When the Procedure Runs Out: Mission Readiness for Aircraft and Facilities Maintenance

This transcript was produced through automated speech recognition and has been lightly edited for readability.

Opening and Introductions

Claudia Hosky: Hi there, I'm Claudia Hosky from FedInsider, and on behalf of the George Washington University College of Professional Studies, Carahsoft, and Legion Intelligence, I'm pleased to welcome you to today's webinar, When the Procedure Runs Out: Mission Readiness for Aircraft and Facilities Maintenance. Please type your questions and comments into the Q&A pod. If we don't get to them during today's webinar, we'll share them with our speakers so they can follow up with you.

Also, take note of the resources tab, where you can download today's slides, and below that, you can join in the chat. We offer live captioning for all of our webinars. To follow along, use the hyperlink under the live captioning or CART tab located above the slides. We'll have the full transcript available for download in the next day or two. To earn training credit for today's webinar, you'll have to answer all survey questions and stay online for the duration of today's hour-long conversation.

And now let me turn over the reins to our moderator, John Breeden from FedInsider. Take it away, John.

John Breeden: Thank you so much, Claudia. And hello, my friends. Good morning, and thank you for joining us. I'm John Breeden, and today we have a really interesting discussion about what happens when a maintenance problem reaches the point where the published procedure no longer provides a clear answer. Whether we're talking about a downed aircraft or an urgent facilities problem, maintainers suddenly need to pull together information from multiple systems, records, and even people before they can decide what happens next.

When mission readiness or critical operations are at stake, every hour spent reconstructing that picture can make a big difference in the outcome. So today we're going to look at how organizations can shorten the path from an open maintenance problem to an informed, approved action, while keeping the ultimate decision where it belongs: with the people responsible for the mission and the work. It's going to be a really fascinating show. We're going to combine technology and physical operations, and I'm really looking forward to it.

So let's meet our amazing guests for today, and then we can begin. First off, I'm very pleased to welcome Brian Dosa, the Director of Public Works at Fort Hood. Brian, it's a real honor to have you here. Thank you so much for joining us, and welcome to the show.

Brian Dosa: Good morning, everybody.

John Breeden: Excellent. Next, please welcome Chief Warrant Officer 4 Reginald Oliver, the Chief Digital Transformation Officer for Portfolio Acquisition Executive Maneuver Air with the U.S. Army. Chief Oliver, it's an honor to have you with us today. Welcome.

Reginald Oliver: Hey, it's great to be here. Good morning, everybody.

John Breeden: And finally, I'm very pleased to welcome Nick Weir, Vice President of Mission Engineering at Legion Intelligence. Nick, I know how busy you are this time of year, so thank you so much for making the time to join us and talk with our audience today. Welcome to the show.

Nick Weir: Thanks. I'm really excited for the conversation.

John Breeden: I am too. Well, thank you all for being here. We have a lot of interesting things to cover, and I'm looking forward to getting some practical lessons from this great group today. Those of you who join us on these regular FedInsider shows know that we often divide the shows into three sections. We do level setting, to make sure everybody in the audience is on the same page about the topic. Then we talk about some of the key challenges. And then comes my favorite part of the show, where we get some great advice from our speakers.

For today's show, however, we're taking in a lot of real-world circumstances and mixing that with technology. So I want to do something a little different in the level-setting section and ask our government guests to walk us through some real-world examples of maintenance challenges they've faced, including what went wrong, what information the teams needed to gather, and what ultimately helped them solve the problem and fix the issue. I think hearing some real-world examples will let us all get our heads around what the people in this field face every day. And I think it'll be interesting, too.

A Barracks Air Conditioning Work Order, From Report to Repair

John Breeden: So, Brian, let's start with you on the facility side, if that's okay. Could you walk us through a real-world maintenance situation at Fort Hood, or anywhere else you've worked in maintenance, that shows what your team deals with when something goes wrong? How was the problem identified and reported to you? How did you prioritize it? And ultimately, what did you have to do to resolve the situation?

Brian Dosa: I would love to do that. Thank you. Greetings from Fort Hood, Texas. We are an Army post located in Central Texas, about an hour north of Austin. We're the largest Army installation. I'll have to admit there are a few more soldiers at Fort Bragg, North Carolina, so if any of you are on from Fort Bragg, I will acknowledge that. But if you look at landmass, we are definitely the biggest.

Here on Fort Hood, we have 2,500 buildings that have a roof and need some sort of air conditioning in the summer and heating in the winter. That's in addition to our family housing. You can imagine Central Texas temperatures are pretty warm in the summer, and one of the biggest challenges we have is just keeping the air conditioning running across the post in those 2,500 buildings.

