The Vertical Space

#68 Don Berchoff, TruWeather: Weather imposes uncertainty and uncertainty costs money

• Luka T • Episode 68

Share episode

Welcome back to The Vertical Space! In this episode, we chat with Don Berchoff, CEO of TruWeather, about the critical impact of weather on aviation. Don highlights how weather-induced uncertainty costs the industry billions and shares staggering numbers on preventable costs. He discusses how aviation has adapted to unreliable weather data and the potential for significant improvements in flight operations and cost reduction.

Listen to Don's insights on the three key elements for better forecasts and the importance of data. We also delve into the challenges and needs of weather forecasting for drones, eVTOLs, and advanced air mobility broadly. Despite a high attrition rate due to weather, the aviation industry's willingness to invest in better systems remains a challenge. Luka and Peter push Don on whether the industry can afford to invest in improved weather systems. 

Don:

So the three things that make a better forecast are data, computing, because you need speed. If you do have the data and you want to run a one or half kilometer resolution model, it's going to take you computing resources, and that costs money and time, so you always have to worry, too, about turning these out fast enough. What good is a forecast if it comes out after the event. And the third one is the science. There's a reason why we can't forecast tornadoes perfectly, and there's still a lot of science. There's still a lot of things to learn. Even in icing, we have a lot of mysteries. So it's those three, but data is the key to improvement.

Jim:

Hey, welcome back to The Vertical Space and our conversation with Don Berchoff, CEO of TruWeather. So I was really looking forward to this conversation with Don. Because as he says, weather imposes uncertainty, and uncertainty costs money. We've known for years, the cost of weather on traditional aviation. And listen up to the kinds of numbers that Don shares as those costs and what percentage of those costs are considered preventable. This discussion is essentially about what's possible in weather. Don says that aviation has gotten used to the lack of weather data and it's gotten used to uncertainty and that there are ways to improve weather today and in the future. Improvements that could dramatically improve flight operations and reduced costs, according to Don. Listen to what a weather pro says about the three things most needed for a better forecast. And according to Don, how data is key to improvement. We get into a discussion of weather on smaller things that fly. Drones and eVTOLs and other vehicles in advanced air mobility. Now Don talks about the need for better low altitude and micro weather and the need for better data for micro weather. He says that with drones, there's approximately a 40 to 50% attrition rate due to weather. But here's the challenge. Even though everyone says they need better systems in this case, better weather systems how much are they really willing to pay for those systems? This is one of the great challenges of selling technology to aviation. And Luka and Peter really challenged Don here. They know advanced air mobility is really thin profit and margins. And questioned their willingness to pay for better weather. But as Don counters though, they can't afford not to have better weather. So take it all in and many thanks to Don for joining us and to our many guests, we hope you enjoy our talk with Don Berchoff as you innovate and The Vertical Space. Don Berchoff United States Air Force Retired is the co-founder and CEO of TruWeather with 40 years in aviation, logistics and weather industry solving, impactful weather problems for the government and businesses. Don built and led a regional weather center that supported worldwide tanker and airlift weather operations and the provision of weather warnings and forecasts at 138 Army and Air Force installations. He was the Base Commander at Manas Air Base, Kyrgyzstan responsible for all ground operations at the Air Force's main staging base for Afghan operations. Don was a Senior Executive Service member at the National Weather Service, where he led the complex transition of$500 million in science and technology projects, overseeing a multidisciplinary team of scientists, data, scientists, and software engineers. At TruWeather, he brought V360 to market an advanced micro weather data and analytics platform producing custom weather insights to serve the emerging autonomous drone industry. Don led the development of a new aviation weather standard under ASTM F38 with the FAA, opening the door for internet of things, weather reports to support aviation operations for part 135 and 91 operations. Don started TruWeather in 2015 to unlock the power of science and technology in the nation's labs and universities and accelerate new capabilities into commercial operations to better serve businesses. Don Birchoff, welcome to The Vertical Space.

Don:

Well, thanks for having me.

Jim:

What's, something that very few in the industry agree with you on?

Don:

Well, what I'd like to say, it's not that they don't agree. I think it's just a lack of understanding about, you know, what's possible with weather. And, so I, you know, when you asked me that question, I think of two industries. I think of my weather industry, which is, my home. And then I think of the, aviation industry, which is who we support and who I'm part of. The weather industry, I think, is in agreement with most of what we're going to talk about today, that there's a lack of low altitude weather data, that impacts predictions and that impacts uncertainty. I don't think anyone would argue with that. We might argue with how to close that gap and what's the right way to do it, but I think most people understand that gap. On the aviation side, the challenge that I've run into is most of your aviation folks who have been out there operating for lots of years have really gotten used to the lack of weather data. And they've kind of built it into their mindset or into their business model. So they're used to flying and dealing with the weather and overcoming uncertainty. And they really, sometimes they looked at me five years ago when I first entered this industry. And I said, this is going to be a challenge that for drones and air taxis, the lack of low altitude weather. And, they said, well, why are you trying to? You know, solve a problem that doesn't exist. That was actually quotes. And you know, it actually did exist, but they were used to it and they didn't understand what was possible. What I'm starting to see now is as folks are flying more in the industry, they're realizing, you know what, we're starting to see and understand the economics around weather data more than the safety component. You can always get around weather by not flying. And, we, we generally had about a 30%. of the time when manned aviation or crewed aviation didn't fly due to weather. They could have flown, but they didn't think of it that way. And so that's really where the mystery is going to get unlocked here. We're going to show them that you can, make more money with better weather data. We're going to show them you can reduce your risk. We can reduce your uncertainty. We can increase your passenger comfort. And you don't have to live with what you had because technology is changing.

Jim:

Don, why is there the difference between those who know weather and the aviation communities interface with weather? Why is there a difference?

Don:

It's a really good question. so I'm going to bifurcate the weather industry a little bit for you. You have those that are scientists folks that are actually in the business of producing forecasts. And they really do understand the things we're talking about. You need more data to make a better forecast. That sounds like common sense, right? But it really isn't as obvious to everyone as you might think even in the weather industry, because we have, the business side of weather. And the business side of weather, came out of really the 1980s and 90s. You know, the government had been the monopoly in weather and it had to be. There was really no options. And what weather companies did initially is they would take that government data and they customize it or make it much more impactful to a decision. In the beginning we didn't really change the forecast. Other than the human would look at the data and have this interpretive skill and heuristics that they can make it better. And that was the customized weather. People paid for that because it was better than the forecast. But then we started adding more data, right? We started getting satellites up there that were commercial satellites. We've added more weather instruments that are, privatized. we have more commercial data. but the business model has always been, and still is, focused on taking data, through apps and advertising, right? So advertisement, I always consider in the weather industry, we have a conflict of interest. The conflict of interest is, if you're in the business of providing data on apps and you're selling advertising, you have a motivation to have people click your apps more, which means now you want to, make data more hyper exciting, right? And then you have the customized piece, which isn't scalable unless you change the technology, around it, which is what TrueWeather's doing. And we have folks that like the current business model. And they don't want to admit that there's a lack of data, because if they do, it says we have to change the business model, which would disrupt the weather business models that depend on apps and button clicks and just forecast or customize data. So there's that middle piece there. So I'm bifurcating, a little bit. Now aviation, I think the challenge is weather's very complex. And, you know, when I was in the Air Force, I was on Air Force bases with F 16 pilots trapped at Kunsan, they didn't want to hear about my excuses about weather. They just want to know why weren't you right? Right. And, there's a tendency to try to simplify this whole process so they can make a decision that they feel comfortable with. And I think, you know, when you start explaining that you could do more with this data or that data, what I've seen in the industry is that they worry about having to pay for the data now. See, they like free data and the problem is free data is not free because it costs money. So to go back to, the edict here, in 1998, the cost of, weather delays by major airlines on the U. S. economy was 41 billion a year. And this is lost productivity, this is, decisions that cost money, it's inconvenience to customers and clients, that number is much bigger today. And to try to explain to the aviation community that's been built in the current system that there's an economic argument to be made, they don't really see it because they're not worried about the bottom line as much as the CFO is of a company.

