Autonomous skies: how AI is powering the next generation of drones
23:41 Share

Welcome to today's episode of Lexicon. I'm Christopher McFadden, contributing writer for Interesting Engineering. Today we're joined by Bill Irby, CEO of AgEagle, to explore the future of drones, AI, and autonomy.

What's next for these high-tech unmanned systems, from military advancements to commercial breakthroughs? Let's find out. Gift yourself knowledge. iPlus is a premium subscription that unlocks exclusive access to cutting-edge stories, expert insights and breakthroughs in science, technology and innovation. Stay ahead with the knowledge that shapes the future. Bill, thanks for joining us. How are you today?

I'm doing fine. Great day and thanks for having me on. Our pleasure. For our audience's benefit, can you tell us a little bit about yourself, please? Sure. You know, spent time, I grew up in the United States on the East Coast. A lot of time going through school and then in the Marine Corps for five years. And then I came into the defense industry working initially as an engineer through project management and just kind of moved into profit loss leadership over time.

Uh, most of my career has been focused on technical development and or profit loss leadership of businesses that had a high content of technical development in them. Most of my time has, uh, actually been with the drone industry or related to the drone industry in some fashion, but it's, it's great to be here with you today and, uh, we'll see where we go. Did you miss the days, uh, your days as Marine? Uh, yes and no. Okay.

When did you leave? How long ago? Oh, long time ago. I don't want to tell you because that'll reveal how old I am. So it was a very long time ago. Multiple decades. Fair enough. Okay then. So Ad Eagle, as soon as a key player in the drone industry, from your perspective, how has the landscape of autonomous drones evolved over the last few years?

Very rapidly. Um, you know, autonomy is, uh, now it's being talked about in terms of artificial intelligence. The, the term used used to be autonomy. Now it's artificial intelligence of all, of all types and sorts. Um, but I think it's evolving very rapidly. There are many companies out there that, that do it. There are a few that specialize in it. Uh,

you know, that are involved in the drone industry that really have a high content of artificial intelligence. So I think you're going to see it continue to evolve rapidly over time. Okay. So, um,

Can I say, so if you found like the civilian sector, uh, non, non-military side civilian sectors, uh, advanced as quickly or the military side more quickly or about the same. You know, I haven't done a comparison, but I would say in all likelihood, it's about the same because the, you know, the military tries to draw technologies from the commercial world, especially in an area like unmanned systems. Um, so dual use has been a common theme in the industry.

I guess it's not cross-pollination because the drones wouldn't exist without military development and vice versa. Yeah, yeah, exactly. Fair enough. Okay, so drones are being used in industries ranging from agriculture to defense. Which sectors do you see as the biggest adopters right now? And where do you think the next wave of drone applications will emerge? You know, it's interesting you ask because if you look historically over decades, clearly the military has been in the lead on the advancement of drone technologies.

In the recent years, though, you see a huge investment in the commercial industry, right? In many states right now are already doing, uh, cargo delivery by drone. Uh, they're, they're allowing, uh, their air skies, uh, to be used, uh,

for delivery of product and also other, other purposes for drones. You know, I've seen some flying over my house doing land surveys and things like that. Um, so it's advancing rapidly in a commercial, uh, commercial environment. And I think there's going to be a lot more money there ultimately than, than what comes through defense. That's not the way it started though. Hmm. So, uh, the ones that he has for the, for insurance, did you say?

Yes, actually they are. They're not assessing my property, but they've flown over our house looking at neighbors' properties primarily for insurance purposes. So yes. Are people handling that, do they find that a bit intrusive? It's just one of those things.

Yes, definitely intrusive, not happy about it. Fair enough. On that subject, where were those random drone sightings? Was it California? Where was it? A couple of months ago, was it California? Did they ever figure out what those were? No, it was the East Coast. It was off the coast of New Jersey. Sorry. And what I have read is it ended up being some government testing that was going on. I forget whatever the agency it was.

I view that as a skeptical comment, though, because it took so long for the government to say anything. And there were clearly people that were recognizing or thinking they were drones when they were actually commercial aircraft. Because when you look at the videos, it was very clear what the units were. So I think there was a lot of flurry of information and misunderstanding out there. I'm not sure that they were all government, though.