To focus in on a particular type of building, let's take our barracks. If you're a soldier serving at Fort Hood, you're in the rank of sergeant or below, and you're not married, you live in a barracks. Think of a dormitory; it's a similar type of facility. Here on Fort Hood, we have 100 barracks buildings and about 11,000 soldiers who live in them. Keeping the air conditioning running for those soldiers, so that when they're not out at the range training or maintaining their equipment in the motor pool they have a place that's comfortable, safe, and healthy to live in, is a big challenge for us.

So let me talk you through a vignette on a hot summer morning. It's finally cooler in Texas now, but let's say it's a July or August morning, and the temperatures are projected to be in the hundreds at Fort Hood. A young soldier, let's just say Private Dosa, wakes up in the morning and the air conditioning in my room is not working. That's a little frustrating, because I know that later today I'm going to the range, it's really hot on the range, and I would really like that room to be cool. So I'm going to take the next step and report that.

We have a couple of different ways to do that. I could use the old-fashioned way and call it in. We have a telephone number for a 24/7 work reception desk. We also have a program called the Army Maintenance App, which is an app for your phone or smart device. If I was smart, I would have gotten myself an ARMA account and I could submit my work order through ARMA. So let's just say I do it that way. I send a work request through ARMA saying that in barracks 1004, room 123, the air conditioning is not working.

Whether it's a telephone call or ARMA, that request goes into our work reception desk, which is manned 24/7. They're going to take that work order, enter it into our business system, and assign a priority to it. Work orders are prioritized as priority one, which is an emergency, a life, health, or safety situation; priority two, which is something urgent that could become an emergency if it's not addressed; or priority three, which is just routine.

An air conditioning issue in a single room in a barracks would be a priority two. So it gets assigned priority two and then sent out to one of our shops. In our DPW, we have an in-house team of government employees, just like me. They're typically wage-grade carpenters, plumbers, electricians, and HVAC mechanics who are actually fixing problems. So that priority two work order would be sent out to, let's say, our central shop, and then handed to one of our technicians who's come in to work for the day.

He or she is going to work priority ones first. But let's say that on this particular day they don't have a priority one to work, so they receive this priority two for the air conditioning being out in Dosa's room. They get in their truck, get their tools together, and head out toward the barracks. We have some habitual relationships that we've formed with the units in the barracks, because Dosa has probably done physical fitness training, taken a shower, eaten breakfast, and is heading out to the range. But we've got some folks around the barracks who can provide access.

So the technician is going to show up at the room, and hopefully he's going to fix an easy issue on the spot. But if not, and he needs to order a part, he's going to put in a work request for that part. We have a supply team in our DPW, about a dozen folks who use government purchase cards and spend their entire day buying repair parts and other supplies. That part would be ordered, and then the technician would circle back and make the repair. That happens many times every day in our 100 barracks at Fort Hood to fix air conditioning issues. So hopefully that was helpful.

John Breeden: That's amazing. I can imagine. I have some friends who live down in Texas, and I know it gets into the hundreds. So I'm glad that guy's problem was at least a priority two, because it could become a priority one. He comes back, he's all hot, and then it's not fixed. That could be bad.

Brian Dosa: Definitely.

An Apache Phase Inspection and a Line That Wasn't Torqued Correctly

John Breeden: Well, thank you so much, Brian. Chief Oliver, let's move to the aviation side. Can you give us a real-world example that illustrates how complicated aircraft maintenance can become when the answer is not immediately obvious? What information does your team need to bring together, and what decisions need to be made, to bring an aircraft back into operational availability? Do you have any interesting stories you could share with us?

Reginald Oliver: I absolutely do. First, good morning, everybody here in cyberspace, from Redstone Arsenal up here in Huntsville, Alabama. It's nice and sunny today; it was rainy yesterday. Before I begin, I certainly want to put out that any anecdotes are personal to me and not representative of the United States government, even though I am in uniform because I have another meeting to go to after this. And of course, I cannot obligate the military to anything. Now that that's out of the way.

The great thing about Army aviation and the way we manage sustainment and readiness for our end items, whether it's an Apache, aviation ground support equipment, or anything special used to support the aircraft, is that there are a lot of moving parts and pieces. It's a huge ecosystem. That runs from what we do over here in aviation acquisition at Redstone, which is modernizing, training, and equipping the aviation enterprise, to getting all of these end items and support equipment through the wickets so they get to the front and to the end users: the great men and women everywhere from Fort Hood, Alabama, to Fort Bragg, North Carolina, to Iraq, Syria, Afghanistan, wherever we may have been or are.

