The Vertical Space

#77 Charlton Evans, End State Solutions: Drone boxes, FAA waivers and exemptions, Part 108

Luka T Episode 77

In this episode, we dive into the FAA's evolving approach to approving waivers and exemptions for complex drone operations, including beyond visual line of sight (BVLOS), 1:many operations, and broadly airspace integration over the past 18 months. Joining us again is Charlton Evans, a trusted expert in FAA drone regulations, who last appeared on episode #35.

Charlton explains the significant shift in how drone OEMs and operators approach the type certification requirement in the U.S. Traditionally, the process involved going through the FAA's Durability & Reliability (D&R) process, which often took four to five years, as seen with companies like Matternet and Percepto. However, since mid-2023, the FAA has introduced a streamlined 44807 type certification exemption process through the Flight Standards Office, focusing on self-declaration. Charlton walks us through this new process, notable waivers, and his expectations for the upcoming Part 108 rulemaking.

We also explore the growing adoption of drone-in-a-box systems, particularly following DJI’s launch of the DJI Dock 2, and finish with a discussion on AI use cases in general aviation.

Charlton:

I think what's improved is we mentioned the exemption process. I think that is the pressure relief valve for type certification right now. Unless an applicant comes to us with a product that's way too heavy, way too complex of a concept of operations, if they're going to try to carry people, then they need a type certificate. But otherwise, the whole enablement through 44807, makes it almost prohibitive to attempt a type certification under at least 3, 000 pounds gross weight for any equipment or any product today, in my opinion.

Luka:

Over the last 18 months or so the FAA has become increasingly accommodating in approving waivers and exemptions for complex drone operations, including for beyond visual line of sight, 1:many operations and, airspace integration broadly. To unpack the details behind this we're talking to our good friend, Charlton Evans, who you might remember from episode number 35 over a year and a half ago. Some of you might not be aware, but there's been a significant change in how drone OEMs and operators to solve for the type certification requirement for certain use cases in the United States, complex use cases in the United States. Traditionally this involved going through the so-called D&R Durability and Reliability process through the Aircraft Certification Service, the FAA's division that certifies aircraft designs, manufacturing processes and airworthiness, in other words, a very design-centric process. But this also ended up being a four or five-year affair think Matternet and Percepto, for example. Much longer than anyone's expectations. However, since the summer of 2023, the FAA has created a more streamlined 44807 type certification exemption process, managed by the Flight Standards Office with an emphasis on self declaration. Charlton walks us through how this process works while also discussing other notable waivers and exemptions and sharing his expectations for the long awaited Part 108 rulemaking. We also discuss drone in a box concepts, which have gained increasing traction, especially after DJI introduced the DJI Dock 2 and we wrap up with a conversation about AI use cases for General Aviation. Charlton Evans is the founder and chief executive officer of End State Solutions, a firm focused on certification and airworthiness for Advanced Air Mobility, drone delivery, High Altitude Pseudo Satellites and Remotely Piloted Aircraft Systems. Charlton was a Harrier pilot and tactical air controller in the Marine Corps and still flies for business and pleasure as a commercial pilot. Charlton has led several successful civil and military drone certifications as well as historic BVLOS flight operations. Charlton and also led historic linear infrastructure and disaster response flights with the type certified Scan Eagle UAS that resulted in the induction of ScanEagle N202SE into the Smithsonian Institute in 2016. Today End State Solutions is considered a trusted advisor by both industry and FAA and engaged across the spectrum of type production, operational approvals, as well as regulatory affairs, building relationships that build trust and products that are trusted through certification. Charlton is a true expert in FAA drone regulations so reach out to him if you need a reliable guide to help you navigate the FAA and get started with complex drone operations. Enjoy the conversation with Charlton after a brief sponsor message. Charlton, welcome back to The Vertical Space. It's great to have you back on the show.

Charlton:

Thank you. Great to here.

Luka:

You know that the first thing we ask is anything that industry disagrees with you on. And last time, I think you mentioned something along the lines of the emerging technology crowd that disagrees with you around locking in the design and just trusting the certification rules and the system. How would you answer the question today?

Charlton:

So, I think there's still a learning curve there, but I would, dogpile on that with, a related element of getting it done. And that's that we can get this done more quickly than anyone thinks we can. So, I mean, it is aerospace. it's intentionally regulated and paced. Aerospace puts safety higher on the list of priorities than innovation or economic return. So all three of those elements, the innovation piece, the economic return, and the safety piece are all interlinked, independent, but their intention, right? Go fast, fail fast, doesn't work well, in an environment where the, results can be catastrophic.

Luka:

Why is it that people disagree with you or what specific aspect of it?

Charlton:

Well, it's the concept that, And I guess our experience is when we push out an estimate or a quote to a, a prospective client, we often, find that they don't understand why is this going to take months instead of weeks or years instead of months. And there are many elements that go into that, but one of them is the baseline that, When you engage with the FAA and, the regulators, their job is to gate this new emerging technology to the point where everyone's convinced that it's not going to have adverse outcomes, it's going to be safe for its intended purpose, and that creates a natural filter and friction against the innovation. Does that make sense?

Luka:

Isn't that a well known fact though? as evidenced by everything that's going on in the industry, why is this something that people wouldn't agree with you on?

Charlton:

I think because in the emerging tech world, investment is key to progress and the pressure from prospective investors creates stories that result in compressed timelines, aggressive timelines, unrealistic timelines in some cases.

Jim:

So, as we're going to ask a lot based on what you said, 18 months ago, what's matured? How has the industry changed in the last 18 months as it relates to this area that people may not agree with you on as much? Have they matured?

