The Vertical Space

#65 Andrew Coleman: Unpacking the Value of Data in Aviation

• Luka T • Episode 65

Share episode

Welcome to another episode of The Vertical Space where we sit down with Andrew Coleman, leader of GE Aerospace's software business. In this conversation, we explore the critical intersection of traditional aviation and advanced air mobility, delving into the effective use of data for meaningful insights. Discover how GE mines, analyzes, and provides data for aviation customers, and how this applies to the evolving landscape of advanced air mobility. From discussing the perception of the value of aviation data to identifying specific use cases for enhancing safety, sustainability, and efficiency, Andrew shares valuable insights that shed light on the future of aviation. Tune in to learn how a focus on targeted use cases and well-defined data can drive success in both traditional and advanced aviation industries.

Andrew:

And so in the ecosystem of aerospace, we've channeled our energy first to operators, next to lessors and third to MROs. And inside that, we really said, where is our permission? Where can we move the needle? And we wanted to think big enough, but we wanted to think so big, it was ludicrous. And so in the space of full flight data and the space of maintenance records, we saw so many use cases that frankly got us very excited. When you contrast that to a technology company, who's looking at large addressable markets, they tend to look at retailers and banks and government agencies, and they get to our industry and we love our industry, but it's not the biggest industry. And so a company like Amazon, Microsoft, Google tends to not want to get down to that level of fidelity of what's aircraft data doing, what are these thousands of parameters I'm pulling from this aircraft, where they truly meet?

Jim:

Hey, welcome back to The Vertical Space and a conversation with Andrew Coleman, the leader of GEs Aerospace' Software as a Service business. So this is another conversation with a more traditional Aviation business leader, like Uma, Paul, Eric, and Lorne and of course, previous guests who discuss the intersection of traditional aviation and advanced air mobility. This intersection is critical to understand for the viability of businesses on both ends of the aviation spectrum. The effective and smart use of data is heaven for us at The Vertical Space, so we were looking forward to our conversation with Andrew to hear how one of the capable blue bloods in aviation, GE mines, analyzes and provides data for meaningful use cases for their aviation customers. It's great to hear Andrew's comments, how airlines buy, who to target and the best way to approach and sell to airlines. This clearly applies to our advanced air mobility customers as well. We start with a discussion around the perception of the value of aviation data in the past 10 years or so. And how under Andrew's leadership GE has identified select areas to assist their core customers, commercial and business aviation. And towards the end of our discussion, you'll hear how some of the same value points may apply to advanced air mobility. You'll hear about what data is being always saved by aircraft, the original eventual use of that data. How it's analyzed and presented back to operators and their pilots, how it's evaluated and valued by airlines and business aviation and different use cases. With GE the use cases start with safety, but listen to how GE has taken that basic safety data and helped to shape and influence his airline sustainability and efficiency efforts, because this contributes to how airlines actually pay for the product. You may be a bit surprised that a company of the size success and prestige of GE is, as Andrew says, not starting with big, hairy, audacious goals, as it relates to addressing problems with data for airlines. But our targeting very specific use cases with well-defined data. Some may be disappointed in this non-sexy or loftier approach, but it's important for all to understand that such focus may be a good lesson for those trying to build a business in aviation. It's also no coincidence that some of the successful companies are those who have real value and applicability in both traditional and advanced aviation. So take it all in and we hope you enjoy our discussion with Andrew Coleman who demonstrates that success includes both people and profits as you innovate for today's and tomorrow's customers in The Vertical Space. Andrew J. Coleman serves as the leader of General Electric Aerospace's Software as a Service Business. In this role, he leads a global team of aerospace professionals passionate about inventing the future of flight lifting people up and bringing them home safely. GE Aerospace's SAAS business supports the aviation industry with software solutions centered around the outcome of flight safety, efficiency, sustainability and reliability records management and transitions. As the business leader, Andrew has global responsibility for leading this dynamic and rapidly growing business focused on a threefold mission. Inventing the future of flight serving as a wingmate to both customers and GE Aerospace and building a team that prioritizes those around them ahead of themselves. Worth noting the business has observed revenue growth in double digits for four consecutive years while retaining over 99% of its clients and seeing less than 1% annual attrition across the employee base. Andrew has an MBA with a concentration in finance, from Xavier University, as well as a Bachelor of Arts from Asbury University where he now serves on the University's Board of Trustees. Andrew and his wife, Angie make their home in Austin, Texas, where they are the proud parents of Jack and Charlie. When not working Andrew is happiest spending time with his family, ideally in an airplane, studying leadership, volunteering in the community, advocating for autism, ADHD and mental health, or standing on the tee box with his family and friends. Andrew Coleman, welcome to Vertical Space. Great to have you on.

Andrew:

Jim, it's great to be here.

Jim:

First question. Is there anything that very few in the industry agree with you on?

Andrew:

There's forever and a day been a lot of language words, passion speech around the value of data. And yet when I look around the continuum of data we play in some cases. It's still viewed as paper, ironically, to be stored in a box somewhere in case a regulator needs it. To the other extreme, where data that maybe five years ago, 10 years ago, we thought, wow, there's a lot of value here, but there's no way we could get a regulator on board, or we could get a pilot's union comfortable with this. We've debunked that. And many of the use cases that people just said, no way, no how, and so I'm looking forward to unpacking that with you all, but in short, I just feel like this data asset that we've known is there for a long time, we're seeing a whole lot of value ranging from flight safety, something that's very top of mind for all of us right now, to sustainability to Asset Lifecycle Management, to even reliability, getting rid of that word delay and cancel when we go to the airport every day.

Jim:

So, Andrew, in some ways, if I am going to interpret your response properly, what you're saying is there's more value in the data than people realized. Is that right?