With all of those moving parts and pieces, when we go through the sustainment phase of the asset's life cycle, there's a lot we have to do. I'll anecdotally give you a story about what we do with all aircraft, but specifically Apaches, because I'm an Apache guy. I'm a maintenance test pilot and maintenance examiner, and of course I've got to show my airframe some love.

With our airframe, every 500 flight hours, you have to do what's called a PMI, or phase maintenance inspection. Essentially, if you're a maintenance manager or fleet manager, you get the opportunity to coordinate multiple 18- to 22-year-olds from all over the country, from different units, different organizations, and different backgrounds, to completely rip this helicopter apart, inspect all of its parts and pieces, replace them if they have to, and then put it all back together.

When I was at Fort Bragg a long time ago, starting to become a maintenance test pilot, I got trained by some really great mentors. They said, "Look, man, you have to make sure you understand the regulations. You understand all of the tasks that have to be done to take this entire helicopter apart and put it back together. And by the way, you have to manage these 18- to 22-year-olds, some are older, and you're going to have to make sure they understand the tasks as well." It's like, okay, no big deal.

Well, the issue with making sure they understand the tasks is that sometimes people misinterpret things a little bit. So I had to make sure, when I was walking them through these tasks, that the inspection was okay. But what happens when somebody misinterprets a task for a flight-critical component? I experienced that. Once we started putting the helicopter back together, some of the components were bad. We replaced them, cleaned them, and put them back on. No big deal.

But when it was time to preflight the aircraft and go through some of the ground checks, I noticed a couple of anomalies, as was to be expected. Then when I did the ground run, I got some cautions and some warnings, so of course we shut down the aircraft. When I got out, I saw things leaking. I could smell fuel and hydraulic fluid. So we brought over our ground crew, and my ground tech said, "Hey, everything's fine. It's just leaking a little bit, no big deal."

Well, it turned out that when we put the engine back in, even though it got signed off and preflighted, something came off because it wasn't torqued correctly. So how do we work through that? The manuals tell us exactly how you're supposed to do it. Unfortunately, sometimes the manual might be incorrect, or somebody might have misinterpreted the way it was supposed to be torqued. That, anecdotally, is one of the ways we work through getting better at repairing.

What I had to do was deal with a line that wasn't torqued correctly. I explained to the guy, "Look, the manual says you're supposed to do an arc torque." For those out there in cyberspace, an arc torque is a creative way of measuring how far you're supposed to turn something to tighten it. They didn't do that. They used inch-pounds, which is another way to measure how much force you have to put on a component to tighten it.

The good thing is that I had a senior maintenance manager with me, because again, I was a young MTP doing something like this for the first time. And my mentor said, "Hey, listen, man, yield to the expertise of the folks who have been doing this a long time. You've got some senior sergeants and some senior maintenance managers who can help develop and train you."

The thing is, when we do maintenance and support sustainment for these aircraft, everything is in black and white. I'm sure Brian can attest to this too. There are lots of regulations in the military, and lots of regulations in management, whether it's facilities or aircraft. But here's what's cool: everything's black and white, but there's tons of gray where they overlap. That's where it comes down to domain expertise, interpretation, and the true understanding that comes from the experience of those around you.

Because I didn't know. I knew the black and white, and I was explaining to the guy working on the aircraft, "Hey, this was wrong and that was wrong." What I didn't recognize was that one person had misinterpreted something, and I, too, had misinterpreted something. So I had to yield to the expertise of someone much more senior than me, who was not an officer, who said, "Look, sir, I get it. This is what the rules are. But this is how we actually need to go about fixing this aircraft and getting it back to service."

John Breeden: Wow. That's really fascinating. I can imagine doing your ground check and starting to see fuel leaking. That's not a good sign. Great story. Thank you, Chief Oliver. Appreciate it.

Reginald Oliver: I got a little sporty. Appreciate you.

What the Two Examples Have in Common

John Breeden: Nick, those are two very different examples, but they had a lot of common challenges, and I'm sure the clients you work with experience similar things. What are some of the key things you heard in those stories that line up with the challenges people face in going from establishing their priorities to getting to an approved action, whether that's getting the air conditioning working or getting that helicopter safely flying again?

Nick Weir: There are a couple of things that really jump out at me from both Brian and Reg's examples. One is just the sheer volume of different tasks that maintainers need to be able to perform, and the complexity of it, whether it's someone who supports a lot of different kinds of hardware in a facility or someone responsible for maintaining something as complex as a helicopter. The amount of information and knowledge you need to have, or at least be able to pull on, to complete these tasks is daunting from the outset.

There's also the fact that, as Reg mentioned, the manuals may not be entirely accurate at times. For many of these large vehicles and aircraft, the manuals are being updated twice a year or something like that. If you're in an austere environment, maybe you don't get the update as soon as you should, and then all of those maintainers are responsible for understanding these changes against what they had historically been doing.