Charlton:

To some degree, I mean, pain is a powerful teacher, right? So if you're seeking investment and you're making promises about, timelines and or capabilities, economics, I think the pool of investors that can be snowed by some of the unrealistic claims has shrunk significantly. Right. So, you know, investment in aerospace it requires that one addresses a realistic approach to design evolution and approvals. And, if you don't, then, prepare to be disappointed. So I think that has definitely happened and, but the symptom of that is that investment has shrunk rather than adjusting expectation on timelines, there's just been a withdrawal to some degree. And I don't have great visibility on how much, but certainly we are seeing that there's less, powder coming out of the cake, if you will.

Luka:

What about more broadly, some of the changes over the last 18 months worth flagging?

Charlton:

I think there has been some maturity in the concept that a configuration is part of the successful formula and entering into service and then iterating on that configuration in a way that doesn't eclipse the progress you can make with a configuration in service doing its intended work in creating revenue. So I think that's, again, probably a lesson learned through some pain. So companies have, particularly on the, well, probably across the spectrum, but we see it most acutely maybe in the drone delivery segment, where companies are now using a device, a tool, a drone that's been approved that they can get work done with, that they can actually start to grow with prior to scaling, and then in the meantime, working configurations in the background, more advanced products in the background to, prop up a future operation that may be based on the previous, approval. So if you've got a 135 certificate today and you're using drone configuration ABC, meaning while you're working on CDE in the background to plunk up underneath your 135 in future operations rather than try to iterate on that in the current operation.

Luka:

Got it. Peter and you cross paths at the recent commercial UAV expo. Any notable announcements that came out of that event or interesting products or, intriguing rumors that you've come across?

Charlton:

Yeah, well the, I think the two things that really jumped out at me where there was an announcement about Percepto's type certificate being granted. That's a pretty big deal for them. I think it was muted from an industry standpoint, and it took way too long. So, we can, I think, talk about that more in depth, maybe, as the questions pan out here. But, It's unfortunate that it took so long because in the meantime, Percepto and many others have gotten exemptions that have the same effect of allowing them to get into service, into revenue service. So it diminishes the value of the TC to some degree, and, yeah, there's a miss that's been made there, by FAA, I think, that, deserves some more attention, and will get some more attention. The other thing that, that was prominent to me, and maybe it's because I was listening for it, but the FAA was pretty clear from the stage about the path they intend to take with 108 the new rulemaking. And, it doesn't take a whole lot of reading between the lines to understand that the operations going on under 44807 today, both the equipment and the equipment approvals or exemptions for airworthiness and the operations certifications are really being used to inform what's happening, in the rulemaking for 108. So when we see that rule come out, you can expect, pretty clear connections between today's drone delivery operations, for example, and, the Mosaic rule making and what's going to be in Part 108.

Peter:

Charlton, what did you think about the maturation of the drone box model and the things that were on display, at the show? I thought it was notable, DJI's DroneBox 2, the price point that it's coming out, the volume that we seem to be seeing, of that moving in the market in just the last few months.

Charlton:

I don't know much about the actual economics or where they are in price point. I do know that the demand for that type of product to do that kind of job for, local inspections, infrastructure work is really high. So oil and gas, all the enterprise level energy, rail, et cetera, are all looking for that solution because the benefits it offers, not just in the data that it collects and can collect, and they're starting to really peel that back, but also the remote nature, right? It can be automated. There are clear paths to, operating it even centrally from a, a command center, if you will, or a data center that collects that without having people on site. So it's leveraging the automation to really start to to scale.

Peter:

Right. It seemed to me like the maturity that, that they were talking about at that show of that it's one of the instances of, okay, this is a system that's at a price point where customers are now thinking seriously about deploying a lot of it, and they are comfortable enough with the reliability that it's exhibiting, that they're willing to, really engage with it in a way that, perhaps before they were, evaluating, the technology and just waiting.

Charlton:

Yeah, well, and the price point is important because in the enterprise industry, oil and gas, et cetera, their bar is really, what am I paying a person to do this for right now. Which when you look at a full time employee is not a lot, relatively speaking, right, compared to what it would cost them to install and utilize a drone. But I think previously there was not a tremendous amount of value placed on the quantity and quality of data that could be collected by these autonomous systems. So as that gets realized and there's value placed on that, then, step back and say, Okay, well, I understand that the upfront costs are maybe greater than, having somebody walk around and do this. But the economics start to pan out really well when it's automated and there's nobody there at all.

Jim:

Charlton, can you just give a little bit of a background on Drone in a Box? We haven't talked a lot about it on the podcast, so, what is it why is it different?

Charlton:

So the box replaces the operator on site. All right, so it's an automated way of launching and recovering, recharging battery, collecting data, the box ends up being, the nest, if you will, for the drone, and, there are multiple approaches, either, fully in house, bespoke units, or some folks are trying to manufacture the box that's agnostic to the drone and use, whatever the best drone or tool for that particular data collection job is. But at the end of the day, the goal is to not have the operator there physically on site for pre flight, launch, recovery, post flight assessment, data collection, all that gets automated.

Jim:

Kind of Matternet like quality to it, but portable?

Charlton:

Yeah, and in terms of, application, it's really more data collection than delivery at this point, or that's most of the use cases. So, Percepto is a good example of a drone in the box system. There are some that have come and gone, maybe were before their time, but, certainly the companies that have started doing it have really paved the way in a regulatory sense, so it's not mysterious on how to get an approval that includes waivers for parts of 107, like 107.15, I think it's 107. 49, that are the pre flight elements, right? So nobody has to be on site to actually inspect this thing and look around into the airspace and say, yep, safe to go fly. They can do that all remotely. So that's the function the box serves and, as well as all the technical, elements of it to recharge, etc. Change out batteries.