Andrew:

Absolutely. Let me see if I can run across the continuum with you for a little bit. I'll start on the record side. Forever in a day, a typical aircraft will generate, believe it or not, about five to 30, 000 pieces of paper a year in the form of a maintenance record. We've all seen this when we're flying, sitting in the, in the Fuselage and then see the technician come on and the pilot's got to sign a couple pieces of paper before we take off. All that was historically thrown in a box somewhere and just kept in the unlikely event a regulator needed to see it, or we needed to pull it. And yet today, because we've digitized all those records and through the power of things like a large language model, we can start to glean a whole lot of insight about what's really going on with that aircraft. And oftentimes we can get rid of a cancel, we can get rid of a delay, we can prove airworthiness in a second. We can look at a bill of materials of what's inside of that airplane in real time. On the other end, forever and a day there was the view as snapshot. Think of it like an old school Polaroid of every flight was good enough. And yet, gosh, in the 60s, British Airways said, I think there's a whole lot more that can be done with this device, the Quick Access Recorder, grabbing a couple thousand parameters, and we today would say that what used to be the Polaroid of a flight, we're now getting a high definition movie of every flight, and when you get a high definition anything, I'll think about the first time I saw golf on television in high def versus the old school. I could see like dimples in the golf ball. I see the word Titleist. I could see the blades of grass and that definition. That's the kind of definition we can now see in a flight. And when you have that level of definition, you can get down to a runway with a certain weather condition. You can get down to a landing gear that we saw land hard three times today. You can get into an APU and potentially how it's being used at the gate or not being used at the gate. All that fidelity that I think maybe people gave up on is now there. And the third area that I point to is airspace. Something that forever in a day, when you look at over the course of a day, a week, a month, a year, all the approaches, all the takeoffs from an airfield, you can start to see a whole lot of inefficiency. And yet forever in a day, people said, oh, the regulators will never meet with us and actually look at how we can raise efficiency here. And I'm finding that myth is just that a myth. Whereas 10 years ago, it was hard. It did require a partnership between airlines and regulators. Today, people are showing a willingness. Let's go have that conversation. Let's go work on this together to not only make more efficiency coming in and out of airports, but frankly, reduce the carbon footprint around airports like we see this right now, we're being done at Kuala Lumpur.

Luka:

Andrew, I like this description of this arc of digitization and, I wonder if we can spend a little bit more time on it. Because if I'm hearing you correctly, we as an industry have gone from having all these tens of thousands of pieces of paper that have been stored in multiple filing cabinets. And then you can probably describe the first attempt at digitization as merely scanning some of these and still having a quote, digital paper record, but not really having the full analytical power of this data. And so that if that's generation one of digitization, the second wave would probably be, making sense of it and doing some level of analytics. Are we still in that second phase? When did that start according to you? And what comes after?

Andrew:

I think you nailed it, Luka. The first phase was how do I effectively create a Google search capability on my records? So instead of sending Luka or Andrew into the warehouse to look through boxes, I could just type into a search bar APU, Boeing, General Electric, whatever I wanted to pull up and then in a nanosecond basically get that record or that series of records. Step two then was we started looking at use cases. What kind of things occur today that are fairly manual and onerous that tomorrow could be fairly automated? And one of the first use cases we landed on was as more and more aircraft are not owned by an airline but owned by a lessor, just like when we lease a car, that lease comes to an end. And our industry, as we know, to transition from owner A to owner B and have the intermediary of the lessor, we've got to prove that aircraft is taken care of and it's airworthy. And so we work with the records and we work with both the airline and the lessor, and in many cases, the MRO, to do that very thing, to prove airworthiness so that we could take the aircraft from operator A over to operator B. And what was taking about nine months, believe it or not a quote from one of our customers was it now takes nine seconds. It's a workflow that's automated, and it was at nine seconds, but as you can imagine, it's a whole lot faster now, proving, yep, that aircraft is well taken care of. Here's all the records. It's ready to go. Operator B, take it, fly it. You're good.

Luka:

What do you think comes next? And if this is now a digital repository of things where you can browse maintenance logs and prove the airworthiness of an aircraft. That still seems like a very rudimentary use of data. What's on the horizon? what are you excited about in terms of the next wave of digital?

Andrew:

Two things, Luka. Thanks to the power of optical character recognition, we're doing a lot of work with Microsoft. They've really invested here. We can take text, handwritten text that's in Mandarin. Handwritten text that's in German, handwritten text that's in English, as well as all the stuff in the quote unquote PDF. And we can start to look for things like, how does that match what's in the log page from the captain and the first officer? And then when that aircraft is reporting a problem, how could I quickly not three weeks later, but say three minutes later, get to a root cause of what the problem is based on all this new data I have available to me that before wasn't available to me. And so it's a pretty easy leap to say, all right, large language models were built with making sense out of text in mind. And again, when you start doing the math of how many aircraft are flying and each one of these aircrafts are generating anywhere from five to 30, 000 pieces of paper per year, we've got a lot of data to train these models very rapidly to start looking at when Andrew, me, I'm put in charge of air conditioning. Oh my gosh. When his fix goes back into production, you may want to double check it. When Luka's in charge of it, that thing's good to go and we never have a problem again.

Jim:

Could we just step back a little bit and GE has been involved with digital for a number of years, and there's been a lot of talk around the internet of things, many years ago, define digital aviation for us, Andrew, and from the big picture, what are the big hairy problems that digital aviation can address and which ones are they addressing today, just to give a good perspective to our listeners?

Andrew:

And so in the ecosystem of aerospace, we've channeled our energy first to operators, next to lessors and third to MROs. And inside that, to your point, we really said, where is our permission? Where can we move the needle? And we wanted to think big enough, but we wanted to think so big, it was ludicrous. And so in the space of full flight data and the space of maintenance records, we saw so many use cases that frankly got us very excited. When you contrast that to a technology company, who's looking at large addressable markets, they tend to look at retailers and banks and government agencies, and they get to our industry and we love our industry, but it's not the biggest industry. And so a company like Amazon, Microsoft, Google tends to not want to get down to that level of fidelity of what's aircraft data doing, what are these thousands of parameters I'm pulling from this aircraft, where they truly meet? And so back to your question. One of my favorite things of our job is GE gives me full permission every day to help every one of our customers with the fleet they have today, operate it far more safely. Our number one use case is actually working with teams around the world called the FOQA team, the Flight Operations Quality Assurance Team. Today, 55 percent of the commercial fleet is trusting us to decode full flight data and help them find patterns, events, things that would lead to potential risks that doesn't need to exist in our aviation ecosystem. And that's far and away our number one use case. And again, the thing I love GE lets me do on our team is this isn't just for customers that bought a GE engine. Or customers just with a Boeing aircraft, it's no matter what fleet they have, we're going to serve them exactly with that fleet. So we're trusted with every airframe engine combination you can think of. The bigger GE bet is by serving customers every day, over time, when it is time to acquire more, customers like the people that are there every day, helping them with their problems versus customers that show up every three to five years for the big campaign or talk about the new cool technology inside of our engine.

Jim:

So you mentioned the use case, you mentioned FOQA. How is that a use case more than it is a dataset that's being used traditionally for safety purposes, but what's the use case that the FOQA data is being used for today? And what problems is it solving for an airline?