Layer on top of that the fact that, as Reg mentioned, many of the folks doing these maintenance tasks are relatively early career, or just aren't there for very long. These are often stints of a couple of years. If you think about needing to take a 10,000-page manual for some vehicle or piece of hardware you need to maintain, and then internalize and operationalize it, that's a really daunting task.

The other piece is the degree of integration it takes across a lot of different elements. Both examples involve folks with different levels of experience, and coordination between the prioritization, the maintenance supervisor, and the supply folks who make sure you have all the parts on hand. It takes a lot to keep a facility running or keep an aircraft in the air. These are complex challenges, and obviously there's a lot of great process that's been built to make this as efficient and safe as possible.

There's also a lot of opportunity to surface the right information to the right folks at the right time and add some of that historical context. Maybe that exact same problem has happened on that exact same helicopter a few times, and you could help folks who may not have been maintaining that helicopter identify the issue and shorten the cycle to repair. So that's a little bit of what jumps out at me.

John Breeden: Absolutely. I have a lot of questions for all three of you. Thank you so much for playing along and doing something a little different in the level-setting part of the show. I thought that was really interesting, and I'm sure our audience appreciated it as well. Speaking of our audience, I'd like to invite all of you watching the show today to submit any questions you might have for our amazing panel. Clearly these folks know what they're doing, so if you have questions in this field, these are the people who can get you the answers you need.

I'm going to try to get to some of your questions live on the show today, but if there's not enough time, we'll round them all up and give them to our speakers afterward. So don't miss your opportunity to get some questions in. Thank you to those of you who are already submitting questions, and for the rest of you, get typing. We want to hear from you. Now that we've grounded the discussion in some real-world maintenance situations, which I thought were really fascinating, let's dig a little deeper into why maintenance problems can become difficult to resolve.

Why Reconstructing the Maintenance Picture Takes So Long

John Breeden: Nick, let me stick with you on this one. One challenge I heard from both Brian and Chief Oliver was that the maintainers, the people actually doing the maintenance, and the supervisors have to pull information from many different systems and records before they can begin the next step. And in Chief Oliver's case, sometimes the institutional knowledge of the people doing the work might be a little different from what the manual actually says. So why does that reconstruction process consume so much time? And where do you most often see information gaps or handoffs breaking down as you work with your federal clients?

Nick Weir: That's a good question. One thing we do see is that the systems that folks in different roles within that process use are often not as connected as you might want them to be. That makes sense given the way we have always performed maintenance and the process we've established. You have the supply depot, with really talented personnel who understand how to prioritize to get the right supply to the right place at the right time and keep a facility as operational as possible or an airframe in the air.

Then there's the maintenance manager or maintenance supervisor's system, where they task all of the different individuals to perform the repairs. And then there's the tool or system a maintainer might have, maybe on a tablet in their hand, to support them when they're doing those repairs. Those have always been relatively siloed systems. There are certainly great examples where a lot of really hard work has been done to integrate them.

But in the setting we're in now, where you can take advantage of AI-powered capabilities and there's a real push to integrate across a lot of these systems, there's a lot of opportunity to make sure the right context gets to the right place. That way the maintainer might know, "Oh, this screw keeps coming loose on this air conditioner, and that's causing it to fall apart. What we really need to do is go out there next time with a new screw," as a very simple example.

When you don't have that context, the maintenance supervisor may have a record in their system that they've repaired this thing four times, while the maintainer only has the manuals and the information they need to fix it that one time they're out there. And then the supply folks know what parts they have to make available to the maintainers. Being able to tie all that information together gives you a much better repair process, one that requires a lot less work from everyone involved and enables you to maintain readiness better. That's something we've seen our clients look to use AI-powered capabilities, built specifically for the maintenance case, to support.

John Breeden: I can imagine that in a field like that, there's a real opportunity for AI and automation to bring some of that together. Thank you for pointing that out, Nick. I'll have some more questions for you about that in just a bit. I think that's a really fascinating area.

Getting Small Problems Reported Before They Grow

John Breeden: Brian, let me ask you. Listening to your story, I'm assuming one of the challenges at Fort Hood is actually getting information about a problem to your technicians. If a soldier wakes up in a barracks covered in sweat because his air conditioning isn't working, I'm assuming he's going to report that right away. But what about smaller maintenance problems that might go unreported until they eventually snowball into larger ones? Why does that happen, and what can you do to improve the reporting process so the users are more in touch with the maintenance people, especially on such a large base as Fort Hood? I think you might be muted, Brian.