Luka:

The cost of weather delays, the 40 odd billion per year, how much of that is due to low altitude weather?

Don:

This study I'm talking about was a 1998 Boeing study that said that they actually said the 41 billion dollars was the cost of weather on the U. S. economy due to delays by the aviation industry. This is what got my attention, they said 28 billion was avoidable. That was the, that's what got me interested in this problem. Because that told me that even back then it was the way decisions were being made that was also causing some of this uncertainty that led to, you know, this 28 billion. So now if you look at today and based on what I've learned through my work in the Air Force and studies that I've done that have been validated, 30%, roughly 30 percent of, I, I think is recoverable, and the percentage of that due to low altitude weather, you have to bifurcate it down now and say, okay, that's general aviation, because remember, big aircraft don't fly at low altitude, right? They come into airports, they're well instrumented. So you gotta look at this and say, who's impacted by it? It's helicopter pilots, it's general aviations, and now it's gonna be drones and air taxis. And what I'm seeing right now based on, I can't really talk about specifics, but we're getting about a 40-50 percent weather attrition rate, meaning cancellation rate right now for some low altitude operations that are happening with drones. And my estimate is about half of that is due to lack of low altitude weather. And about half of that's due to just decision systems that aren't able to take this data and present it in ways that help to make a better decision, right? I don't want to get into those details, but that's about the way I see it right now.

Luka:

Yeah, I'd love to uncover a little bit more of this because it's intriguing your comment about how there is perhaps still pushback from the aviation side that, solving for this low altitude weather is, not really worth the effort, right? Yet the impact is monetizable, some of the studies that you quote. And so is that because really where the money is, the weather is not as big of a factor on the low level, But the majority of the impact then becomes on GA and yet this is an area right now where the cost of not flying is not as high, because if you're a GA pilot, you look out the window and you decide that you won't fly because of weather and it's not like there is an economic impact as a result of it for the most part, right? So can we maybe dissect this business problem a little bit more with some examples and some additional feedback and commentary that you're hearing?

Don:

Sure. So, let's take the helicopters, for example, right? If you talk to Rex Alexander, who is a significant advocate within the helicopter community, who also is trying to work with, you know, Oh, good. So, yeah, so Rex and I, you know, we have these conversations all the time. He's worked very hard to increase the ability of the, emergency helicopter HEMS, right, Helicopter Emergency Management Services, to improve their, readiness and their ability to respond, and do that in a way to try to improve weather information through a system that has been developed by, I think, NCAR and the National Weather Service have developed a system by which they try to better pull together the data we do have and try to better inform whether a helicopter should take a mission or not. Because remember, if there's an accident on a highway and you know, you have to decide whether you're going to send an airlift out there or you're going to bring an ambulance out there what you want to make sure is that if you're going to send that helicopter, it could get to where it has to go. And many times they will not take a mission because, remember, most flying helicopters does not occur at an airport. The airport's the only place that we measure weather right now. You don't get the measurements between airports. People don't understand, like, cloud height, visibility, measurements only happen at airports. That means 97 percent of the country, you have no measurements. There's no measurements. Satellite can't do it. There's nothing. Same with winds. Winds only get measured at airports when aircraft land, or with two balloon launches a day. We have a vast, desert of data below 5, 000 feet. The satellites get down to about 5, 000 and then they attenuate. There's some that get through, but it's not really relevant. This micro weather problem is the problem. We have always known it is a problem. There was a study done from the ground up. This was developed in 2008 by the National Academies that said we needed 400 profilers around the country to measure what we call the boundary layer up to 5, 000 feet on a regular basis, right, to get this sensor data so we can improve our modeling. So the helicopters may have to go to a hospital somewhere in the middle of an urban canyon that doesn't have even, it only has a windsock. There's no digital information and if there's fog or stratus they have to fly VFR. You just don't know if you can do it because we don't have data. So they'll drop that mission. They'll say, we can't take it and it could risk a life. And so they send an ambulance out. So that's an example of just one use case, but that magnifies itself across the board, right? Into both public safety, emergency response, and business, because we haven't talked about the business side yet of not flying that helicopter to the helicopter company or the hospital. That's just an example.

Jim:

What other industry is more dependent on good weather than aviation?

Don:

That's a really good question. I would say it depends on how you define good weather, right? If you're a farmer, you want rain. That's good weather, right? If you're a renewable energy farm, you want winds. That's good weather, right? But it may not be good for your picnic. So, I always say it's the use case that drives that question. Now, you're talking about a really nice day where that's probably tourism, Because, you know, people want to go to the beach.

Jim:

How about the ability to predict weather accurately, whether it be good or bad? What industry is more dependent on that, or is more affected by it? You just said there's a 40 billion study on cost and delays and cancellations. That's a significant cost to the industry. What other industry is more affected?

Don:

Well, I'm gonna blow your mind now. In 2017, the cost of weather on the US economy was 634 billion a year. And that covers agriculture, renewable energy, construction, logistics, transportation. These are all industries that are affected by weather, and they all need micro weather, by the way. But the challenge has been is we haven't been able to build the business case to close these gaps, to collect that micro weather, because these industries, there's no real coalition. The other problem is the government can't afford to put that out, that infrastructure. I was a science and technology director at the weather service for four years. I had a 130 million budget, and I learned pretty quickly that the money that we're spending on infrastructure in the government has to be on strategic infrastructure, such as satellites and things of that sort. And we have gotten much better at long range forecasting because of satellites. But when it gets to the microweather, we haven't really invested in that. And so that's like the nearest target, but yet the hardest to solve because of the resources. So if you think about it, agriculture can use microweather, renewable energy, right? They want if, you know, every time winds are off by one or two knots from the forecast, the renewable energy farm could be hit with millions and millions of dollars because they think they're going to generate a certain amount of power. They've got a contract to generate that power, and they may have to go to the spot market now to buy fossil fuels to make up for the difference they're losing. And that costs them lots of money.

Peter:

So Don, this deficiency microweather out weather that you talk about is it really a statement that we need more weather observations, in order to have a better picture of micro weather or do we have enough input data and the answer is somewhere else in this chain?