Andrew:

So today we've, harnessed, worked about 55, 000 years of flight data. Every one of those flights we're analyzing through a series of events. I'll give a couple of basic ones. Unstable approach rapid taxi, runway exceedance. And so in each and every one of these flights, we're looking for any of the parameters we set up and these parameters evolve over time to be violated or miss, triggering a warning that we would then share with our customer. Hey, at this approach, at this time of day, with this kind of weather, we see a lot of crabbing of the aircraft coming in, a lot of unstable approaches. At this approach, we see a high risk of runway exceedance. With this aircraft, we're seeing this kind of risk in this kind of condition. Each airline that we can think of with any measurable size of tail. So I'd say over 10 tails today, and I'll make it very U. S. specific, has an apartment that in many ways looks like a group of actuaries at an insurance company. They're just looking at all these reams of data, trying to find needles, if you will, in the haystack. Of data patterns, things that would indicate, wow, we've got a potential unknown risk that because of this exceedance, because of this miss in this event, we need to change the way we do our training. We need to change the way we operate this aircraft. We'd change our standard operating procedure.

Jim:

So you're taking the FOQA data. Historically, I know once the aircraft lands, they take the FOQA data and they try to process it, they try to find the relevant information. Is this something you're doing post flight? Are you extracting it in real time during the flight? And then number two how is that information going back to to the dispatcher. How is it going back to the air traffic coordinator? How is it going back to the pilot? In what form are they receiving the information so they can do something different?

Andrew:

So to work through your question, great question, where a wireless quick access recorder, you know, you call it a QAR, exists, we can get the data into the hands of whomever needs it 10 so it's not real time, but it's not really far off from real time, because a lot of aircraft still have a quick access recorder that requires a technician to actually go with a stick to the aircraft and pull the data. In some cases, we'll have that data a day, two days, three days, five days out. And so there's a range from, I'd say 10 minutes to about five days is what we typically see in the range of how often this data is being harnessed. To your point, where we've really seen the power is not just having this data sit inside of the FOQA department for them to analyze and share with the chief pilot or share with the head of flight ops, but it's to close that loop and put it in the hands of the actual pilot. So in 2017, Qantas had the bold vision, hey, let's go put this data right in the electronic flight bag. So that our pilots can see both post flight, how did I do across the series of safety dimensions, but now pre flight, based on my track record flying this mission, what are things I ought to be aware of? What are things our fleet's seen? What are things I've seen to really prepare for each and every mission based on a series of historical data? One of the powers we've then been able to tap into is every pilot that we work with tends to be very type A, very driven, and very serious about their profession. And so when they start to see not only how have I done over a series of time, but we layer on top in a de identified way, how all other pilots completing the same mission have performed. It drives interesting behavior characteristics where I want to be the best, and if I'm finding when I land at Hartsfield Jackson Airport, I'm coming in more unstable than others, it may lead to some time in the sim. It may lead into this questioning how I approach that airport in the future.

Luka:

So, Andrew I think a part of your response, you were alluding to the things that Flight Pulse offers to the pilots on the dedicated app and I've taken a look because it, it sounded interesting and to your point, people can compare in a, post flight debrief type of scenario, how much fuel they used in landing, how much time to spend taxing in with an engine off what the touchdown distance was the pitch rates at takeoff, et cetera, et cetera. And so I can see the gamification aspect of it and how that would be engaging for a type A personality. But when you sell a piece of software like that to airlines, how does that customer quantify the value? How do they see and how do they realize that particular value.

Andrew:

I would say for 9 out of 10 customers, it's just a commitment to continuing to raise the bar in flight safety and not having to put an ROI, just knowing by the amount of events we've seen and the reduction in events that this was raising the bar in flight safety is good enough. However, I will say, we're seeing a lot of enthusiasm from Qantas, this was actually in their sustainability report, but also from a lot of our European customers, saying, I'd also like to use the capability and the power of the tool to lower my carbon footprint. And so we've built a series of dimensions in Flight Pulse to help a pilot see where carbon is created and opportunities they have throughout the mission to lower their carbon footprint. And so by giving a sustainability score at the end of every mission and then seeing that across 10, 15, 25 dimensions, however many the operator wants to show, we're also seeing CO2 reduction, which is a good thing.

Luka:

And how can these improvements in metrics, whether this is in sustainability or safety, how can that be clearly attributed to an app like that, as opposed to, a dozen other initiatives that an airline might have?

Andrew:

So what we're finding is I'll lean on research that came out actually from BYU about what a pedometer does. The second you put one of these things on your hands, whether it's an Apple watch or a Fitbit, science says you take about 500 more steps a day. Just having that visual reminder of how you did with real data. What we're finding with Flight Pulse is the same thing, by providing highly accurate personalized data in a very near real time way, because we think it's very important for pilots to see this within a week, not two months after I completed the mission. Here's how you did. And here's some of the things you could have done better. We do see pilots over time choosing to improve the way I taxi. A real basic example, performing a single engine taxi when appropriate. One of the powers of this high fidelity data is let's say there was a lot of rain on the taxi approach. That was not a safe time for a single engine taxi to occur. And so in that case, we're not going to punish the pilot for a bad sustainability score because they would have risked safety in doing so. And so that's the level of intelligence we put into the software.

Peter:

Are you sure that the data you're measuring can be connected directly to the overall end safety outcomes? If you measure a certain piece of data, and especially if you gamify it, isn't there a risk that pilots are going to fixate on maximizing performance against that data measure, but through that fixation, isn't there a risk that they're diverting their attention away from other aspects of the taxiing or the approach that might actually move you further away from ultimate safety? How do you know that you're measuring the right thing and that you're encouraging the right behavior?

Andrew:

We're up to about 30, 000 pilots using this technology now, so our sample size is getting pretty meaningful. And to your point, there was a season in our industry many decades ago where compensation was tied to fuel savings. We've intentionally done none of that in this work. And again, the decision tree is always safety events override any sustainability events. But as I shared, in our 30, 000 pilot sample size, I've not personally seen every one of these, but I've yet to see an example where that was risked or put in jeopardy. I feel like just the pilots we work with understand safety comes paramount to any kind of sustainability savings. And on the back end, I've not come across an airline yet that's saying, you've got to lower your carbon footprint at all costs. Every airline we work with is abundantly clear safety first.

Luka:

The 30, 000 pilots. What is that roughly as a market share? Is that about 10 percent of commercial pilots?

Andrew:

Exactly. The last study I saw was in 2023, CAE said 351, 000 commercial pilots, or about 10 percent of the commercial pilots.

Luka:

I understand.

Jim:

Luka, that's a pretty good guess.

Luka:

Yeah, come on, Jim, give me some credit

Andrew:

Not bad.