Brian Dosa: Thank you. Hey, John. That's a really great question and observation, and it's exactly right. A lot of times, the biggest challenge for us in fixing facility maintenance issues is actually learning about them. If we don't learn about a problem when the first evidence of it appears, it can become a larger problem very quickly, and much more expensive and much more impactful on the occupant of the room, the soldier, or the unit trying to use the facility.

Probably the best example of that is water inside buildings. Inside a building, the only water that should be present is in a toilet, a sink, or a water heater. Any stray water that's not reported is very problematic. Let's say you have a small roof leak and there's a drip, or water coming down your wall. Another common example is a water leak underneath the bathroom sink, where you might not even see it.

If that leak isn't addressed quickly, the water will keep leaking and saturate the wood or the sheetrock, whatever the case may be. Before you know it, we're growing mold, and it becomes a much larger, much more expensive problem to fix. That happens way too often.

So yes, reporting problems is a challenge for us, particularly because Fort Hood is a very transient organization. Our soldiers are here for two or three years. During that time, they probably deploy, they go to the field, and then they move to a different location and new soldiers come. So educating soldiers, and civilians and contractors too, on how to report problems to us is one of our core functions.

We do that through media, social media, messaging, and simple posters on bulletin boards in facilities like barracks, to constantly get the message out: if you see something, report it. You can do it by calling this telephone number, or by getting an ArMA account and doing it on your smartphone or device.

John Breeden: It reminds me of the civilian side. If you're in an airport, they say if you see a bag or something out of place, you need to report it. So you're telling your people, if you see a leaky faucet, report it.

Brian Dosa: Very important, yes.

John Breeden: Absolutely. That's interesting.

Rewriting Technical Manuals for a Technical Force

John Breeden: We're getting some really great audience questions, and Chief Oliver, I want to bring one of them to you. This comes from U.S. PACOM. The person says they help write training manuals, and they find that "dumbing it down" often helps: telling people step one, do this; step two, do this. They wanted to see if you have a similar opinion. Do you think some of those technical manuals need to be rewritten or dumbed down? Or what do you see as the best way of transferring that institutional knowledge to people who are just getting into the field, like the 18- and 19-year-olds you were describing, who are just getting started in this career, so they can make the right decisions?

Reginald Oliver: Thanks, John. That is a really great question, and thanks for the question from PACOM. I know there's a lot going on out there. Sometimes SOPs have to be rewritten, especially depending on the environment, because different environments affect our systems in different ways. Not just the physical systems, but the way we interact with those systems and the way we use them.

So what I would say is this. A lot of the technical manuals we use are written by engineers, and not everybody can understand engineer-speak. I believe there's merit in making sure the technical manuals are explicit enough that the task is conducted correctly to get the system back to its manufactured state. But I wouldn't say dumbing it down is necessary, because the great thing about our armed forces men and women out there maintaining and sustaining our equipment is that they are a phenomenal group of folks. Our maintainers are extremely brilliant. They really do understand what they're doing, and they have to do these things under very adverse conditions.

But making it step one, do this; step two, do that, so that everyone understands there is a process by which the maintenance must be conducted step by step, is absolutely paramount. I completely agree with that.

So to our person out there at PACOM, as you're rewriting your SOPs for training and exercises, here's what I would ask. I had an opportunity to write an SOP for Expeditionary Sea Base's third class when we had attack aircraft out on the USS Lewis B. Puller in the Northern Arabian Gulf. Shout-out to all those folks on the Puller, an absolutely phenomenal bunch in the Navy. But there wasn't a manual for us to execute attack aircraft operations off of an ESB.

So we had to talk through what the domain knowledge was for the Navy, the domain knowledge for the Army, and the domain knowledge for aviation operations in both the Navy and the Army, because they are very different. One of the things we had to do for our SOP was make sure the terms we were using were the same, because the Army and Navy might speak a different language. I mean, the Army is much smarter. Go Army. For all the Navy folks out there, it's just a joke.

But seriously, we had to make sure our language was the same and that everybody understood what a widget was, or what a screw was, because it might mean different things. I say screw, and it's something really small that you twist into a panel. To the Navy, a screw is on the bottom of a ship, you know what I mean?

So whenever you're rewriting your SOPs or doing training exercises, I wouldn't say dumb it down. I want to impress on anyone listening that it is extremely important, especially in the digital age we are encroaching on, where the way we receive information is much faster. As our Secretary of War says, we have to move at the speed of relevance. The people who are going to use your SOPs and do your training, and everybody, need to recognize that maintainers have become much more accustomed to a propensity of information very quickly, and they can process lots and lots of information very quickly. So while there might be some instances where we should use layman's terms, I think we've got a very technical force. So just consider that.