Don:

If you talk to folks that are trying to demonstrate that they are developing a micro weather forecast because it is a requirement, and people do need it. They're using techniques that downscale models, but not adding new data below 5, 000 feet. Just so you understand, the global models are like 9km resolution right now. That's the best. That's European. It means that every grid space is 9x9km, which means everything in that grid space looks the same. You're not resolving anything in that other than the average. Our short range models in the United States are three kilometer resolution, which is probably like a mile and three quarters or two miles. Those are in those grids you're not resolving anything below that. So if you try to continue to drive that resolution higher by getting down to one kilometer by one kilometer or 500 meters, you're not adding new data into the set. you're just, all you're doing is you're making it look better. You're making the pictures look really better, but it doesn't mean the data is accurate. You need real data just like anything else. You have to have data in order to do improvements to forecasts. There's also modeling. The modeling has to get better. There's physics and then the computing. So the three things that make a better forecast are data, computing, because you need speed. If you want to, if you do have the data and you want to run a one or half kilometer resolution model, it's going to take you computing resources, and that costs money and time, so you always have to worry, too, about turning these out fast enough. What good is a forecast if it comes out after the event. And the third one is the science. I mean, we're not, there's a reason why we can't forecast tornadoes perfectly, and there's still a lot of science. There's still a lot of things to learn. Even in icing, we have a lot of mysteries. So it's those three, but data is the key to improvement. You can't do machine learning without data. What I'm seeing in the industry is people are doing machine learning, but they're using analysis data. So the way we do historical weather in the weather community we don't have data everywhere, so you have to build a grid. You have to take the data you have, and then you build a grid, and you match that up against your physical models, and then you estimate what you think is happening in between those, and then you get a grid of historical data. But guess what? That's not real data. And people are using that kind of historical grid today, which is 15 kilometers or more resolution to run machine learning. And after they run the machine learning, they can't even validate it's right because there's no data. So, you know, that's the dirty secret in weather. Now, there are machine learning techniques that make sense, that you can do, like if you have weather radar, that you at least have real data, right? You're getting data. Although, keep in mind, weather radar is not necessarily always what it seems. Because you can have, areas that the radar beam goes above the precip, and it may be precipitating, and you're not going to see it on the radar. Right? So these are the, this is all the problems with weather instrumentation. There's no perfect solution with instruments. You have to build a village of sensors. Satellite, radar, lidar, remote sensing, you know, that kind of thing.

Peter:

So two questions. One, is is there a well executed accuracy assessment in the weather industry to track these various forecasts against what actually happened, so that we have an objective understanding of, okay, you have these large grids, we don't have differentiation inside, people are doing interpolation, they're training models off of those interpolations rather than off of actual observed data. But who's measuring how accurate that actually is and do we even have agreement on that, within the sector, within the industry, whatever you want to call it?

Don:

So, the, global models are being actively validated by the MET services. So ECMWF, National Weather Service, they are validating their models. The usefulness of that depends on what you're trying to measure. So, if you're trying to understand how accurate your winds are at 20, 000 feet, you're going to use a combination of satellite observation, and, balloon launches and maybe aircraft observations. And you, again, you're gonna map that out and potentially put it in a grid again. And you may not have a perfect real historical data set, but you have what's the best you can have. That would be on the average how you would, validate a model, using an analysis of the best data you can have and fitting it to a grid. Now, if you're talking about an airport, and you're just interested in validating your forecast at an airport for ceiling height, that's the best example of a really good validation, right? You can say, okay, I have measurements of ceiling. They come in every hour or more often. I can take a grid point out of my model and I can say, what's the ceiling forecast, what's the visibility forecast, the temperature for at that location. And that would be, that's your best set of data to do validation. So the problem is in between. So you have a point, and then you have an area. When you start looking at area, then it becomes more difficult to put a lot of confidence in all your validation. And so what TruWeather's doing, so, you know, just to give you an example, the low altitude has no validation except at a ground level at the airport. And that to me doesn't matter because drones don't fly at airports. And eVTOLs and air taxis mostly won't be flying at airports. They'll be flying out of urban centers that we have no data today. We don't know what the winds are doing in the urban canyons. We have no validation of that. All we could do is run models of what we think is happening. And they're not very good because there are a lot of estimates going into them. So, what we're doing is saying, look, we can't solve world hunger, but, when I saw the drone and air taxi industry coming up in 2017, I said, this is the best business model we've ever had for microweather. Now, you might say, well, what about agriculture, what about, yeah, they're all important. But when I say business model, I meant the ubiquitous nature, the amount of traffic that's going to be flying in certain areas. And the ability to amortize that cost of better services across a much larger array of companies and use cases. And so, my thesis has been and still is that we fill these low altitude weather gaps with ground based infrastructure and with drones that can collect weather data. And we use the satellites and we use the radar and we bring that together around corridors that are going to be your highest use corridors. Let's focus on that. We get real data now. When we run our models, we can validate that data, and we can then do real machine learning. That's the thesis.

Jim:

You've painted a picture of the weather gaps that I was not aware of. I did not realize it was all generally around the airports. And, but I want to go back to, before we get to the microweather, talk about the use case, talk about the gap that you're filling. I just want to pull it right back just a little bit to commercial aviation. You mentioned the 40 billion dollars in delays, diversions, and cancellations due to weather, you're saying, and you're saying a large portion of that is preventable. So let's eat into that a little bit. You were part of the NextGen weather ConOps. What could we do today to put a dent? If you're saying a large percent of that is preventable. What could we do today to be able to help address those preventable weather delays? How could better weather, better predictions, more accurate predictions help? And how do we do it?

Don:

Yeah. Well, it's a very good question. I was involved in the JPDO back in 2005 for NextGen. I helped write the weather ConOps I had my, you know, we worked with a couple of folks. We had a vision that, you know, we would have to run more modeling, collect more data, improve the microweather. Cause keep in mind, even around, even at an airport, you're going to have microweather challenges, right? If you've ever been to San Francisco, you know about the fog that sometimes really messes up Air New Zealand coming in. And the challenge is that, you've got to improve the micro weather for the current aviation to help with that understanding of how airport delays are going to impact the system. Now, when Boeing said, because it wasn't me, 28 billion was recoverable, I mean, obviously that's a perfect system, right? We built this, NextGen weather. We couldn't get DOT and NOAA to agree who was going to pay for it. That's how simple this problem was, right? You try to go back to OMB, and I have all this experience in government, so, you go, we went to OMB, we said, look, we need to make this a, interagency kind of thing. I'm going to be frank, DOC, Department of Commerce, they've got their budget. And, NOAA has their budget. And Weather Service has their budget. And, in order to improve and get to what is really required, requires big investments in computing, big investments in other things. We couldn't even get them to agree who's going to pay for it. So that's the first challenge is just the bureaucracy. Then you have the challenge of the current systems that operators are using, we're using a lot of legacy systems in aviation that, NextGen ran into this problem, right? They said, we, you want us to equip our aircraft and do what and pay for what, and then we have to change our software. So the problem is, I made a decision that problem is intractable, right? Let's go after the low hanging fruit. Let's go after a new industry. I think the only way you're going to get to what you asked me, is by demonstrating it. And showing people how bad it really, I mean, how much it's really, costing you and then demonstrating there is a solution, right? And so then I started working with the ASTM. We rewrote the FAA to their credit, said, yes, we got to change the way weather data is collected. We've got to open up the spigot. We can't just collect data at airports anymore and say, that's the only approved data. We rewrote the standard. We got it through. It's published. FAA is working with us right now on how we can test and demonstrate that. Which means then we'll be able to get more data collected through IoT and other things as long as they follow certain rules, right? So that was always my thesis. I gave up on the big picture, right? I said, let's focus on what we can control. Helicopters have always been the bastard childs of weather. So I always knew they would be a fan. And this is where it's going. So that's kind of my answer. I know it's not a great answer, but I'm telling you exactly why I'm doing what I'm doing.