Jim:

Let's say we're going to go talk to the CEO of United Airlines and your pitch is, we're going to better mine, tell me where I'm wrong, Andrew we're going to better mine the FOQA data that's coming from the airplane in order to be able to improve the safety of flight. And it's presented back to the pilots in some delayed form, a week, perhaps longer, in some de identified way, and in order to be able to help them fly a safer route. Is that largely the value proposition?

Andrew:

In their case as with about half the operators, they've been trusting us to do this for 5, 10, 15, 20 years already. In the flight safety side, what's new is we're taking it to the cockpit.

Jim:

Let's say it's a Scott Kirby. Scott's going to say, well, that sounds interesting. I would assume a CEO is going to very quickly say, what's the return on investment for me? And sometimes safety's a tough sell. Before we move off the opportunities that would come beyond the FOQA data on the airplane, what's the pitch for justifying the expense, Andrew? And give us any numbers on, what's the quantifiable value, tough to quantify safety, and then what's the dollar amounts these guys are potentially paying, and what objections would you get from a CEO of a major airline to this kind of capability? Right. So the beauty is you get a 3

Andrew:

for one special with this data. It does start with the FOQA team and the flight safety department. It then quickly plays into as more and more airlines really care about how I get to net zero and not just talk about it or do marketing around it. It does start to play a role one to two percent of your carbon emissions can be reduced when you apply this technology. We've seen that. Qantas has been very vocal as an example in their sustainability report. So this isn't magical make believe numbers. They're real numbers we've seen. We then take this same data, because it's such high fidelity data about every flight, and help the tech ops team gain a lot of clues about what's going on with their aircraft. So all of a sudden I can start to predict failures. In many cases, before they occur. Failure is as basic as the example I used before. I saw three hard landings this week. Probably time to check the landing gear. Things around pneumatics, hydraulics, air conditioning, APU. Just all the things that drive us as passengers nuts. All these things are being performed with the same data set. So you get a three for one. On the case of price point with pilots, so I won't get precise, it's less than Netflix per month per pilot. So it's a very affordable price point.

Jim:

How much of the data that's being thrown off by the airplane? And I think you guys said by 2030 there could be six million terabytes of data that are being generated, how much of that data is coming in the form of the FOQA data? And then if it's not via FOQA, how else is it being generated by the aircraft? What other data is being generated above and beyond the FOQA data?

Andrew:

So your typical flight data recorder today, you're talking about tens of thousands of attributes. One of the challenges we have, we're starting to talk about the Internet of Things and the quote unquote edge, is the amount of time the aircraft's actually on the ground. And you start talking about downloading the terabytes you could download, just bandwidth is nowhere near there in that near real time fashion, so using a wireless recorder. It's absolutely there if you want to go take the technician and download the data that way. And so we have to be, part of our secret sauce, if you will, is to be very thoughtful about the couple thousand attributes we know Flight Ops and Tech Ops are going to badly want, we know is going to help the operator really understand the safety, the reliability of the aircraft. And so again, because we've been at this for a while, we know what attributes tend to lead to good prediction and good outcome and what attributes we leave alone. And then to your point, when do we go get those at another point in time or include those in another environment.

Luka:

Andrew, what exactly is the data that we've been talking about? Can you give us some concrete examples of the data that operators, maintainers, lessors can get out of these aircraft and who owns this data?

Andrew:

I'll work backwards, the stance we've always taken is we earn the right to work on this data each and every day. This is our customer's data, not our data. And so it's the operator, it's if the lesser owns the asset, it would be the lessor's data? but it's not our data. Again, we earn the right to work on it each and every day. And so this data Is the diagnostics, the health of the aircraft. One of the use cases that we are really proud to power is GE and CFM currently have 44, 000 engines flying all over the world each and every day. And so we provide as a service back to GE and CFM the ability to decode all of their engine data from the same spot, flight data, recorder. And then as a result, when we've got potentially an aircraft with a problem, we can quickly get to root cause on say a fuel nozzle and what's going on for that particular aircraft and that particular engine versus having to say, Oh my gosh, we've got a systemic problem across the entire system, across all of our engines. That power to get to root cause really quick. Our customers appreciate it. It keeps aircraft flying and allows us to keep our fleets of engines as reliable as possible. Well, also on the backend planning for maintenance. When we see things starting to wear down beyond cycles, we're able to plan for, all right, the shop visit, you're going to need these materials there in time. So the engine's not just waiting in the parking lot for all the materials to show up to be fixed.

Jim:

Is that engine data, Andrew, traditionally within FOQA data, or are you getting data, because you're the manufacturer, are you getting data above and beyond what would be traditionally in FOQA data?

Andrew:

Historically the engine business prior to us and the advent of our digital business would just use ACARS snapshot data. To my earlier analogy, what they quickly discovered is while that snapshot gives you a good picture in time of how the engine was doing, it doesn't give you the full picture. And so we're coming along as the complement with what we say here in GE is full flight data versus snapshot data.

Luka:

Andrew, given the massive amount of data that you described as a, as an HD movie of data being owned by the customer and broadly available to third parties who want to apply layers of software and machine intelligence of different sorts to derive some useful applications and value to the customer. Where do you see the opportunity is for large established incumbent organizations like yours, as opposed to, young, nimble, very tech savvy innovative startups.

Andrew:

Love that question. Let me hit it from a couple fronts. My background was about 20 years in the tech industry before the last eight years here at GE. And I honestly, when I started here eight years ago, did not appreciate the amount of depth you need to have to understand how an airplane works. And while there are a lot of brilliant data scientists in Silicon Valley and Austin, Texas, where I live and other places like that, all the physics of an aircraft married with this data science slash digital and now AI knowledge is a real powerful edge. And so the profile we look for actually, when we hire and interview engineers, is about 80 percent of their strength comes from depth in our industry and 20 percent depth in technology and depth in emerging technology. Because we find the technology is getting easier and easier to work with, requiring less and less domain knowledge, whereas our industry just begs of that deep domain knowledge. To your question though emerging use case that's been very exciting for me to be a part of is a couple years ago we joined the Digital Alliance. Airbus took to market a platform called Skywise, where they made a very large bet saying we're going to make it very easy for our operators to have a single place to store data about their aircraft. And by the way, much like our stance, they're aircraft very agnostic. Whatever their fleet is, Skywise is ready to accommodate it. They then approached us and said, we feel like there's a domain knowledge you GE have about how aircraft work. And we also think there's a domain knowledge you Delta have about how a world class operator runs an airline with, probably the best completion factors on the planet. And so while never easy to get three companies of this size together, we did work through all the things that needed to happen to form not only a charter, but an alliance, and now we have 10 customers live benefiting from our predictive maintenance alerts. And so in many ways, while it would be maybe unthinkable for these size companies that nimble and fast and agile. We're actually doing that every day and customers like SAS, customers like EasyJet, customers like Allegiant have been very vocal about what it's meaning to their aircraft reliability.