John Breeden: Absolutely. Thank you. I love hearing your perspective, and I'm sure our audience member did as well. Thank you for addressing their really excellent question.

Separating the Authoritative Record From Inference

John Breeden: Nick, this dovetails with what you were talking about earlier. When information is being pulled together from all these different sources, like maintenance records, previous repairs, and manuals, how difficult is it for the maintainer to distinguish between what the records actually establish and what a system or a person is inferring? Why does that distinction matter, especially when someone is being asked to act on a recommendation?

Nick Weir: That's a great question. If we go back to the example of the fuel leak from the engine that Chief Oliver mentioned earlier, you may have a very experienced maintenance leader who knows that 90% of the time, this is the cause: the seal is cracked in this one spot within the engine, or something along those lines. But if you look at the historical maintenance records for that particular engine, you might see that one bolt has not been torqued properly in the past, which may have caused some cross-threading, and that's causing repeated issues within that engine. Then you might take a very different approach to repairing it.

So having the ability to integrate what people know from their past experience, and what they're inferring from that experience, with the actual authoritative record of what has happened for a given vehicle, or what was reported by the person requesting the repair, is really important.

When a lot of folks think about using newer technology or AI to help solve some of these problems, the AI is not going to actually solve the problems. The maintainers and those very talented individuals are the ones who are really keeping the birds in the air. Folks assume that if you're using AI, you're relying on the model to make a guess or an inference.

But particularly with a lot of the newer agentic systems out there, these systems can go and get you actual authoritative information to help drive your decision-making. I think that's where we're finally getting to the point where you're no longer relying on guessing, or worrying about ML hallucination, when you're trying to support tasks where you really need to be backed by the facts and understand what is inferred and what's not. That's where a lot of the opportunity is now, as we try to build more confidence in systems that can help maintainers, or folks in other fields, where the facts can impact life-or-death decisions.

John Breeden: That makes a lot of sense. Thank you so much for that.

Open Communication With Leaders When Resources Are Tight

John Breeden: I'd like to move into the practical solutions and advice section, because this discussion is fascinating and we could talk about the challenges for a long time, but I think getting some great advice from you all would be most helpful for our audience. So let's do that. Brian, let me start with you. You talked about some of the basic challenges of maintenance operations, the blocking and tackling.

What practical steps have you found most useful in making sure that problems are reported clearly to your staff, that work is prioritized properly, so that air conditioning system gets assigned priority two instead of priority three, for example, and that the information and parts get to your people so the repairs are done efficiently? What practical steps have helped you that might also help the people in our audience facing similar challenges?

Brian Dosa: That's great, John. I think my mute button is not on this time. We always have more problems and more challenges than we have resources. That's true at Fort Hood and wherever you all are this morning. We're always in a resource-constrained environment. So one of the things we work on is making sure we have good engagement and good, open communication channels with leaders.

I'm the DPW. We're part of the city staff of Fort Hood, which we call the garrison, and we live to serve the soldiers, units, and families stationed at Fort Hood. So we keep open communication with those leaders because, yes, you're exactly right: we don't always prioritize the work orders that come in correctly. They aren't always reported to us correctly or characterized the right way.

That's why leaders get paid to watch their soldiers and their facilities, and then engage with me and the leaders we have in DPW when something doesn't seem right. We've got 350 people here at DPW working hard every day to serve those soldiers, units, and families, trying their very best, but we don't always get it right. So we always tell folks to communicate with me and our senior leaders and ask questions if something doesn't seem right, because we have the ability to redirect some resources, or look at a problem a different way than we were looking at it, to get something solved.

John Breeden: That makes sense. Thank you so much. Appreciate that, Brian, very much.

Turning Maintenance Data Into Usable Information

John Breeden: Chief Oliver, let me ask you a similar question on the aviation side, especially based on the stories you've told us, which are really fascinating, by the way. What can maintenance organizations do to make sure the people making the decisions have things like aircraft history, repair history, parts information, what's available and what's not, technical guidance, and maybe the broader fleet context, before they have to get in there, make some repairs, and approve the next action for the aircraft they're working on?

Reginald Oliver: Great question, John. With access to the copious amounts of data we have for our aircraft, there are a couple of ways you can approach this, and it really depends on your position and where you are in the maintenance chain. If you're a maintainer on the flight line, on the ground, in the trenches, you're going to need access to some very detailed information about the system.

We have what we call technical inspectors in the quality control office. They're the ones who are going to have your aircraft historicals and lots of other things. But in the logbooks for all the aircraft, the vast majority of the maintenance data you would need to do an inspection or a task correctly is already there for the maintainer. That's on the front line, in the trenches.