Peter:

Charlton, any thoughts on, from an end user, how they're thinking about using a drone in a box systems for doing inspection versus using an aircraft that has long range to do, say, long linear inspection. So let's say a customer has a pipeline and they need to inspect that pipeline. How are they going to evaluate the choice between putting a dozen drones in a box down the length of that pipeline or whatever the number is versus, using an unmanned aircraft that has the long range to go fly the entire length of it? What is that trade off looking like?

Charlton:

Yep, so the data collection piece is part of it, right? It's very, just very efficient to collect, that kind of data with a fixed wing platform that's constantly moving over a long linear piece of infrastructure. But the regulatory hurdle has been higher, and the technology required to get that done for BVLOS is more intense. So the regulatory approvals for the drone in the box are local in nature. they are going to leverage shielding or operating within the obstacle, if you will, as a mitigation. And that's a low tech, high operational mitigation. We used to call it a soft or a strategic mitigation, right? And then if you want to fly a Cub like or a Cessna like aircraft over a bunch of pipeline, then you're going to have to have technical solutions that are equivalent levels of safety to get you integrated into the national airspace to do that, which is a higher technical bar. And making the safety case for that same system is a higher regulatory bar. So the trade off is if you're willing to push on those levers to, to get those regulatory and technical pieces solved, then you can reap the benefits of a more efficient data collection over linear infrastructure.

Luka:

And when you speak to end users in some of these segments, energy, oil, and gas that are drone in box customers, what's important to them, obviously cost is. The expertise required to run these programs is another one, so, regulatory, gymnastics associated with it. But if we look at the state of the industry right now, a lot of the waivers recently, and we can speak to that in more detail, if you don't mind, for some of these drone boxes involve ADS B in, and then infrastructure shielding or masking. And so that seems to be a template that works. Now we have boxes that came down in price from a few hundred thousand dollars to, teens, twenties, thousands of dollars. So that now becomes a compelling business case. What else stands in the way of these customers embracing drone boxes in greater volumes?

Charlton:

I think it's important to define who the customer is, who the end user is, right? Is it the local facility manager, who's trying to solve problems that he's relying on people to do right now, or is it the more enterprise level data folks who need to meet regulatory inspection requirements, or have specific problems they're trying to solve with off nominal behavior, meaning oil and gas is an example, leakage, right, or methane detection, where it's not supposed to be. Right now it seems to be the more acute cases where we're looking for off nominal activity, meaning we've got a regulatory baseline of inspections we've got to make, but we also want to know where we've got leaks, as quickly as possible and why they're occurring and how to address them in a very, acute way, emergent way. So, I think, the end user being the data, the customer within the enterprise is probably the growing end user. The facilities manager who needs to know things locally about his facility is a beneficiary of that, but I think he's probably a secondary end user. So sensor development to solve those problems is finally, I think, gaining traction, The tool or the, our little pickup truck to drive around the sensors being the drone and the box to enable it. That's all coming along really quickly, right? Better, battery technology, more, advanced boxes for the drones to operate from, less human interaction. That's all happening pretty steadily and the regulatory approvals to do that are no longer mysterious.

Luka:

Where do you see the product stack being the weakest right now? Is it in the payloads and the data collection? Is it in the data flow between the drone to ultimately whatever software these end users are consuming the ERP systems and the like.

Charlton:

It's, seems like there's a lot of development going on in, methane detection, OGI, that, there doesn't seem to be a go to. That seems like an area where there's room for, leaders to make progress against that sensor problem. Then I think we're going to run into really the data management problems. I think that's going to be reactive. So, how to handle all that data, in an enterprise way is going to be driven by how much data is collected. And I know that's happening now to some degree, but I think it's going to be a bigger problem once they've really found a way to collect specifically all the data they'd like. And the data is really, the it's, It's addictive, I think. The first one's free, and then as the folks who start to collect it recognize the value of it, and not just how to meet regulatory requirements, but also how to prevent, cost avoidance in, in either spillage or efficiencies wherever they can find, incidental uses of the data that become not incidental, but actually evolve into requirements for them. That's going to then push how they handle that data.

Luka:

How do customers use those systems now, do you understand the flow of data today with some of these early adopters?

Charlton:

Not in detail. So my work while I was at Insitu with the ScanEagle, the closest we got to that was really doing wildfire monitoring, where we'd fly all night mapping a fire and produce digital maps throughout the night. In a pretty, we could do it basically on the hour because it was mostly automated. Which was a much faster pace of creating data than that particular end user was either accustomed to or really willing to accept, right? So a little bit ahead of its time, maybe, but in that case, the data flow was a major part of our effort because we had the sensor, we had the system, the drone, that was all well proven out, but creating a path for all the imagery we were pumping down from the system to then, in the background, create those maps that were a usable product for the end user was a major part of that effort, but it was solved within weeks, not within months or years, so it's definitely doable.

Luka:

Speaking of waivers and exemptions, what are some of the ones that you would highlight, or maybe what's worthwhile mentioning as it relates to the trends that we're seeing with waivers and exemptions?

