes. You know, this is the big one that we're all excited about, right? Integrating those two systems for the maintenance that's being tracked on the aircraft, work that's being done on that aircraft, two separated systems, we're bringing those together. So, we already have a really good feature to very easily import those maintenance tasks into WorkCenter, But where the extended process and that comes is getting that finished, you know, the data from that finished work back to the maintenance tracking records and the log books and things like that. So, you know, that's where we're gonna really make that process really streamlined is at the click of a button, you know, say you've finished the work in work center, it's all gonna be staged to just, you know, click a sync button or something like that to get it right back over to tracking. All seamless. It removes all that data entry that we saw from earlier in the presentation where the Tom scenario, it's taken him over an hour to do that work. You know? This will be reduced to to minutes. So Yeah. So all of, you know, maintenance tracking to your works... To your work center work orders and back, you know, now connected, which just makes everything, you know, more efficient. You have AIRE with your pubs there. You know, it's just making things as easier as possible. And, again, just to reiterate, you know, about, you know, Mike and Tom is just reducing that time spent not being efficient, not working on making money, not working on being safe, you know, all those things that AIRE and Veryon can help you, you know, with. So that's it for the questions. You know, again, just one more time, you know, thank everybody for joining us today. You know, like Andrea mentioned, you'll bring your hardest problem to the booth, Booth 645 next week at BACE, and we hope to see you there. Yeah. Thanks, everyone.