From a fleet management perspective, the type of data is different and where that data comes from is different, especially in this digital age, with the advent of AI and the implementation of artificial intelligence for sustainment of systems, and especially as we work through how we can use that for predictive maintenance. I say this a lot when I go to different conferences: I know we use the term data a lot with these maintenance managers, field commanders, and senior leaders. Data, data, data. Big data this, big data that.

But what I've started to recognize, at least personally, is that a lot of senior leaders who have to make very strategic, large-swath decisions don't really care about data. I say that tongue in cheek, because it's not data we care about. It's information. Data points are data points. At the end of the day, data points have to be distilled into actionable information that's usable to make decisions. We have to focus on that.

We have lots of different data points, from the aircraft status, to the supply chain that's going to sustain the aircraft, to how we're going to continue to manage the operations we're responsible for. What does the data tell us? How do we distill it down into information? At a strategic level, we have to recognize that we have different sources of record.

As Brian said, and as Nick mentioned, we have all of these different sources of truth and systems of record that we have to pull information from and consolidate, and these are all disparate areas. What we have to ask ourselves, especially as leaders, whether in commercial industry, government, or the military, is how we take all of these disparate data points, consolidate them, and distill them down into usable information, so we can continue to gain and maintain the advantage when we're using our systems. I hope that's helpful.

John Breeden: It's super helpful. Thank you, Chief Oliver. Appreciate that.

Where AI Realistically Helps Maintainers

John Breeden: Nick, I want to ask you about AI again. In the last section, you mentioned there's a lot of opportunity to use AI and automation to support decision-making. Where can those technologies realistically help with the work? Obviously, they're not going to go out and turn a wrench for you. But maybe you could give us some examples where you've seen AI and automation work.

And could you explain to our audience what a good AI solution in this area would actually do for the maintainer, rather than simply giving them another tool or another interface to manage? What should it have to actually make their jobs easier and make maintenance more efficient?

Nick Weir: Absolutely. I think it dovetails really well with what Chief Oliver was just talking about. We're in a phase of technology development now where more and more data is being made available for all the different aspects of the maintenance life cycle, as Chief Oliver was describing. Many times the data is so overwhelming that it's unrealistic to use. You don't know how to operationalize it and turn it into the information people would need to make an intelligent, informed decision.

So that data either doesn't get used, or creating a solution that transforms all of this new data, maybe all of the telemetry coming off an aircraft integrated with historical maintenance records, takes a lot of work with historical automation technologies and approaches before it becomes something you can use.

That's where I think AI presents a lot of opportunity: the speed at which you can create a new solution by tying together different data points that are all referenced back to their sources. That gives you confidence that you're getting factual information backed by the data you're receiving. Then it brings all of that information together to paint a coherent picture for the individual doing whatever task they need to do, whether that's prioritizing different repairs, maintaining the right level of supply for different parts, or turning the wrench and making the intelligent decision about what to check first.

That's where AI presents a lot of opportunities: bringing all the right data together, transforming it into the information you need, and giving us the ability to harness this ever-growing data landscape much more quickly than robotic process automation, which is the term used in the industry for a very regimented, coded-out process. Ensuring you're getting the right data to the right person to back the right decision is where AI can really help here.

John Breeden: That makes perfect sense. Thank you for explaining that for our audience; I think that was very helpful. I'm looking at the clock, and I can't believe we're near the tail end of our time today. But we have enough time left that I can sneak in one more question to each of our amazing guests before we have to say goodbye.

Keeping Preventive Maintenance on Schedule

John Breeden: Brian, for a final question to you, first off, thank you so much for being here. It was fascinating listening to your stories. I want to circle back to something you said about working with constrained resources, because I think that resonates with almost everybody today. And I want to bring back the concept of preventive maintenance. I'm guessing preventive maintenance can sometimes lose ground when you're constantly responding to breakdowns, urgent work orders, and all the tier one and tier two things that keep coming in.

So how do you, and how can people in our audience, make sure that important preventive maintenance still gets done consistently when day-to-day repair demands are competing for the same attention and resources? As you said, you may not have the resources you need to do everything. How do you make sure preventive maintenance doesn't get overlooked?

Brian Dosa: That's a great question, John, and definitely a pressure that public works directors in the Department of War face, probably in the civilian world as well, and probably units and soldiers too: services versus repairs. For us, we've tried to make it simple. We focus our in-house DPW team on breakdown maintenance, and we have a separate contract for preventive maintenance. We've found that's a really good way to discipline ourselves to make sure we're doing preventive maintenance, because it's so important.

For your own cars and vehicles, for Army equipment, and certainly for facilities, buildings, and the mechanical systems in them, you've got to do preventive maintenance. You've got to invest in that. So again, our in-house team focuses on breakdown maintenance, and we have a contract that does our preventive maintenance.