Charlton:

Yeah, so there was a denial of a petition for exemption by Percepto that I found very interesting because, they didn't deny them the operation. They denied them the exemption because, they didn't need it. Right. So this is evidence that the FAA is learning, right? The petition was for, 107.15 and, 107. 49, which is the preflight, and in condition for safe operation rules under Part 107. And they petitioned for an exemption saying, hey, we've got an equivalent level of safety because we've got cameras, we've got other means of meeting these remotely. And the FAA said, well, we agree. You're meeting the rule. So therefore you don't need an exemption. Whereas previously the FAA would have said, yep, okay we grant you an exemption because the spirit and tenor of the rule was that there would be somebody on site doing those things, but that indicates that the FAA is actually, has an understanding that these, safety measures can be accomplished remotely and in some cases in an automated fashion. So I will expect to see the same sorts of things occur over time in beyond visual line of sight applications for the technical mitigations like radar or optical. If those technical mitigations are given waivers consistently, and we can bucket those waivers to say that these particular technical mitigations, are an equivalent level of safety for a human, either on the ground or in a cockpit to see other aircraft and therefore avoid them. Then I will expect that ultimately waivers and exemptions won't be required, by systems that meet a standard yet to be published.

Luka:

I was expecting that this would happen already with some of the waivers that we've seen with, BNSF and Skydio remote BVLOS operations and yet it's been, you know, several years and I haven't seen any meaningful, traction, to use that as a template for other waivers and exemptions.

Charlton:

Yeah. I think that's a cultural, what you're seeing as a result of a cultural resistance to those kinds of broad approvals. I think that is changing as well, and I'll call out AFS 700 as a very forward leaning office within flight standards that is intentionally making attempts to bucket these sorts of things and get out of the business of waivers and exemptions. There are multiple efforts there. One in particular, or one aspect of the effort that they're undertaking is to actually transition from an exemption process to a waiver process, which reduces the burden both on the FAA and the petitioner, because when you're doing an exemption process, you're actually petitioning for a change to rules.

Luka:

Yeah, I was going to follow up with that. Why don't you explain the difference between the two for those who might not understand.

Charlton:

Yeah, there are rules that are, eligible for waivers and then the rest of the rules have to be exempted. And, where the basic tenant for which ones are exempted and which ones are waiverable came from, I honestly don't know. It may have to do with the level of perceived risk in waiving or exempting. And, in order to exempt a rule, it's got to go into a rulemaking process. Therefore, it's got to be posted for public comment to change the rule. And it's a specific exemption for a specific time frame on a specific rule. Waivers can be granted without that process, so 91.113B is a good example. It's waverable. ATO is the owner of that rule. So ATO is the authority or the office of primary responsibility for waiving that rule. Flight Standards informs them on the technical, elements of why an application is credible. And then ATO makes a decision that they are, or are not going to waive that rule. But that's done outside of rulemaking, and therefore you don't typically engage the aviation rulemaking office or FAA legal or all the subject matter experts necessarily within an office like AFS 700 to have to review that. So it reduces the burden of process significantly if it's legal.

Luka:

And so exemptions take longer, but do they pave a better path for the next company to then obtain something similar?

Charlton:

Well, because there's a public, notification of denial or grant exemption, and that's available to everyone, it does help template things from an industry perspective. But honestly, the waivers and exemptions serve the same purpose within the FAA of granting authority for something that's outside the box to begin with. So, those can be used to inform the bucketing of operations that should be exempted or waived, and should be. I think the siloed nature of the FAA works against that in general. There are elements within the FAA, both within flight standards and I think within ATO, that would love to see fewer exemptions and waivers come in the door and would love to bucket it. I agree with you. I have not seen a concerted effort to put that all together. I'm not sure that's on anybody's, list of accomplishments that they have to achieve for this year to get that out the door as a, Hey, we've assessed these operations and we think we should bucket these approvals into something that is an equivalent level safety and here's the standard one can expect to have to meet. As an industry applicant.

Luka:

The FAA has been accelerating their waivers and exemptions since probably early 2023, which was a great sign. Where do you see, over the last several months, most of the waiver and exemption activity falling? Is it around, BVLOS? Is it around one to many type of operations? Are there any notable, things that you might want to highlight based on FAA's recent activity?

Charlton:

Yeah, it's a couple of areas. One is the 44807 exemption path, and that's for airworthiness. So that's the airworthiness of the reliability, durability, operational safety of the equipment. To go make revenue outside of Part 107, has to have an airworthy vehicle to go do that with. And, the type certification process has been, belabored, to say the least, and has not been, reactive enough or agile enough, to meet industry needs to get products out the door that can be used in revenue operations outside of Part 107. So, the FAA has taken the 44807 exemption process, which is an exemption for airworthiness, and applied that in a much more, I think there's still a methodology that's appropriate applied to it, but they've taken the criteria for type certification of smalls in this case, up to 1320 pounds, maybe even larger, and applied that type certification criteria, but instead of through a design certification lens through a flight standards operational lens. So rather than a full test and analysis of a system that goes down into the sub components of the system. It's really more of an assessment of the company will self declare, the manufacturer will self declare, yes, we have all these artifacts that show that we've done enough company testing and we documented how the system works adequately. And when the FAA looks at that, again, from a flight standards perspective, they then can agree without further test or further analysis that those artifacts are there. And it's basically putting the trust and confidence in the company that they've done their due diligence to have a product that's safe for its intended operation, rather than prove that out through a much longer process.

Luka:

That's a pretty big step for the FAA,

Charlton:

Yes, to put the responsibility back on the manufacturer and in kind back on the operator to go operate that said same said product, this was all the mantra for the durability and reliability type certification process was it was low risk. These were small systems, being used in concept of operations that, started out with low, risks, low hazards posed to people and property on the ground and other aircraft. Well, the, that under the AIR system, it never scaled back far enough to make it, really workable in a timely way. So Flight Standards has taken that and is doing it with it today.