Jim:

Andrew, there's two questions I want to ask. One is, what's the alternative? If you were speaking to the CEO of United Airlines, and I'm not sure if they're customers of the product, what will they be doing alternatively if they're not using your product?

Andrew:

More often than not, I would say it's what we started the podcast with Is just accepting the status quo and letting this data live in a department for a specific use case, not get into the hands, in the case of the pilots, the people that could really move the needle and impact the results in each and every flight. What we've found is the paradigm used to be you'd find an event in the FOQA department, you'd go to the chief pilot, and then over time it may make its way down to the line pilot. We flipped that model on its head because now pilots pre and post every flight are getting direct feedback on how I'm doing. And often times they're now going to their chief pilot, hey, this is what I've seen. My neighbor actually flies the A320 for Delta and he pulls me aside all the time. Hey, here's what I'm seeing and here's what I showed my first officer on my last mission, what we can do better. We're really excited about just changing this paradigm of an insight being lock baby in a department or just shared at the highest levels. Now it's at the front line with the people that can really move the needle.

Jim:

Thank you, Andrew. Is there ever a gestalt here? Because you're working with 10 percent of the pilots, for example, and you're pulling all the data. I know you don't own it. But is there ever a greater value of working with the number of pilots you're working with? Can you aggregate it beyond the operator itself? And is there any insight that comes from the value of working with somebody one time, or are you just literally analyze the data from that one operator and the conclusions come from the operator, or are there any other greater conclusions because you're working with, arguably dozens of other operators. Can you use the data to bring about greater conclusions?

Andrew:

So because our industry has declared safety is just not an area we compete, we actually have user groups. I'm headed to Dubai in two weeks for our European Middle East one, have one later this year here in the Americas. Where we get together and do that very thing in a very informal way, but United to your example, we'll get up and show, here's what we learned this year. We want you all to understand and benefit from this, so our industry is safer. Beyond that the FAA has been a longtime customer of ours through the ASIAS program, which was created, gosh, over a decade ago to start sharing this data across operators so we could learn from each other. And so credit to the FAA to have this vision many years ago. We serve as the technology backbone for this, but they were the ones that said there's just benefit in bringing together all the data of our airspace, of our industry, and starting to find patterns beyond a single operator.

Luka:

Andrew can you tell a few stories from how the MROs are using data and how they're leveraging data to improve their operations?

Andrew:

Absolutely. And so, historically, an MRO would have a series of steps from induction to post inspection returning that asset, whether it be an engine, an aircraft, into the field. And yet, unfortunately, what we found, unlike a order you would place on FedEx or Amazon, where you could track every step of the visit, and you could pull at any point in time records for the parts being applied to this was done in a very ad hoc manual or paper based way. And so as a customer of the MRO, I had very little visibility into what was going on in my shop visit. It was more done with phone calls, or an email every once in a while, or a quick audit as to where things were. Now, just like the Pizza Tracker, just like the FedEx Tracker, I can go in Delta Tech Ops is one of our customers of this technology, and see at any point in the shop visit what's going on, what's been changed I can see a record to that, and I can see exactly when I'm going to get my aircraft, my engine back. And so again, it's just a level of visibility that didn't exist before. The ultimate use case or ROI point would be turnaround time. Instead of waiting and hoping and wondering when I'm going to get through, I see that turnaround time get fidelity and that turnaround time be honored.

Luka:

Have there been any notable success stories or use cases where the value has been financially so massive and significant. Implicit in my question is an observation that we have been hearing about digitizing aviation and mining data for, a number of years. And yet it feels like it hasn't lived up to its promise yet. And I'd like to dissect this phenomena. Is it because of the difficulty in, grabbing, processing data? Negotiating data rights? Is it about somehow feedbacking that back into existing workflows. Is it about the rest of the ecosystem that might not be ready? Give us some more granular insight into why this aspect of technology and aviation hasn't really accelerated in the way that everybody hoped.

Andrew:

Luka, I think you nailed it. All of the above of what you just said, and I will point to a recent event that is a very unfortunate one, but it's forcing some very fortunate change. And that's the AOG Technics event, where there was a very bad actor putting bad parts into the aviation ecosystem. One of our customers, when that challenge confronted them, because they had gone through the process of digitizing all their records, and making them incredibly easy to retrieve, Within 60 seconds, they knew across their entire MRO network where their vulnerabilities were, and they were able to get the bad parts out of the system. When you contrast that, others that had not gone to that step, in some cases it took months and potentially just over provisioning, taking good parts out to be safe and be conservative. And so that event, while unfortunate, really did quickly prove digitized maintenance records.

Jim:

I'm going to, I want to step back and we're talking to the CEO of a major airline and you gave them the value proposition, largely safety based. There's some efficiency components to it. Who's the person that you next want to speak to within an airline that's going to make the decision for this kind of capability, Andrew?

Andrew:

Head of tech ops, head of flight ops. Oftentimes the COO of the airline and on the back end of your point, the CFO want to know what kind of savings we can drive through efficiency, through reliability of the aircraft.

Jim:

Even though this safety is the largest piece of the value prop, is it largely paid for by operational efficiency improvements?

Andrew:

Yes. That's almost always where the business case is hung from

Jim:

So you're pulling the data off again, week or later from the airplane. And, we just had Lorne Cass on And he talked about choke points. He talked about, at the end of the day, in the United States, it comes down to New York and we want to be able to find ways to be able to improve the throughput of aircraft, let's say, in New York. Is there any possibility that you take the data that you have and you and in addition to being presented to the pilots via the EFB and the like, are there any other opportunities for operational improvements by using the data for other systems that are being used by the airline?

Andrew:

All the time. And so the airspace one is one we're very passionate about for some time now. We have had the Naverus business as part of our business, which goes way, way back looking at very efficient ways, not only to operate airspace, but to land in places that were historically very difficult or challenging to land. A lot of work many years ago with AirAsia allowed us to make the KOL or KL Kuala Lumpur Airport much more efficient, getting aircraft in and out more rapidly, but also cutting on the CO2 footprint. When you look at that airport, we tend to target a 90 percent efficiency rate. And I think not that long ago it was at 60, 63, something like that. That thinking we're applying to the state of Florida right now. We see some more challenge to what you mentioned in New York at many of the airports in Florida. We see some other airports here in the United States where there's an opportunity. actually just had a meeting this morning with a big operator in the Philippines and just seeing with the I believe it's 33 airports they have across all those islands, the tremendous amount of inefficiency that exists today and an opportunity to make that better.