Jim:

Who's most benefited from low level weather and who really is not being affected at all. There's a part of me that would say commercial aviation is probably the ones who are least affected by better low level weather and probably the drone community is probably most affected. Is that the case?

Don:

Yeah, I mean, it really has to do with the size of your aircraft, right? The engineering, has to do with, keep this in mind also, people forget this, what's the best weather sensor we have today on an aircraft? Does anybody want to take a shot at that?

Luka:

Pilot?

Don:

The pilot, exactly.

Luka:

We had a very simple rule when I was flying in the military. We would look out the window and there was a smokestack. If you could see the smokestack, you were good to go. If you could just barely, you know, pick it up, then you knew you were getting close to the limits. And if you couldn't see it, then the birds are walking. Even the birds are not flying. Right. So you're grounded for the day. And,

Don:

yeah.

Luka:

But you know, I get it. I get it. The thing is, I think where Jim's question was kind of pointing towards is, so what. what

Peter:

I mean, look, for a lot of drone ops, you still have a pilot standing on the ground watching the drone, right? Not BVLOS,

Don:

Want to make

Peter:

but For a

Don:

If You don't want to scale your bis..., If you don't want to scale your business, yeah, you can have a pilot on watching the drone, but that's not the, it's not, that's not what's going to work here, right? What I was trying to get at is that you can take a lot more weather on an aircraft that's engineered, it's heavy, it's big, it can carry regular fuel, not battery power. You can put more gas on it. You have a pilot who can deal with, unexpected circumstances that will come up because we don't forecast perfectly. And you have a pilot who has flown routes that they know the local weather effects better than the weather people because you don't have the data and they never come back and tell us what they experience because they want to have that as a secret so they can get us later when they say, Oh, I told you it would be that way. I'm kidding. But, the bottom line is that you cannot scale this industry if you don't have better digital information, because that pilot today might be sitting a thousand miles away from where they're flying their drones. They want to fly one too many. There's no way the pilots can even know what's going on. They don't even wake up in the same environment that they're flying in. All those heuristics are gone. I'm sorry to say, Mr. Pilot, it's useless. Because now you're flying somewhere where you've never flown before, you've never experienced the weather, you don't even wake up in the same location that you're flying you're weather blind. And if we don't backfill that information with real data, you're going to have more uncertainty and you're going to be cancelling missions more than you have to. And you cancel them today when you could fly. 30 percent of the time when you cancel a weather, a mission due to weather, you could fly. I know you know that because you always get mad at us when we say you can't fly and then you go, I could have flown today. That's money. And that happens in manned aviation. Imagine uncrewed. Imagine lighter aircraft. Imagine a consumed data source. I mean a conserved data, a power source. You can't add more power on a battery. So now you have to make decisions. How far am I going to go? How much weight can I carry? How much recharge time do I have? And if you're flying people, how bumpy is it going to be? Now this is a totally different environment. So that gets back to your original question. What is it that many few people agree with you, Don? They disagree with me about what I'm saying because they can't see the alternative reality that's going to occur in five years. Every knot of wind is going to be battery power, just like the renewable energy. It means weight, it means distance, it means something. And, that's the bottom line. I mean, I don't see it any other way because I'm so deep in it and I know it, you know, I could see the future on this.

Luka:

The tendency to cancel due to weather, being, too conservative today. I totally agree with you. I was frustrated with that when I was flying. But at the same time, the cost of traditional manned aviation missions, especially in military is much higher, right? And there's human lives that you're putting at risk. And so there's a tendency to err on the side of safety. But when you're dealing with drones, whose flights are really short to begin with, who don't necessarily have a high cost of returning home, if the weather turns out to be not good enough. That calculus is a little bit different for that segment of aviation. I guess what I'm trying to unpack a little bit more is who sees the business value in this? There's no doubt that there is a gap, but how do you monetize this? Who is it that will take that data, that new weather, granular micro low level weather and have a disproportionately higher impact on their business?

Don:

There's a lot to unpack there, right? So, we could talk about the military. Military is about bombs on target, reaching objectives. It's a different calculus, right? It's also about asset efficiency, asset management, because if you have limited resources and low density, high demand assets, if they're getting canceled due to wet weather, unnecessarily, you're not covering an area over the battle space that needs to be covered. So I can make the argument from a military perspective, which I'm very good at because I was in it about how, you need to be considering efficiency and using your assets in the military environment because they're not unlimited, right? And I proved that at the TACC. So, in the year 2000, I got to the Tanker Airlift Control Center at Scott Air Force Base. I was asked to build a regional weather squadron. I was the first and we built it to 108 people. And then I was at the Air Operations Center. We also supported TACC. And the first thing I noticed is that I was watching how planners and dispatchers were making decisions about how to allocate those airlift assets in those refueling assets around the world. And I couldn't believe what I was seeing. Weather wasn't even in the calculus. They would just plan and say we don't change based on a weather forecast. You've heard that before. I know that because you're in the military, right? So I watched this for a while and I said to my boss, the two star General Woolley, I said, there's low hanging fruit here. You know, we're not perfect at forecasting, but I could tell you. We are certain about certain forecasts, and if we're certain enough about those, why don't we take that into consideration when we're planning these missions? It doesn't mean we change everything. Let's just take a very focused risk management approach to this, where if we're really confident you change that mission profile, you're going to save resources. So he, he was game and we did that. And in one year we cut the delays. They were averaging 5, 000 delays a year from weather out of 100, 000 missions. So 5% In one year we cut it to 1800. So AMC says, what does that mean for us? And I said, well, let's take it to your financial analysts. How many more aircraft do you need to make that up? And what we found was that we saved them 200 million with that. So that's what's falling on the floor today. So we got that study. I'll share it with you. It was validated and we were nominated for the Chief of Staff Team Excellence Award and AMC made us validate everything because they didn't want to be embarrassed, right? So fast forward, that's what I'm using right now for this whole industry. But that was a military example. Now let me give you a commercial example. So I was up at eVTOL Insights two years ago and I had been asked to make a presentation. I was going to be one of the later in the day presentations and there was an OEM, one that everybody knows and is well familiar with that got up in the morning and said, I just happened to be in New York City where I grew up. So you're really in my backyard and I understand weather there. And they said we're going to fly air taxies and we're going to fly eight missions a day with one eVTOL. I said, okay. And, he said, we're going to generate 14 million a year. And I sat there and said, no, you're not, you haven't included the 30 percent weather tax. And it was at that moment that I trademarked the term weather tax because it hit me. People have to understand who are the people that care most about money in a company, the CFO, right? If I use the term tax, that's going to get their attention. So I got up in the end of the day and I said, look, You got a 30 percent weather tax in New York City, and people look at me like, what are you talking about? And I said, that's the price Mother Nature imposes upon your business, both real or perceived, because part of that tax is the time that you ground yourself when you could fly, based on uncertainty. And I said, so you're really going to generate 9. 8 million per aircraft. That's what's going to happen. And, of course,

Jim:

Why 30%?