Luka:

I'd like to know maybe, as a final question on this topic elaborate a bit more, who needs 44807? My understanding is that whoever intends to operate under Part 135, primarily this is for drone delivery type of operations, they need a 44807, otherwise, if somebody is operating under 107 and has a waiver for BVLOS operations, what does that open up in terms of commercial use cases, as opposed to going down the 44807 path?

Charlton:

So the Part 107 is the box, and it's driven by risk. And part of that risk is weight. So for better or worse, 55 pounds, right is the high end on weight under Part 107. If your piece of equipment, if your drone weighs more than 55 pounds, then to use it for commercial purposes, there's a trigger that pushes you into another realm of airworthiness, which would either be a type certificate or an exemption for airworthiness. So 44807. So weight is one of those. The other is, complexity of operation. So if you want to be up and away above 400 feet BVLOS, and you're going to integrate into the national airspace in a way that's much more like a manned aircraft today, then the complexity of that operation could drive you outside of Part 107 as well. And there are a number of factors that weigh into that, but basically it's one, revenue operations, and B does not fit into Part 107 for one reason or another.

Luka:

And so if the operation involves flying below 400 feet AGL and below 55 pounds, what is it that one can do commercially? Assuming a BVLOS waiver, what can be done commercially without requiring a 44807?

Charlton:

All the inspection work, right? Collecting data. but the, and I guess I should mention the other trigger there is if we're going to invoke Part 135, then that requires this airworthiness basis, even if it's below 400 feet or below 55 pounds, because 135 requires an airworthiness degree of, design assurance that's greater than 107, which has no design assurance associated with it.

Luka:

And so, this new path of obtaining a 44807, instead of going through the D&R process and the airworthiness crowd, the design, drill down with the FAA, but instead going down the Flight Standards and self declaration paths does that also resemble some kind of D&R or elements of it?

Charlton:

Yes, the criteria are actually exactly the same.

Luka:

And so, what are some of the lessons from that exemption history that the FAA is, taking notes on as it starts to form the contours of Part 108.

Charlton:

Well, if you recall the precursor or the vehicle by which the FAA approved operations prior to part 107 was part 333, right? It's section from the Re-Authorization act that Congress said, Thou shalt go do, make this possible. Same thing for 44807, and now 44807 is informing Part 108.

Luka:

Any more specific bits of information that you, maybe this is a good way to transition to, to Part 108.

Charlton:

Well, all of the, all of the exemptions that have been granted thus far, and you can pull all those down on the Federal Register, right? Or in DRS, the Document Resource Service within that, that points to all the FAA documentation. If you pull those down, you can see a very clear trail from the early exemptions on through subsequent exemptions, wherein the FAA in each grant of exemption references previous exemptions and what they decided. So there's actually, it would take some time to put it together, but there's actually a very clear trail of, we made a decision to grant this. And therefore we're going to grant this next one based on that. And occasionally they even, put in language that indicates that they learned something from the previous ones that they don't need to grant an exemption anymore for. So slowly but surely the layers of, oversight that the FAA wants to apply are being peeled back. And again, more responsibility being laid on the manufacturer and the operator.

Luka:

So based on all these exemptions, what do you expect part 108 will include?

Charlton:

I think it will refer to industry consensus standards as the basis for airworthiness, along with some basic criteria that we've seen come out of the D&R process. I think self declaration is going to be a major component of the airworthiness applications. That doesn't mean that you just toss out the information and the FAA agrees that you've declared. There's still a degree of oversight and approval. However, again, the bar is going to be different. It's just approval in a different manner, with less upfront oversight and probably more oversight on the back end. So that's the airworthiness piece in terms of the, airspace integration pieces that we may see out of Part 108, I expect that there'll be some kind of a nod to airspace below 400 feet or 500 feet, some low altitude, not a, maybe not a carve out, but an indication that there are appropriate uses for drones that will be effectively blanket approved if the operator meets certain criteria. And again, not sure what those criteria are going to be, but it's going to be a much more clear and easier to navigate path to getting those approvals.

Luka:

In terms of use cases, who do you see the most likely winners?

Charlton:

Drone delivery, is obviously, an easy one. The drone in the box folks, again, also, that will be covered, I think. And then, there will be some linear infrastructure pieces that are probably, gonna be beneficiaries, but I will say that just because of the sensors, at play, and the altitudes that they make sense to be used at for long linear infrastructure, that's probably still going to fall outside of, the BVLOS enablement that Part 108 brings to the table. That's my guess. And I have a a lot of curiosity myself about how the Part 135 world is going to be addressed through 108, because everyone agrees there are elements of Part 135 that are not completely appropriate for drone delivery operations, but I have yet to hear anybody say that they're willing to back off of the requirements in a, in any sort of detailed way and that sets up a tension between the current manned 135 world and drone delivery. If it's easier to get a Part 135 certificate under, Part 108, is it, should it really be as hard as it is today for a single aircraft operator under Part 135? I'm equally curious as everyone else is to see what the proposed rules are for, the operations certificates.

Jim:

First of all, I hope there's no quiz following this podcast, and then note to self, never get on a podcast with Luka asking questions. That was great. Hey, Charlton, it's been a year and a half. How have the use cases changed the most in the last year and a half? What's changed? What's been meaningful? Has the market grown or has it shrunk since we talked last and why?