Jim:

And is that through this product, Andrew, or is it a different product? Or are you powering another product with the data you're getting from the FOQA data?

Andrew:

Yep. Powering another product would be the way I would characterize it.

Jim:

And that other product is a GE product. It's called Airspace Insights It's interesting that you mentioned that at the end of the day, it's flight ops, it's tech ops that are paying for it through operational improvements, because that's how they have to pay for it. It's difficult to justify payments for safety only. So I guess you've addressed the operational improvements that a United Airlines potentially could get that would help to pay for the program.

Andrew:

Yeah, Jim, we had a first in the last year without naming the operator, we had to head of sustainability actually by the technology. She believes so much in what it could do to reduce their carbon footprint versus some of the other schemes we've seen, like offsets and that kind of thing, that was the buyer of the technology, not FlightUps or TechOps, which is a first for us, but we think it's not going to be the last.

Luka:

And Andrew, was this for the Flight Pulse product or something else?

Andrew:

Yes.

Luka:

I see.

Peter:

Hey Andrew, it seems like there is such a potential for correlating the data that you collect with the weather data at the time it was collected. How are you guys approaching that and what are you seeing in terms of the opportunities that you can derive from doing that?

Andrew:

We're harnessing full flight data from the aircraft, but think of it almost like making a cocktail. We're mixing in data we know about airspace, data we know about runways, data we know about the weather, data we know about the flight plan. All of that is also making its way into this environment.

Jim:

So let's say you have your whiteboard up on the wall and you're saying, here is where I see the next 10 years of GE Digital Aviation. What data would you want and what problems would you like to be able to solve that you're not doing today that's part of your plans that you can share with us?

Andrew:

Yeah, the number one thing we're working on solving around that is, is we've discovered, again, what was initially intended for the FOQA department. Progressive maintenance departments are dying to get their hands on this data to do a lot more work around fleet reliability. And so our more progressive operators have rolled up their sleeves and trained their teams on how to use this because it's a relational database. It was built with a belief you have to be fairly technical and know how to write standard query language, SQL, to retrieve data. Thanks to the power of what we're now seeing through our partnership with Microsoft, I see very quickly the ability, just like you ask Alexa at home, tell me the weather for me to tell our EMS environment, our measurement system, tell me about the last 10 flights into Jackson with A320s where we saw an unstable approach. And just that easy. Verbally say it. We just have to know how to write all that language and pull that data out. We think when that becomes reality, the amount of users in this environment goes from, our typical customer will see anywhere from 10 to 150 users, to potentially 300, 400 users. And not only those 400 users, but that data making its way to the head of flight ops, head of tech ops every day.

Luka:

What's FAA's perspective on the use of some of these AI tools, LLMs and other methods in touching, data that ultimately has real safety implications. How do you think about dealing with incoherence and hallucination of some of these LLMs?

Andrew:

So, the overarching challenge has been, we'd like you to step up innovation GE, not slow it down, I really applaud that. I think it'd be easy to say, slow down, let's lock it down and be very conservative. Part of that is the track record we've built. Being at this for over 20 years, there's a lot of credibility that comes with. We've not had, knock on wood, the big events you read about, newspapers, of data breaches, of lack of privacy, lack of security. We never take that for granted. However, because we've earned that permission, we find the FAA saying, we trust you, GE. You're not going to, in the software vernacular, fake it till you make it. When you bring us something, you're going to harden this, you're going to make sure it works, and it brings real value. But as you do, we're ready. Please, bring on that kind of thinking and take advantage of the technology that's out there.

Jim:

Andrew, let's, let me go back to the 10 year whiteboard. Let's say you're 20 times the size you are 10 years from now. And. a sizable value enhancement has been made based on, whether it be FOQA data or the integration of weather data, going from a week post ops, to potentially real time, to potentially predictive. What would have to happen for you to be 20 times the size you are today?

Andrew:

We have to get the data off the aircraft quicker, which technology is already starting to address that. We really like without overhyping it, the eVTOL space, we see a tremendous amount of opportunity first to make sure that mode of transportation is very safe, but also over time they're going to need the same data to make sure that these aircraft are reliable and able to be dispatched the way they're being sold. And ubiquitous, almost like an Uber. We like the role we get to play in sustainability. We think we are sitting on just so many clues today: how to help an airline lower their carbon footprint, not a hundred percent, but gosh, if we can knock down one, two, three, four, 5 percent of the carbon created in this industry, that's a pretty big move in the right direction. We also like, frankly, the opportunity we see in the military side, something that's very early on, but we see a tremendous amount of opportunity for an aircraft to land. And let's just say within 15 minutes, the technician knows exactly what I need to fix on that aircraft for the next mission, not let me go look at a bunch of logs and diagnostics and then call the supplier and order some parts, and maybe it'll be ready next week.

Jim:

You would think that maintenance piece would have a terrific return on investment as long as they're doing something different than they would otherwise do today. You would think that would have a good ROI. And let's talk a little bit about the eVTOL. Who is the customer? Is it the OEM? Is it the operator? And then how would you get the data and what's the value proposition to the customer that would buy that data?

Andrew:

Without breaking a lot of confidentiality, it is the OEM today. We're working with all the players you would expect. And a lot of them early on are asking the question, do we want to build this capability ourself, or do we want to lean on a GE who's done this in commercial aviation for quite a while, to be our partner? And we're finding about 80 percent want to lean on a trusted brand like a GE in the history we have, versus do it themselves. There's still a lot of smart technology people at these eVTOLs that want to do it themselves. And so 20 percent are saying, Oh, we think we'll just do this ourself as a core competence.

Jim:

And let's say it's the head of, let's say, Joby and you're talking to him and you're talking about your value. What's the pitch to the head of Joby and how it would be different to then the head of, let's say, United Airlines?

Andrew:

So an important segment of our business today is business jets. And while oftentimes it's hard to track the quote unquote value of safety, what we can show with a business jet is from the day they start with us to five years later, on average, 80 percent of their safety events are reduced. And so let's say in year one we see, I'll make this number up, 50 safety events, by year five we'll see 10. Just by again, harnessing this data, putting it back in the hands of a pilot to show when you fly into Teterboro, here's what we see going on these missions. And so I take that over to the eVTOL space, again I feel like, a critical point in adoption is going to be as a passenger knowing I'm getting on to something as safe as, or safer, than my car, my bus, my boat, any other mode of transportation.