Don:

Oh, well, I took my lessons from the TACC, and I took, other data, that we've collected to show that, about, on the average, depending on the city, you know, it all changes. I also said all the weather tax is all local, like politics, right? It changes based on the regime and the weather. But, generally speaking, when I look at all the studies that have been out about AAM, they just look at actual real weather impacts. They say, oh, the weather will be 15 or 20 percent, 50 percent of the time you won't be able to fly, right? But I know that you have to add that uncertainty piece in. So I take that as an average. I feel pretty good about it, right? 30 percent of the time, you're gonna be, impacted if you don't build a better system. So I said to him, look, I said, 30 percent of that's recoverable. I'm pretty confident of that. Probably more, but let's just give you a low hanging fruit. 30 percent is recoverable. So I said, first of all, you're really going to only generate 9. 8. Which they didn't like. And I said, but I could get you back 30 percent of that. And I said, for each aircraft, if I have the right systems in place in an urban area, I could get you back 30%. Think about what that is. That's like 1. 2 million a year, right? Then I looked at him, I said, how much will you pay for that? I said, don't you tell me that, I'm going to talk to your CFO. I said, if I could show you that, how much would you pay? I maintain they'll pay at least five to one, four to one return on investment. We could generate two to three hundred thousand per aircraft for weather alone. We're working right now with NASA, We're implementing, Urban Weather Sensing testbed at Hampton, we're bringing it out to Dallas, it's going to be around Alliance Airport, and we're going to do more of the economic studies behind this. But my argument is that you will pay me for that, if I could show you that I can get you more flight time, and by the way, more reliability for your customer, and if we have the better data, we can keep you out of rough air, we don't have to get people sick. Who are not going to come back on these small vehicles if they're getting sick. All these things. So, it's really hard to believe, right? I mean that this is possible, but that's kind of where my whole career has taken me to this point. It's all been a lifetime of work to get here and then to see the possibilities and now we're acting upon it and hopefully, you know, we'll be successful.

Luka:

For me, it's not that difficult to understand the value prop as you're describing to me, what I'm curious to learn more about is how do you convince those customers that this actually is achievable and is real, right? Because it kind of goes back to how do you convince people in one weather app being more accurate than another. There's no easy way to validate it. You need to run, large scale, experiments and testing and retroactive analysis. What is the second and third and the fourth conversation and the conversation with the CFO organization look like, can you give us a sense of their skepticism? You're having to convince them how this discussion evolves.

Don:

No, I mean, that's the big problem, right? So that's why I didn't go into conventional aviation because that would be an impossible thing to prove. When I saw the new industry, I said, this is all new software. It's all open data systems. We have ability now to actually make a new market around weather that is working in concert with the industry in understanding how they're using data in all areas of the industry, not just weather. How the FAA is moving to a performance risk based standard for weather, but for everything in aviation. So this is really being embedded and baked in right now into the industry as we go along. NASA's given us several SBIRs and we're going to be going out to Alliance Airport around Hillwood and, and we're going to be putting in three scanning wing LIDARs. We're going to put in the infrastructure that we think is required to get the data we need to run the high resolution modeling and to then be able to run those models every 15 minutes with winds that are unimaginably accurate compared to what you have today. And we're gonna, we have an economist, and we're gonna work with the operators, and that's the next step in the process. I realize you have to prove it. The thing I'm saying is I have no doubt that once we collect that data, if somebody wants to deny it, then it's not going to be rational. It's going to be because of their own personal culture, what they've been raised in, and how insistent they are that the way we used to do it is the right way to do it. I think we'll be able to show that. I just have to sit down with the number counters and show them the results. We have to have a clean slate, we got to throw away the past, we got to look at the future, and then we got to realize the software that's being built in this new system is very easy to work with. So what I did with the ASTM, as I said, all weather data has to be translated into JSON, GeoJSON, not weather geekism. I mean, one of the problem with weather data is it comes in so many shapes and sizes that people can't even get to it to use it properly. So I'm taking a holistic approach and saying this industry is the test for what I'm saying and NASA's been behind us and the FAA's behind us. I'm not going to tell you yet that it's assured. But what I will tell you is that I know you can't do better without better data. There's no data in the urban canyons. You can't run CFD modeling without real data to know what's really happening.

Peter:

So in the air taxi use case, like you go back to the New York scenario and you postulate that with these micro weather forecasts, predictions into the future, what the weather will be at that very granular locale, that these operators will be able to, fly more often, but how does that fit into their ability?

Don:

Or carry more weight, or carry more

Peter:

weight. Okay. But how does that fit into they're improving their profitability on their bottom line because how many hours or days ahead are they typically selling a ticket for that scheduled flight wherever the passenger is going and are you going to be forecasting, so far ahead of time that they can, cancel the flight before the ticket is sold? Or are they selling a ticket and then canceling a flight before departure? How does that time window work? And the contrast in commercial aviation is really quite different because. Well, number one, commercial aviation is so weather resilient that they only cancel when the weather gets to a real extreme. People are buying those tickets, days, weeks, even months ahead of time. But, you know, in, in the overall spectrum of weather, they're, they'll fly in instrument conditions, they'll fly in visual conditions, they'll fly in a lot of conditions and the transportation mode is pretty resilient to the weather, until you get to an extreme. But in these air taxis, from what I've gathered so far is a lot of these are going to be VFR operations. And so they themselves are just inherently more sensitive to weather conditions. If you realize this ability to predict whether it's not a weather observation, but it's the ahead of time factor that it brings, how does that match up with the eVTOL passenger when they plan to buy that ticket? And how does it mesh with, the operators, need to cancel and their ultimate profitability?

Don:

Very good question. I'm going to give you a certain amount of information, then we're going to have to sign an NDA after that, okay? Because some of this is my secret

Peter:

sauce. We'll sign up the whole audience of the show.