Charlton:

Well, I think it has grown, maybe not in the ways that folks expected it to. Certainly there's a, informally referred to as the Key Site in Texas, where several drone delivery companies have collaborated, with some FAA involvement, to really, leverage or exercise what they envision as a viable way of coordinating, maybe a first touchpoint at a UTM in practice where they are, deconflicting each other's traffic and also with the waivers they've got for, beyond visual line of sight operations, employing all that to a very small scale, operation that is proving out the case that this can be done safely. And I think that indicates that there's growth in that segment, right or the potential for true growth in that segment. Once that's proven out there, I'm certain it's going to expand. And it's not really the drone manufacturers and operators that are waiting for this as much as it is the enterprise, consumer market, right? So Walmart and others that want to use this service are watching keenly to see who comes out on top and who is the best potential provider for them to use as a tool to get consumer products to market faster, cheaper, all the things that they, greener, all the things that they propose technology So the market

Jim:

hasn't necessarily changed in size, but provider's ability to address the market has matured because of what's going down in the Dallas area.

Charlton:

Yeah, I think it's a great use case that's demonstrating that given enough reigns the industry can solve these problems. The Regulatory Problems as well as the Technical Problems and prove what works and what doesn't.

Jim:

Have there been any other use case changes in the last 18 months? Any other applications for UAS that we didn't talk about a couple of years ago?

Charlton:

I, I don't know that there's been new ones, but there's been progress in gaining approvals for some of the traditionally sought after use cases. We've been involved with the Chevron company in, oil and gas and helping them get approvals for BVLOS waivers for long linear infrastructure using AATI's AI ranger system. And, that's not a very talked about waiver and exemption, but they can go fly in the San Joaquin Valley like any other airplane can, and it's a big piece of sky that they're approved to fly in. 4, 500 square miles, and no reason to believe that won't be expanded. So, that, to me, and maybe it's because we're so close to it, but that's a watershed event. Other approvals like that have occurred in the past, but they haven't been, replicatable, if you will. They haven't been, for either economic reasons, use case, end user reasons, or, regulatory reasons, there was too much headwind, too much friction to actually. make those operations scalable. this one may be different.

Luka:

What makes this one different?

Charlton:

Well, the onboard DAA, we proved the safety case for radar as the mitigation for 91.113B. So, I'm not aware of anybody else who's been able to do that, to date, and what this basically means is if that company can apply that system to solve that problem, and that's, a, replicatable, system and approval, then what's to stop these systems from going and flying in other similar airspace, other similar con ops, anywhere?

Luka:

That's really interesting. Can you speak a little bit more as to the FAA's threshold of, risk appetite or level of safety that this company with your assistance had to go and prove in terms of, hey, you need to be able to detect X percent of traffic out, X number of miles? How did you shape that conversation with the FAA?

Charlton:

Well, I can't speak to the details of the technical element of it. However, I will say that, the safety case met, what the FAA considered to be an equivalent level of safety to a pilot in the cockpit today. And that's still ill defined. You can't ask the FAA for what that bar is and get a straight answer. It's not published. That probably will be spoken to in Part 108, but I don't know if it's actually going to be addressed specifically, for operations above 400 feet AGL.

Luka:

Yeah, this is tangentially, this is so frustrating that these things cannot be talked about. I mean, those are safety level questions, considerations, thresholds. In whose interest is it to lock that information in a vault somewhere?

Charlton:

Well, from where I sit, it's all company proprietary information. So as first movers in this space, if you will, or leaning forward, Chevron has an advantage over their competitors if they can leverage this technology. There is a desire within the FAA to, through these projects, inform themselves in a way where they can, agree on a standard internally. The, one of the challenges is that the industry groups that, that want to, that have proposed standards, RTCA and ASTM, those haven't been accepted by the FAA formally for detect and avoid, but they are used to inform decisions. And I think what you're going to see as a result of 108 being published is the FAA is going to be compelled to come to some conclusions about what equivalent levels of safety are, and use those standards, and maybe other industry standards, to actually publish what's good enough.

Peter:

Charlton, is there historical precedent or examples in the past in the aviation industry where safety has been a secret and it's been a competitive edge for the players?

Charlton:

I, I'm sure there, have been, I would imagine that with, propulsion technology, as both internal combustion engines and, jet engines, turbine engines evolved that the folks who were the first to employ those in meaningful ways, the airlines as end users depended on the manufacturers to come up with solutions that gave them a competitive edge, but ultimately, in the same way that I think this will pan out those products became well known certified and the standards for by which they were, certified became regulation.

Peter:

I can see both sides of the argument, obviously, and I can see how, in a very sort of transitory way, you could have a short period of time where safety is proprietary because it's evolving so fast. But if the regulations that make the same standards available to everybody take so many years to roll out, we're not in a transitory phase. We're in a five year plus drawn out period of time where, safety continues to be treated like a proprietary advantage and held close.

Charlton:

Well, maybe this approval is also an indicator that the FAA has, at least internally, decided that there is some kind of level that we can, seek as a standard for this type of operation. I would say that's gotta be a symptom of some corporate acceptance of an equivalent level of safety that can then be translated into rulemaking and, but time will tell, there, there are still economics to work through with this operation like any other operation until it matures and becomes an operation that's routine, recurring, and non dramatic in any way. It's still, emerging technology.

Luka:

What airplane was this with? Is this one of the, tweeners as you would define them?

Charlton:

Yeah, the, AI

Luka:

AI Ranger. Okay.

Charlton:

AATI's product.

Luka:

What's the latest in that category of aircraft in terms of their integration into the NAS, any notable, other than this, any notable developments, waivers, or path forward that you anticipate these kinds of heavier aircraft being incorporated into the NAS?