Luka:

So, in order to fully exploit data for better maintenance, whether this is better visibility, or more predictive and proactive maintenance, or, anything else for that matter, but what's the impact on, or disruption on the existing practices, on the existing processes, workforce, logistics, warehousing. In order for this technology to really reach its potential, it obviously cannot be done in isolation, so what's the impact of the larger ecosystem and how quickly do you think it can move to, to embrace this promise?

Andrew:

So ironically, a lot of the impact isn't for the future aircraft coming, although we're very excited about that. It's the reality, given all the challenges we know about, and I place an order for an aircraft today, it's a commercial example, I'm 20 30 something before I can take delivery. And so accepting that reality, I've got to take the aircraft I have and make sure they're flying longer than maybe I anticipated. And in doing that, it's begging of us. How do we find more and more clues to help through a summer schedule where I thought it was going to have brand new A320s and I'm actually going to fly a legacy model for a couple more years? And so we actually like this challenge a lot. How do we keep the fleets our customers have today knowing the life is going to be extended on them in the air longer? Doing a lot of projects right now with customers you would imagine even heading into this summer, knowing the fleets a little older was anticipated to be refreshed a little faster. How do we keep these fleets as reliable as possible as we see just unbelievable demand heading into the summer?

Jim:

When we talked to Lorne Cass, former VP of the Integrated Operations Center of American, and we said, what's the greatest operational challenge you have to address? It was predictability. He needs to be able to better predict what's going to happen. And there's all kinds of things he can do with better predictability. The second one is just real time. What's going on. So Lorne, going back to that discussion said, surveillance is really important. I need to know where the planes are for a variety of reasons. The third element is. we'll call it, there's a POET tool out there, what is it? Post Operations Evaluation Tool, so POET. And people have said, listen, there's all kinds of value of historical post operations information in order to be able to better plan the operation in the future. How do you think an airline sizes up the value of those three elements. And where are you within, I assume you fall into the poet component to it, but how do you think they size up the value of the three? And do you see yourself moving into any one of the other two buckets?

Andrew:

So the predictability one, we feel like we jumped in head first with our Digital Alliance. With Airbus and Delta for what you just described. And I'll quote David Thompson from Delta Tech Ops. I need to know, and I thought this was really profound. I need to know yesterday what I need to run my airline today. Not six months ago, because then I'll overstock and just do a lot of crazy things, but also not in real time. Oh my gosh, this part's falling apart. This problem's arising. I need to know yesterday what I'm going to do with my airline today. And so that was the problem. We've really been hovering around for the predictive maintenance work we're doing with the Alliance. But absolutely, to your point, while we live in that third leg of the stool today, we're going to absolutely need to live across all three if we're going to be the trusted holistic provider we're aiming to be and frankly, keep the seat we've been honored with. I think if we just stay in the last one, we're probably relegated at some point.

Jim:

Now let's, you know, for our listeners who have technology that may apply to advanced air mobility, it may apply to the airline. What advice would you give to our innovators on the podcast who have an interesting technology they think would apply to the airline, but they've never sold to an airline before. What advice would you give them?

Andrew:

As much as you're taught to have a BHAG and to think gigantic, I would think small to go big. We tend to, in our world, think about a problem for I'll point to Qantas. They were very passionate putting data in the hands of pilots is going to help us raise flight safety. We had a strong hunch there was repeatability, but instead of day one, let's go sign up 50 operators and really go at it. We really stayed in stealth for about 24 months to make sure that thesis was true and to make sure we run the trust, not only of pilots, but also the pilots union and the leadership of Qantas. So what we were doing was good for the airline and not creating an environment where we saw unintended consequence or people trying to win on one metric to advance another metric. That's tough if you're in a VC world where you're asking for money from investors. But as you know, you've worked in this industry for a long time as well. The amount of credibility that is given to, you really move the needle at a Quantis, honestly and I don't want to diminish our sales team because I think the world of them, they do so much selling on our behalf post that experiment that I don't know that had we rushed to market and tried to go solve this for 50 airlines at once, we'd sit here today with 30, 000 pilots. As we talked about earlier, the addressable market in this industry is not as big as banking or telco or government or retail. And so that's why you don't see a lot of tech startups go pure play into the aviation space. We had the benefit because of who we are, GE Aerospace. This is the industry we care about. And so with that, I get a lot of freedom to just focus on this industry versus Andrew, could you be a transportation business and go help boats and help cars and help trucks and other things that move. Nope. We help the aviation industry.

Luka:

Andrew talking about other players who are in this space, do you know how active Palantir is in this area? And the reason I'm saying this is obviously their mission statement is to, and I'm paraphrasing here, but enabling better decision making by, analyzing data They've been very successful serving the government customers in this and increasingly so, corporates. And there have been initiatives between Palantir and Airbus with Skywise a number of years ago, but I haven't heard any PR or any good case studies coming out of Palantir that are discussing this problem and said, unless I missed it how active are they and how do you make sense of either outcome of this question?

Andrew:

Yep, I was actually at their offices a couple weeks ago. We hosted our board meeting for the Digital Alliance at the Palantir office in New York and to your point they are the technology backbone for the Skywise platform the platform We're building all of our predictive maintenance capability against. So we've gotten to know them very well, and being very direct, one of the decisions we've all made is, we have deep domain knowledge in how aircraft work, and the problems of our industry, whereas Palantir has deep domain knowledge on how to build a platform that ingests data from a lot of different places using a framework, they'll talk about it as their data ontology, making it very easy to not only build that environment once, but keep that environment current and bring new data sources in. And so we feel like we're getting the best of all worlds the intellectual property, the thinking, the depth of a Palantir on the platform, married with our depth on the industry, our customers and the problems they care about. So very familiar and very integrated with the work of Palantir.

Luka:

How quickly do you think a third party can get up the learning curve on that domain knowledge? And again, I'm thinking of Palantir and how they started with no pedigree of DoD, it was a startup, but they quickly got up that learning curve. And arguably, somebody like them could probably do the same for aviation.

Andrew:

Absolutely. We take it very serious and we're very humbled to that and very paranoid to that. But also as I've gone through growing, scaling, and hiring our business, as I shared before, the depth of knowing how aircraft works, and not just looking at parameters coming off of an airplane, I can't emphasize enough how much that served us in serving our customers. Early on, we did experiment with just pure data scientists that didn't know how the industry worked. And in many cases, it was just a bridge too far. Some cases they were nimble enough. They did learn the industry. And I'd say that about tech companies we do see. But it's not as easy as you would think.