John Breeden: That makes sense. It's a good way to separate them. Appreciate that, Brian, and thank you so much for being with us today. It was really fascinating having you.

What to Preserve as Maintenance Modernizes

John Breeden: Chief Oliver, we've talked a little about bringing in AI and automation tools as part of modernization. Everywhere across the government, everybody's modernizing, and I'm sure maintenance organizations are too. As that happens, what is the most important thing leaders should preserve about the way experienced maintainers make decisions and support mission readiness, even as these new tools come in? Is there anything people should keep in mind so they don't throw the baby out with the bathwater?

Reginald Oliver: Absolutely. And again, I had a great time here today, so I really appreciate you all having me. I would certainly say that domain knowledge is always going to be the number one thing you have to maintain. Foundational knowledge of your craft is most important, because while we have lots of technological innovation in this space, these are just tools. They are not the end-all, be-all. They're just things to increase productivity, improve workflow efficacy, and make sure things are more accurate. That's really it.

So how we use and implement these tools in our actual workflows is important. So is how we adjust those standard, everyday workflows we're accustomed to, transforming them into workflows that include AI, automation, or other workflow improvement injects. We have to figure out what that's going to look like, but ultimately we have to make sure that foundational knowledge, understanding systems and systems of systems in maintenance and repair, remains key.

The other thing is data literacy. Data literacy is huge, because at this point, if you're not on the train, you should probably start running, because it's leaving the station. The genie is out of the bottle with the drop of artificial intelligence models and now harnesses. So I tell everybody that data literacy, for any industry, is the number one thing people have to focus on.

John Breeden: Absolutely. Thank you so much, Chief Oliver. I can't tell you how much I appreciate you being here today. It was really a joy.

Reginald Oliver: Of course.

Starting an AI Pilot With a Specific Outcome

John Breeden: Nick, that means you get the final word on the panel today. Thank you for being here; it was great hearing your experience as well. As organizations look at using AI and other decision support technologies for maintenance, what is the one thing leaders should keep in mind if they want to make sure those tools are integrated properly, improve mission readiness, and make life easier and more efficient for the people doing this really varied, really complicated, and really critical work?

Nick Weir: That's a great question. Thank you, Brian and Reg, for chatting with us here. I've learned a ton over the course of the conversation, so I really appreciate that. I think the most important thing when you're thinking about integrating AI, or newer technology in general, into your processes is to really think about the problem you're trying to solve. There's a lot of "I want to bring AI to my workforce" and "I want to modernize my workforce or my capabilities."

The challenge a lot of the time is that you get some compute, you get some AI, and then what? That's where a lot of pilots, and a lot of organizations attempting to adopt AI capabilities, have really struggled. Where it goes much better, where organizations succeed, is when they identify the problem they want to solve with that pilot at the outset.

What is the specific outcome we want to achieve? Do we want to improve fleet readiness? Do we want to improve the efficiency of our maintainer staffing? Then go into a pilot where you're testing against that very specific outcome with the set of tools you're being provided. That gives you the opportunity to evaluate the system you're bringing in much more effectively, and it lets you pick the right tools.

There are so many different technologies out there in the AI space, and the breadth and scope of what they can do for you is growing every single day. If you come in with the specific problem you're trying to solve, not only does it ground the organization and help it identify where to start that transformation journey, it also gives you a much more effective way to evaluate the utility of the technology.

John Breeden: Perfect. Thank you, Nick. That was a great way to wrap up the show. Really appreciate that.

Closing Remarks

John Breeden: Thank you all for talking with me about this really interesting and really important topic. I learned so much today, and I'm sure our audience did as well. Thank you to our superstar speakers, Brian, Chief Oliver, and Nick, for joining us and sharing your experience and perspective with our amazingly attentive audience. I really appreciate you all being here. And thank you to our amazing audience. As always, I'm so glad you're all here with us today.

If you liked today's show, which I really hope you did, please check out our scheduled future shows. We have so much great content coming up right around the corner, and I hope to see all of you there. Now let me turn things over to FedInsider's Claudia Hosky for some final thoughts.

Claudia Hosky: Thank you, John, and thanks again to our panelists. To our audience, thanks for your participation. Training certificates will be emailed after the webinar ends. And don't forget to log back in in the next day or two if you want to download a copy of our closed captioning transcript. As John mentioned, we have many webinars in the works, including another one today at 2 p.m. Eastern time, so we hope you'll join us for that.

Visit FedInsider.com to see what else is coming up and sign up for multiple webinars at the same time. While you're there, you can check out our podcast, beds at the edge. And with that, this is FedInsider, signing off.

‍