Don:

Well, they start paying and then we'll sign them up. But here's the thing is that, remember I said, you know, I'm not trying to solve the long range forecast problem. That's beyond my pay grade, right? As you mentioned earlier, you have to run large simulations. We're running a very localized microweather model that could be a 10 by 10 kilometer area or even less. We're going to use the initial conditions that come in from the bigger models. Then we're going to have our data coming in our sensors that nobody else has. And then we're going to nudge that and rerun that model with our data. And then we're going to run a short range forecast, probably out to six hours. Now, you might say to me, well, Don, is six hours enough? Well, the answer is, we're not trying to solve the whole problem. We're trying to go after the most impactful part of the problem. People are going to buy tickets when they buy them. The question is, and I always tell this to people in the military, because remember, I went to National War College, I have a degree in National Military Strategy, and I always brought weather in as part of the campaign, right? I believe weather is a weapon, weather, anticipate and exploit, you know, using it against the enemy. And I also ran an Air Force base. So I understand the logistics. I ran the main staging base in Afghanistan. And so I, I practice a lot of what I'm talking about today and I tested it in real life, but what I'm saying is that last six hours is the key, right? Because now you still have options at six hours out, or even three hours out. The key is not committing somebody to come to a VertiPort where they cannot take ground transportation to get to the airport. That's the key. Period. And that's what we're focusing on. And we're also building an intermodal system. We just won a DOT Smart Grant with Fort Worth, City of Fort Worth. We're going to bring weather into this, as well as air. And we're going to start looking at how, if you can't get to the VertiPort, or if the VertiPort cancels two hours ahead of time, can we get you in an Uber Black, what those road conditions are going to look like. All this predictability, we are just a chaotic transportation service today, right? This is about just bringing more discipline and better data into the system into these decisions.

Jim:

Don, help me out. So, one way to answer Peter's question, I believe. Again, you got me at today, most of your sensors are at the airports. So what you're saying is, I would assume with drone operations, drone delivery operations, eVTOL, those things that fly low altitude it's going to be, really important to have good weather. Peter's asking, so what? Peter's saying, well, what difference will it really make? So it's possible, even though they plan that advance. Very good point Peter on the commercial side. People already bought their tickets and they're already at the airport. Now it is true. They may not cancel because of weather, because they can fly in almost anything, but there'll be massive, unnecessary delays because isn't good weather prediction. And so I think that's a very big part of the 28 to 30 billion dollars. If I only knew really what was going to happen. So in your case, it's possible what you're saying is, because of the big chunks of grid today that weather is done on, if I know much more accurately where the real weather would be, because of your sensors, for example they're less apt to cancel flights that they otherwise would have cancelled because there is no weather in that spot. And I'm guessing that's where your three or four million dollars, a year per operation would come from. Is that right? Is that kind of the so what, the way I'm reading

Don:

it? Yeah. You got it. So what is, today you would can't, you know, you might have, I've seen this already. I've worked with some companies where they would have to cancel going back to Peter's point. I was driving away from the long range into the short range. So what was happening is they were doing testing, of some of these concepts. And they would have to cancel the mission two or three hours in advance because they didn't want to bring the people out to the VertiPort only to be stuck there. Or in the opposite state, you're at the airport and they're expecting an air taxi, and now you're not going to have it. So they're going to have to find another means of transportation. That's not as bad as the other one, right? Where you can't get to the VertiPort and now you've waited. You're only an hour and a half before takeoff because you thought you were going to fly there, and now you can't get to the airport now because traffic in those areas won't allow it. So, they were making this no go decision about two, two and a half hours before. Well, a lot of times they could have flown, and this is something they told me. So that's where the 30 percent comes in, right? So now we bring them out 30 percent of the time more than we did, and they're going to fly and they're going to make money. And they're not going to be sitting there with the empty aircraft sitting on the VertiPort, not flying and all the costs that goes with that. Does that help a little bit? I'm glad you helped me tease that out.

Jim:

It does. Help me out a little bit with, drone delivery services. Our sense is that eVTOLs are still several years away at scale, potentially more than that. some would say, but drone service delivery is happening today, and it's going to happen at scale a whole lot faster, arguably, than eVTOL operations. How does micro weather help, now you got me talking only micro weather, so you've trained me. How does micro weather help make a difference to the drone service delivery Does it really, will they make any different decisions based on better weather.

Peter:

And I would say just on, on Jim's question, when you're thinking about drone delivery, and let's just take last mile drone delivery, because so many people have a pretty vivid picture of what it is, that's going to operate more based on microweather observations than forecasts. And, you know, consumers will open the app and they will either be able to place that order or not, because this is all taking place within a matter of minutes. It's, you know, it's not like I'm ordering something three, four, five, six hours ahead of time and making some kind of a decision based on that. It's immediate. So it's just, okay, what is the weather along this two, three mile route right now or 15 minutes in the future? And can we go?

Don:

One of the things I want to point out is micro weather doesn't become as important if you're not flying scheduled missions. If you're not depending on velocity, where your business model depends on having a lot of flights, right? If you're doing surveillance on a pipeline And you could wait another day. It doesn't become as much of an issue. So, the drone delivery one is one of those that really depends on volume. And so every mission you don't fly you're dropping revenue that probably hasn't been told to your investor about how much the weather challenge is. So there's a lot of pressure on these companies. You got to remember, it's not like the current airlines who have baked in weather already, right? These are all folks that are going after VC money or other money. And they're, you know, they're embellishing a little bit of their abilities to fly, 95 percent of the time when it's really maybe around 20, right? So now they're already behind the power curve. The challenges with drone delivery is the winds above the tree level is not measurable and we see a lot of difference in the winds above tree level than the models put in there. The battery power issue, the weight carried issue. It just adds chaos into the system as well as the problem of you have to have three miles visibility you have to you know, be 500 feet below a cloud deck, all those things. Right. So you notice, I even talk about a thunderstorm because the thunderstorm stuff we see coming. It's the micro weather that's invisible or not notable that's going to hurt their business model. So, So, you're right. Now, one thing I will say, Peter, is that you really need about an hour, ahead of time because, you gotta remember there is an operation, there are people loading drones, they're running around the store picking stuff up, so, you know, if somebody puts in an order, it's not like it's gonna happen in 10 minutes, generally, right? So you have to think about the flow. So, if you have, I don't know, just to say you have 15 drones at a Walmart someday, right? That's 15 drones. That's a lot of drones? So every 10 minutes you're not flying when you can fly, you've just dropped 15 drone flights. So even the short range, and you do need the data, as you said. So what we've been advocating for is let's get some wind sensors on towers at 200 feet, 100 feet, 150 feet. Let's get above the tree line. Once we know what the winds are really doing, which we don't have today, that takes away some of the mystery immediately about power management, distance traveled, wind shear, there's other things I could talk about. And then let's get a ceilometer at every Walmart or wherever you're flying from so we can see the cloud height, measure it, because right now they're using the nearest METAR. And the nearest METAR could be 15 miles away and it may not even be relevant. You've been on airports where one half of the runway is IFR and the other half is clear, right? I used to, that happened to me at Kunsan. I used to get yelled at all the time by the wing commander because we would call it seven miles visibility but the approach end of the runway, which was less than a quarter of the area drone, was socked in but I couldn't call that visibility. He threatened to put me on a boat at the end of the runway to take my weather observation. So, my point is that all that comes together. So you're right. The data is important. Having that extra information is important. And then you still need a short range forecast for your flow to manage your logistics, right? To make sure you're not taking orders that you know you're not going to be to

Luka:

meet. Don, you mentioned earlier that about 40 percent of drone flights are canceled due to weather. Can you talk more about that? And, you know, maybe be a little bit more specific who do you refer to when you mentioned this, and if you talk to drone delivery organizations that are really starting to have some meaningful and interesting scale, companies like Wing, like Manna. Do they have this problem? Do they, complete your sentences when you talk about the issues of winds aloft and how they're not able to predict those and how this impacts their customer service ultimately. Do you have an agreement on the impact that this is having on their operation and how do they think about this? Do you work with them? Do you provide services to them?