Charlton:

Not that I've readily witnessed, I think the vast majority of that sector has turned to DoD. It turned to DoD some time ago, and is relying on that as their target market. we've even seen that shift in the advanced air mobility sector, right? Where, getting to market in the civil context or commercial context, has driven them to actually see what's going on the DoD side, and the use cases are not so dissimilar. There are additional and different challenges, but the use case is there, and the DoD has a need, and they're willing to pay for it without as much rigor put into the regulatory side, right? There's probably going to be a lot of fast followers. Right, so once it's public knowledge that ATI has this approval and at a very high level, what they did to achieve it. So, I would imagine that there are other folks out there working on similar approvals. I guess one, one other product that comes to mind now that you mentioned it is the, Phoenix has used the Swiss drones and I can't remember the designation for that, but that's long linear infrastructure with rotocraft that uses shielding of the power transmission lines as its primary mitigation for BVLOS. And that has its limitations as well, but it's, it had been good enough for that use case. And that CONOP, and apparently it gets the data that, the end user

Luka:

needs. Do you see anyone taking the approach that MQ 9s in the Air Force and what is it, California National Guard and the NY National Guard, where they're using ground based radars to spiral up into Class A airspace and then, get integrated a lot more easier, it seems like it's a problem that can be streamlined before the, below 10, 000 foot and above 400 feet airspace integration

Charlton:

There are approvals like that exist for the high altitude pseudo satellite sector. So, getting to upper E, above flight level 600, is enabled through technical means based on the ground rather than trying to carry that equipment on the aircraft. And then operationally, waiving 91113B in Upper E, solves the rest of the problem for transit, which certainly I think is an appropriate level of safety and an appropriate solution today. I think it's going to be a long time before Upper E becomes crowded enough that the risk of collision is as high as it needs, it needs additional mitigations, operational mitigations, communication between entities that are using out airspace today should be very doable.

Jim:

Charlton, when we spoke last, you gave some really good balanced, measured advice to, those who are going to the FAA to get certification. You gave from the FAA's perspective and from the operator's perspective. So since we talked, Mike Whitaker is now the Administrator and he came from our industry. What's changed with the FAA? What are the, what, how would you change your advice or alter your advice, to, those seeking certification from what you would have said 18 months ago, especially with Mike's, leading the helm. I think one piece of advice you mentioned was there's leadership that's needed there. And, well, clearly there's a leader and he's doing a great job and he came from industry. So how has that manifested to the industry? What's improved in and perhaps what's degraded?

Charlton:

I haven't seen anything degrade. I think what's improved is we mentioned the exemption process. I think that is the pressure relief valve for type certification right now. Unless an applicant comes to us with a product that's way too heavy, way too complex of a concept of operations, if they're going to try to carry people, then they need a type certificate. But otherwise, just the whole enablement through 44807, makes it almost prohibitive to attempt a type certification under at least 3, 000 pounds gross weight for any equipment or any product today, in my opinion. So, what that means is that there still has to be some way of addressing why AIR is stuck, if you will and the type certification for highly automated, lower risk systems, is not happening through AIR currently. The Percepto's TC that they just got was a five year process. So that's an indicator that it's still not where it needs to be in terms of right sizing the process to the risk, to the application, to the equipment. And I hope Mike Whitaker can pay attention to that. Obviously, there are a number of other topics that I'm sure are on his doorstep every day, probably take priority over whether or not the drone industry is getting type certificates, particularly when there's a path via exemption that everyone anticipates is going to flow into a rule that will then alleviate some of that pressure through alternative means. Now that doesn't solve the problem for products below what we currently refer to as advanced air mobility, where part 135 operations to carry people around is the goal, below that, but above, the drone delivery, the linear infrastructure, part 91 world, part 135 world on the smalls. There's a gap and until such time as that's fully addressed, I think there's still going to be a lot of friction for products that, are, that need a design assurance level that's higher, but, going through a type certificate process is going to be, too high of a burden currently.

Luka:

Charlton, what's your assessment of the agriculture use case? Is that something that you're seeing on the rise? What are the drivers?

Charlton:

I think it's a, an unsung success story from both that world and the FAAs. I don't remember what the numbers are off the top of my head, but as I recall, and this is a question for Joe Marra because it's his office that spearheaded this. Again, AFS 700, but they've approved hundreds and hundreds of exemptions, for agriculture in a very efficient manner because if they fit the bill, if they fit within the box of what they've seen historically as a low altitude, low risk operation, it sails through. Point being that, that agriculture, I think will be included also. It's a Part 137 and that the airworthiness for that will be rolled up under Part 108, quite readily.

Luka:

What are the waivers associated with an agricultural use case? those guys are mostly within visual line of sight, so there's no need for that. Maybe some one too many many type of operation if there's some swarming involved, but otherwise it seems like a, relatively simple use case.

Charlton:

That said, of course, the industry is going to grow the products in size and they're going to want, expanded operational approvals. So you're going to bump up against part 91 for the airspace approvals and see and avoid will become an issue. And then, okay, maybe that can be done with, ground based, systems in a localized area. Does that really scale? I'm sure that industry has designs on covering vast areas of farmland in a much more efficient way. So again, we'll bump up against trying to bucket approvals based on historical approvals to go through. What's the equivalent level of safety going to be? Hopefully that's informed by some 91 operations. But same problem to solve, you've got to avoid hitting other aircraft. In this case, the non cooperative aircraft in that segment are also competing for that airspace in a very real way, and they're not typically, or often they're not, electronically conspicuous, if you will. So there's no actual requirement for crop dusting to use ADS B out, and since there's no requirement, sometimes they do not because there historically haven't been a lot of other airplanes down there to bump into. And locally those operations know more or less who's doing what, so there's a community of, organic communication that occurs. If you roll into that same area as a drone provider of a precision agriculture or spraying, and you're not integrated with that same community, then there's, there could be an operational hazard created by the introduction to the new technology.