Jim:

And I would assume when you're talking about managing FOQA data, having a trusted partner like GE who knows aviation would be a big deal. Now, if Palantir could augment what you're doing, all the better, but being GE I would think would be a big advantage.

Andrew:

And we see the same, I'll put it on plug for Microsoft, see the same thing with Microsoft. While they know optical character recognition incredibly well, they don't pretend to know what kind of things we're actually looking for, from a maintenance record. And so that's where we come together. We say, here's what we're trying to do. Can you unleash your brilliant engineers to help us do it a lot faster with a lot more languages? So we're finding the tech companies want to build platforms that scale across all industries, and then putting that domain knowledge of the industry on top is something we're enjoying doing, Airbus is enjoying doing, but we don't see a lot of tech companies wanting to go all the way to that end.

Luka:

Andrew, are you exploring any use cases in the unmanned aviation domain and autonomy and perhaps leveraging data to inform the development arc of autonomy?

Andrew:

The short answer is yes. It's another reason we like the eVTOL space a lot because of the things they're trying to do over time without getting too precise. Yes.

Jim:

Can you get more precise? Now this is interesting. So, let's say, Wisk from Boeing brings about, from Boeing, brings about an autonomous vehicle in the next five, ten years. How much of their bringing about that unmanned eVTOL will GE be a part of, and why?

Andrew:

It's really OEM to OEM in this space. Some of them are asking, GE, we want you here every day. Help us with all of this. Some without naming any are saying, we got this. We don't need any help from anybody. We just want to do this all on our own because we think it's so unique, proprietary to our business model. I'm headed over to Europe this weekend to meet with one of the eVTOL companies early next week. That's been very much of the GE come in here and help us with all you got. But there's others that have said, nope, GE, we've got this. Thanks, but no thanks.

Jim:

How much smarter has your capability become? I loved your answer, by the way, to the, our innovators on the line of, how to approach the market. Go with one really deep and prove the value. Boy, if everybody could listen to that's great advice. But let's say you, you developed a capability with your product through Qantas. How much smarter did it become because now you worked with 20 other airlines? Or was the value proposition more or less the same as you went to each incremental airline?

Andrew:

The value proposition was effectively the same, but to your point, the argument could be, well, Qantas helped fund a lot of the innovation. What's unique about Qantas, and you guys know this, is they're incredibly proud of their safety record and what they've done to make our industry safer and safer. So in this case, frankly, it wasn't a concern of theirs. The other trade off is in tech speak, when you're an early adopter or part of private preview, so before a product is generally available you tend to get the best, the deepest, the richest engineering team possible because we want to get it right. And so the benefit they gain is they're getting our best, brightest, deepest, and get to drive a lot of the requirements for the product versus if you're in more of the mainstream or late adopters. You're getting something that's hardened that works great, but you're not able to influence the roadmap nearly as much.

Luka:

So, Andrew, what's the most common criticism that you have received? One that is maybe, the fairest as it relates to your efforts around data.

Andrew:

The most common, and I thank GE for letting me be this proactive, is whose data is it? And very early on, we landed on, this is not our data. We are in the right to work on it each and every day. The second common criticism I get a lot is, why should I trust an OEM with all of my fleet data, not just their engines. And again, from day one, this data is our customer's data. This data is not shared in a secret way with our engine team. So we can see how the other engines are working, this sort of thing. And never for a second have we violated that and we never will. And so those tend to be the two biggest objections. And again, I'm just really proud of working for a company that said no, just serve the customer with the fleet they have and the tools you have to the best of your ability every day. And by demonstrating that over time, when it's time to grow and add, I think it puts us in a pretty unique place.

Luka:

What innovations are on the critical path to making digital aviation, live up to its fullest potential. You mentioned better connectivity. What else?

Andrew:

I want data faster. I would like to widen the aperture to get more attributes off of the flight data recorder, because I think as more and more of these attributes are becoming available, while we may not find these things, I do believe in the power of artificial intelligence. The I believe with time, we'll be able to glean a whole lot more insights than we have historically. And so to me, it's getting the pipe bigger, making the latency lower, and then really leaning into our relationship at Microsoft as an example, and the work they're doing with not just occupational character recognition, OCR, but also AI.

Luka:

When startups reach out to you for partnerships or, other commercial agreements, what are those conversations typically look like? What would you call out as, good or bad strategies?

Andrew:

We love working with startups, and I frankly spend a lot of my time working with emerging technology and four or five people with a really good idea. Where we tend to have a very grounded conversation is early on, they want us to put a GE logo on it, which implies a level of quality we're really proud of, and take it to every airline around the world. Back to our use case, our example before, we try to start with one customer. And one very discreet problem before we go promise or tell the world, Oh my gosh, look how neat this technology is, or overhype something. And so again, in the tech speak, we're probably too conservative. We're not hyping it enough. We're not faking it till we make it, but that's something we're proud of. And so that's the number one thing when we see a neat idea come our way is if there's a true interest to channel the use case to a single operator. Knowing there'll be repeatability down the line, we'll go for it.

Luka:

Excellent. And just to close it out, what's one point that you'd like to leave our audience with?

Andrew:

I think the biggest point I'd love to leave our audience is so many people join our industry with the passion for I want to be a pilot, and I love and respect pilots a tremendous amount. And then they get into a Embry Riddle or University of North Dakota, to universities that actually have our technology, not only to monitor their fleets, but teaching classes on it. And they start to fall in love with the data and analytical side of our industry and feel like they're at this crossroads. Do I want to go work in the digital world or do I want to go be a pilot or be a technician? And I think that decision doesn't have to be so acute anymore. I believe you can have an incredible career working in our industry, but working with the data, the insights that are just now starting to make themselves available. And so I would tell our audience, no matter where you are in your career, if this is something that raises your interest and raises your curiosity, give it a shot. I am surrounded by a team of wingmates that just feel like we have found our calling on this earth and I don't care what industry you're in, when you're that passionate about what you do and you see directly every plane I know that I see going over my home, more often than not, it's one of our customers and I take a level of pride knowing that landing's going to be a little safer and those passengers are going to get to where they're supposed to a little more reliably and we're having a hand in that.

Jim:

It's great, Andrew. Thank you for your time and I appreciate everything you've done for us and If there's no other questions from the team, we thank you for being on.

Andrew:

No, thank you guys for what you're doing with this podcast. It's a really groundbreaking podcast with some great thinkers and ideas I'm honored to have joined.

Luka:

Thank you,

Jim:

Thanks, Andrew.