Don:

Very good question. So what I will tell you is I, we are a Drone Ups company. I'm proud of the fact that they hired us, over another company after they realized that what we're producing is meaningful to them. They've acknowledged the challenges of micro weather, keep in mind that, when I say 40%, it doesn't mean the weather was bad 40 percent of the time. It means that you just don't know. So, that statistic really depends on where you're, what your weather challenges are, right? Some people might think, oh, in Phoenix, you probably don't have a 40 percent attrit. Well, I don't know the winds, right? The winds above the ground, you know, they can't handle 30 knots of wind necessarily above the ground, right? So your weather attrit is not seemingly the same as what you would think about for a regular aviation, right? That number seems high if you're not really on the ground and understanding what's happening. What we tell them, you know, is that we'll get you that, some of that back. And it means that you have to buy into the whole system. You've got to buy into the system of better sensors above tree level, some ceilometers, the way that we, aggregate data and present it through custom, outputs for locale and your decisions. You've got to make very systematic approaches to how decisions are made. You can't leave it up to a local pilot, to make that decision because they don't have the big picture. So this is the kind of work we're doing with DroneUp is trying to figure out how do we, really, you know, it's not just the data collection side, it's also the decision side, right? And how you build the system so that those decisions are made in a systematic fashion, that there's a stable process. I could talk about other drone companies. I don't want to do that because some drone companies don't want to admit there's a problem yet because they're still trying to, raise money. Others, are doing certain things that they feel they can handle it themselves, which may be possible. I just know I've been in the business long enough to know that you can't, if you're going to be a successful low altitude drone company or eVTOL company, you can't just look at weather as a data supplier. You have to have partnerships. There's even a meteorologist piece to this. There'll be times when you're going to have an area that has lots of drones that you want to fly and there's a system coming in and the models are going to say, you know, that system will be, you'll be getting rain at 30%. You know, one of the things I'm finding is what does a 40 percent chance of rain mean? Do you cancel? How do you make a decision with that, right? So, what we say is you talk to our meteorologist, we'll help, we'll get you an extra hour of flight time, potentially, because we say, oh, that 40 percent is happening north, it's not going to move down into your area for another hour. There's no automated weather today that could be the heuristical part of that, like you used to get when you talk to a forecaster. There's still going to be a requirement to have meteorologists in a loop until we're perfect at forecasting.

Luka:

Assuming a drone delivery operation is happening from a hub and covers, let's say a four mile radius, how much does it cost a drone operator to instrument this particular area with the necessary sensors, and then also get the meteorological service from you. What is that burden on an annual basis or on a per flight basis, and how does that fit into the overall economics of those delivery operators?

Don:

Yeah, so I don't want to give you all those details because that's my IP, but what I will tell you is this. Our goal is to get them four times return on investment. When we talk about sensors, when we talk about integration, when we talk about APIs, and when we talk about getting those APIs properly integrated into, their decision support systems that they're doing. And when we talk about even having a forecaster in the loop every now and then, you don't have them every day, you only have them on the edge cases, you call them up. The goal is a 4 to 1 return on investment. So I'll give you an example. If they're losing, 40 percent of their missions today, we want to get it down to 30%. And if that revenue, we want to get you, we want to get you, you know, we don't want you to pay more than 25 percent of what that savings is. And that doesn't even get into the cost avoidance piece. I'm just talking about revenue right now. We're not even talking about people that are standing around and I've seen this doing nothing because they're waiting for the rain to end, right?

Luka:

It seems like, the customers who could benefit the most from this, if I'm interpreting your, arguments correctly are the, high frequency, low altitude operators, drone delivery operations, for instance, right in the near term. Those are also, customers who are very pressured by, cost. And they are looking at every single cent and asking themselves the question, can I afford to spend this cent because my unit economics are really challenging to maintain and achieve. We see this often with, all kinds of supplemental data providers that yes, have a lot of value, but, you know, drone operators might only have the capacity to pay cents per flight for some of this data. So in that context, how do they prioritize the value of weather data and their willingness to, give up margin for it, as opposed to best effort, if I need to cancel a flight, I'll cancel a flight, I'll send it a minute later or a couple of minutes later, because it's really at a small scale.

Don:

Well, first of all, all those other methods cost money. So if you've canceled a mission because you didn't have good data, and you then had to drive that or find an Uber Eats guy to take it, you've just spent money. We're looking at this the wrong way. You're looking at weather as a cost of doing business. Weather is an enabler to make money. That's the problem I run into all the time. I'm not criticizing. I'm saying that what we're up against is a culture of how people think about weather. Everyone thinks about the cost, but nobody recognizes weather's costing them. You know, there's no such thing as free weather data. It's costing you. So, what I would tell you is this, and again, some of this has to be proven, but I'm, you know, there's an old saying, I don't know this to be a fact, but I know it to be true, right? Weather may cost you overall today, 60 cents a drone flight when you amortize all the impacts of weather on your ecosystem. If you amortize that cost, it might cost you 60 cents a drone flight. Alright, if you start paying me 15 cents a drone flight, I'm not, I'm using this as an example, and I save you 60 cents, I don't understand where the problem is. Why are we prioritizing? You're going to have to lay somebody off instead, or you're going to have to cut costs somewhere else, or, and this is what I say all the time in weather. It's amazing how much value there is here that people haven't realized yet, and they're actually cutting costs in areas that they don't have to if they just have a smarter weather system.

Peter:

Yeah. These drone companies that, that we've talked to, they have weather minimums where they are flying with about a 50 percent safety margin on wind conditions, rain, etc. So they have a pretty wide safety margin for their go no go. But obviously they want better dispatchability. They want to fly more often.

Don:

So we want to get that safety margin down. You know, we want to get more certainty so you can bring that down to 20 percent.

Peter:

I don't know if that's the direction that they're going to go though, because the drone platforms themselves are improving with weather resilience, so much. I mean, and a lot of it has its roots in the amount of energy that the drone can bring along with it on the mission. And so we know batteries are getting better. We know drones are being built that can handle stronger winds. They can handle more moisture in the air. Icing will always be a challenge, but short of that, I think we see the weather performance of these drone systems improving a lot in the coming years. So, how does that trend fit in with what you are offering to them on a forecast basis? And I would just assume that they're going to, keep the same 50 percent type safety margin.

Don:

Well, if you're assuming they're going to keep the same 50 percent safety margin, then there's no business model. Because that, they're, for some reason, they're doing that. But back to the, the question, the bottom line is this, is that we have very sophisticated aircraft today that are manned and they, and we still ground these aircraft 30 percent of the time unnecessarily. The same thing is going to happen when you have your very sophisticated drones. If you ever use the term all weather, be careful, because there's no such thing. I've been told we can have all weather aircraft since 1980s. There's no all weather aircraft. The more you engineer, the more weight you put on an aircraft, there's always a trade off between engineering, weight, and everything else. So I've been in this a long time. It really comes to this way. You have to improve your engineering, but you have to improve your weather resiliency from a prediction standpoint. You have to work them both together. There's no one or the other. You're always going to have weather impacts and you're always going to have margins that you're going to give up if you don't improve the data. I mean, why wouldn't you want to invest in better data and get more margin and get more flights. I mean, this doesn't make sense to me. I mean, maybe I'm, I don't know. I don't get it. Right. I don't get why you have to have a 50 percent safety if you're got better weather data.