Peter:

And incidentally, on the, AI stuff, if you want to work that into the conversation,

Charlton:

From my perspective, my, the context that it fits in with me is that, just with this prompt, we can ask AI that comes up with solutions that aren't that far off the mark. So if we do product development, offline with products that are designed like Garmin's Autoland to automate, off nominal and emergency situations. It's not a stretch to see how that can be incorporated. into any of these highly automated products, AAM drones, whatever. and I, as I mentioned, if I had AI driving me around when my motor pooped on me at 400 feet, it's a technical term, by the way. maybe it would have, maintained my airspeed a little better than I did, and maybe it would have managed the energy of the aircraft and gotten me back to a predetermined landing spot that it had already thought about. For me, it's looking out the left side of the airplane saying there's a runway and that's a better alternative than anything else. And that all happens very quickly, right? It's not a stretch to see that if we can make a simple prompt and AI comes up with solutions that are at least in keeping with the practices that we would expect a pilot to come up with today, then refining that isn't a huge stretch. So that applies to detect and avoid. That applies to navigation, just basic navigation. Companies like the Daedalean are doing that right now.

Peter:

Yeah. I guess the caveat being that you would need a AI that is not just a large language model, like for instance, GPT 4 or Claude, because those are just trained on text and, but you need an AI that has a true understanding of

Charlton:

Terrain, navigation airspace, all the Yes. But

Peter:

It, it needs to understand that innately. And so you can prompt Claude and you can say, Hey, I'm 15 miles from a certain airport and what should I do? And it will come up with an answer that's mostly right. I guess what you would want to aim for is a spatially aware AI, and then also one that is respectful of the checklists because the checklists are probably not something that you want to go and edit, midway through an in flight emergency.

Charlton:

absolutely.

Peter:

You wanna have the checklist kind of hardcoded into it, for those procedures, and then have the AI kind of sit above that and help exercise the other aspects of judgment that are involved in handling that situation.

Charlton:

Well, yeah, and it's, and the result needs to be a, deterministic or a predictable outcome, right? So I think a lot of the development that uses AI will happen all offline in advance, so that what actually resides on the aircraft. It's going to be predictable and how that covers, 90 X percent of the use cases. Or the functions that need to occur from startup to shutdown. it's not a small problem to solve, but the tools are emerging that indicate it is solvable.

Peter:

Are you following, the development of tools like what Beacon AI is bringing forward?

Charlton:

No, my exposure to this is really mostly through our work with the Daedalean.

Luka:

When I think about this, I mostly focus on what the product needs to look like and what the price point needs to be in order to, generate demand on the GA market. Because, there's a lot of innovation that we've seen, that is targeting general aviation and rightfully so, because GA has probably the most acute safety problem. It's also one segment that doesn't want to pay, even if it's, even if it's mandated, as we've seen.

Charlton:

Well, the forcing function is really AAM, right? If the AAM model that's been proposed is going to work, then everything's got to be automated. And what does everything, if you're going to yank a pilot out of that aircraft that's carrying passengers, all those functions have to be accounted for. And so, There's no way to do that without applying AI to the problem.

Luka:

In the closing minutes, what are some of the things that you are excited about in the coming months or opportunities that you would vector our crowd of innovators towards.

Charlton:

I'm going to come right back to the 44807 exemption process. Cause that's really in full swing. It's probably an underreported success story. It's a clear path. It's predictable, not mysterious. Whatever evolution is happening on it because it is evolving and changing is generally friendly to industry and not more restrictive. So again, the concept of putting more responsibility on operators, putting more responsibility on, manufacturers that's happening. So anything that doesn't fit under part 107 now really has a, as a clear and arguably easier path than it may ever have to getting into service. 107 may or may not make things easier. It may become more predictable, or better understood, but maybe not easier because I think the FAA is very open right now to operations that are different. So that's probably what I'm most excited about. And really the 1, 320 pounds that people talk about may not be the upper end for all that. So, the bigger gross weight platforms that we see coming to the fore. If you don't have to get a type certificate and you can get it into service under an exemption and operational approval, then you can get some work done.

Jim:

Last time I asked you a question, I'm going to ask the same one. So you're going to invest in an area of advanced air mobility, UAS, UTM, where would your money go today?

Charlton:

I would seek out companies that were trying to solve only parts of the problem first. So, any of the advanced air mobility companies that are starting out with conventional takeoff and landing, and maybe just solving the propulsion piece first. Manned aircraft, electric propulsion with a path to eVTOL and a path to higher levels of automation, with removing the pilot well down the road, right, because that step function is hard to measure but a new way of, of applying electric propulsion. Got it. we can do that. Then if you want to add on top of that, we've wrapped our heads around propulsion. Now let's really apply that lesson learned to vertical takeoff and landing. Okay, that's a small enough piece of the puzzle that we can tackle that in a meaningful way that's more predictable. Structures is not mysterious. so, advanced composite construction is known. So that shouldn't scare anyone. It's really the removal of the pilot. If that's the goal right off the bat, then I'd say that's a lofty goal and, challenging and scary thing from an investor perspective as a regulatory guy.

Jim:

Charlton, anything else you were dying to tell us we haven't asked you?

Charlton:

Yeah, there's plenty of stuff we didn't talk about, but I don't think any of it's super critical. there's just a ton of, emerging topics, so maybe we come back in a year or two and see if anything's actually changed.

Luka:

Well, Charlton, thank you so much for carving out some time to come back and chat with us.

Charlton:

Yeah. Thank you guys.