Jim:

Are there other alternative funding sources that are available for your capability or your competitors? Beyond just the organizations themselves paying for your services, will the government underwrite your capabilities? If they're looking at every cent, which they are today, and you're saying, but there's real value here, and they're just prioritizing their dollar, is there an opportunity for the government, any kind of consortiums to help underwrite the costs of the microweather?

Don:

Yes. To answer your question, I don't expect the operators to pay, the full cost of this. There's two things I want to mention. Number one is once we get these low altitude weather networks out, they're going to be able to serve other verticals that today are underserved. So if you think about the urban canyons, there's a lot of air pollution. There's a lot of concern about how air pollution moves around transport through urban canyons. This data is not only going to be valuable to aviation, it's going to become valuable to many other verticals and they will buy it at some point. This just hasn't been a business model. The second thing I want to mention is that the states, can invest in infrastructure. So the way I break it down is this way. You have strategic weather infrastructure and then you have tactical weather infrastructure. The strategic weather infrastructure would be like those scanning wind lidars. You don't expect companies to pay, anybody to pay for that, right? Either that's going to be private investment, helping to fund that, that would then, you know, buy into that based on the business model, or we look at public private partnerships like toll roads. Right? So we could see that the states, there's like five business models where the states can actually buy some infrastructure, but then just contract out the maintenance, the data processing, and then, you know, you work out the business model with the state. Maybe, they get a discount for all public safety to use the data, right? And then you charge commercial some fee. There's a bunch of different business models that can make this work, which is why, I'm not landing on one solution. I'm still figuring out all the things that you guys have brought up, I'm well aware of. And so I've been working on the approaches to try to get around those, objections, right? So to speak, to try to find a way to make this work, to make it profitable and make it work. So I don't see the, drone operators having to buy all the infrastructure. I see them buying a service. And that's really where we want to get to. Now, they may still want to buy some of their own equipment for takeoff and landing location that only they have, right, like a ceilometer. But, if we can get everybody to share their data, you've got real data now coming in where you can improve cloud ceiling forecasts over an area to get a much better understanding. Today they don't know what they're going to fly in if they can't see it. There's no data. You don't know if it's IFR around that ridge. You don't know. So that means they have to stay grounded until they know and things like that.

Luka:

Don, can you share some, thoughts, observations, what it's like being an entrepreneur and building a, a weather business thoughts for aspiring founders in this industry.

Don:

Well, you know, you guys got an idea of what it's like, because you've been asking me very good, challenging questions. And by the way, I do appreciate it, because, what we're really trying to do is very hard on a number of levels, right? You got the policy. Which I've worked on with FAA and ASTM to try to, you know, make it more favorable to bring in third party weather providers into the aviation industry. You've got the science and technology, which is all cutting edge. Everything we're doing is on the edge of technology and science. It hasn't been that this stuff hasn't existed before. It's been in the universities and labs a long time, but no one's ever pulled it together into a, you know, into a production system and make it into a business case that could work. And then you've got the culture in aviation, like, the questions you guys are asking. These are all good questions. They're all real problems. They're all real objections that I have to overcome. and that's hard. That takes a lot of energy, right? If I could have probably invented the pogo stick and it would have been a hell of a lot easier, right? But I like hard challenges because I've been in this industry so long, I've been fed up that we haven't been able to really change the aviation system, right? It's been the same system for years and it just doesn't make sense to me that we're kind of happy with it when it's so disruptive and costs so much. So, and then selling my investors. Right? So, this is not an easy sell. You have to find people that really get in and understand that it's really about the data. It's not about weather, right? It's about the data. It's about what the data can do. Why is it every other part of the industry says we need better data to make decisions, but weather, eh, we don't need better data. This will work fine. I mean, I just, it's so baffling to me. And I understand it. I think it's the culture. I think it's the way people have been brought up. I think it's just the system. And it's also, really difficult to understand what we're trying to do and understand why it will work. So I get all that. That's, what's been the hardest part. I mean, this, building the business, hard because you need money, hard because we need customers. What I am confident about right now is we will get them. So what's the second problem? The second problem is the pace by which this industry is actually growing. And I would tell people that are, trying to, build a business in this industry, not just weather, is you gotta have patience and you gotta have recognize it's never going to go as fast as you think it is. Don't dilute yourself thinking, that you're gonna, you know, do something different than everyone else, because you can't if the policy doesn't change, you can't if the FAA doesn't change the policy. And the other thing is these, all these other companies I work with, they're all doing new software too, which is what I love, by the way. And they all have their challenges. Like you said, Peter, they may not be ready to invest in weather, right? They may not be ready to invest in something else. So, I think, the key is, you have to know what you're getting into. You got to understand that it's a different animal. We're basically building an aircraft and flying at the same time in this industry writ large. It's not an established industry where you're trying to do something that's entrepreneurial and innovative, and it's challenging.

Luka:

And as we wrap up, what's, one message that you would like to leave with the audience?

Don:

Well, I think, you know, again, I think this has been a great conversation. I'm really appreciate you guys, I felt like I was, on, trial here and it was good, right? It was good.

Luka:

I hope it wasn't like that. We like to think of it as more of a discussion.

Don:

Well, you know, I'm kind of built for the kind of this kind of discussion never going as easy as I want it to go right. I do believe that what we're doing is the right thing I'm always more than happy to have more conversations about this. I want to you know, there's things I still have to learn There's people that probably can help me figure out how to crack the nut Right? So I don't want you to think that I have all the answers. what I'm just going to say though is weather imposes uncertainty. And uncertainty in anything costs money. And, the only way that you're going to resolve that is by having real data to fill in that uncertainty gap. And if the data can give you the ROI and you can demonstrate it, believe it. If I can show you this, believe it, right? The implementation part is hard because the companies we're working with have to bring their company along with this, right? Who, again, have a lot of people who would say, Oh, why are we doing this data? I get it from ForeFlight. I can see my weather, you know. So, this is a multi faceted problem, right? That's why you have to come in at the CFO level and in the COO and CEO level. Because, you really got to have a buy in by the company up and down to say, yeah, this is the direction we're going to go in because it makes sense. It's the future of weather. It's the future of, networked information and it's the future of analytics. So you want good weather data in there. Why are you going to spend a billion dollars on all these other systems that are going to give you perfect data and then you're going to use weather as your weakest link in the tent? It doesn't make sense to me. Why invest in all that if it's, this is your weakest link in your chain, right? So that's my takeaway.

Luka:

Well, we certainly enjoyed it as well, Don. Thank you really for a great conversation and look forward to having you back and talking more about this. So really appreciate your time.

Don:

Thank you. Really, thank you.