So these, one of the craziest ones I heard is they were looking for a broken arrow, a fallen, uh, nuclear missile or something. Yeah. That's, that's a, that is definitely a conspiracy theory of some sort. Interesting. Anyway. Okay. Uh, so how is, uh, ag Eagle balancing innovation with regulatory challenges, particularly as governments worldwide refine their approach. It looks like we have. Yeah. I'm sorry. I lost you for a few seconds there. What was the question?

So how is AgEagle balancing innovation with regulatory challenges, particularly as governments worldwide refine their approach to drone laws? Yeah, really, really good question. So, you know, regulations are advancing at the speed that regulations do. And when I say that comment, I mean, you know, through EASA and FAA in terms of airspace regulations.

Not only that, but there are new regulations being put in place to secure the supply chain. So I would say all these are advancing at a certain pace. We are pushing hard to keep track of that. The biggest challenge for us has been we're a small company, right? We have limited resources to go advance the way we develop the technology and to pay attention. Whereas the bigger companies are better equipped to actually influence very directly the regulations.

We are involved with AUVSI. We're a member of AUVSI. In fact, I'm part of the board as the chairman of the board of that association. So from that perspective, we do have a voice that is influencing the overall community in the way regulations are formed. I see. But mainly you're kind of at the mercy of the regulators, really. You've got to adapt to change, any changes. Of course. You're going to see that come, you know,

all the time. You know, when you talk about people like FAA or defense authorities that, that issue regulations, you're never going to get away from that. So we are going to have to respond when we don't see, you know, when we don't foresee changes. Um, but either way, we're doing our best to stay prepared. Fair enough. What can you do really? Um, right.

So, uh, AgEagle, uh, appears to focus on professional fixed wing drones, uh, high resolution sensors and intelligent software. Uh, what are the biggest technological breakthroughs you're most excited about? Yeah. Interesting. Um, great question. Cause our drones typically came out of the photogrammetry market. Uh, we did a lot of, uh, you know, mapping missions with our previous products. Two of our existing products are called the EB-TAC and the EB-X, uh,

Those started as photogrammetry drones, and then the EVTAC moved a little bit forward with doing some other surveillance activities, surveillance missions. And then we developed something called the EVVision, which was targeted specifically for military use, military and public safety. But it is a real-time, you know, full-video, full-motion video surveillance drone, both in RGB, you know, full-motion, full-color video, and thermal video.

So that system is the most advanced that we have. We actually have

It may not sound like a big deal to most of the market, but for a drone of our size, we use a payload that's actually retractable into the fuselage and deploys. And we've actually put some water protection and sand and dust protection around it. So pretty advanced. We're very happy with how that system is achieving sales. So that's primarily into military, as you said.

Primarily, yes. And we've equipped it such that it's a low visibility drone. It's hard to visually detect when it's flying at altitude. It's also low profile, a low noise signature. We've got commercial encryption in the drone to protect the communication link. We're doing many, many things like that. Presumably as FPV. It is FPV.

is not what you formally consider an FPV drone. Those have a certain consideration these days and normally used with systems that, you know, turn into ordinance of some sort. That's not what we're doing. Yeah, fair enough. Well, that kind of answers my next question. I was wondering if you've sold outside of your customers, most of them, if you could tell us, is primarily U.S.-based or do you have international customers? International customers. Yeah, we've sold our systems in the U.S.,

uh, for some time, but, uh, actually our, our factory that builds the drones is in Switzerland. So we've got a more direct presence with European nations and, uh, we've sold a pretty high quantity of these things to France. We've sold to UAE, we've sold to Germany, we've sold to, you know, a number of countries in Europe. Excellent. What about Ukraine? No, um, no, no. Interestingly enough, you know, because of the Swiss is a new Switzerland's a neutral country, right? Because they've declared neutrality. They've been hesitant to let

their systems be sold to any country that's in the middle of conflict and particularly they've blocked Ukraine. Fair enough. Fair enough then. So one of the main concerns with drone adoption is autonomy and reliability. What advances be made in AI and automation to make drones more efficient and independent? Yeah, great question. We've got three modes in our drones right now that are, you know, called autonomy at some level, right?

We have an ability to do pre-programmed routes, but we also have an ability to fly a drone that keeps its target in view. In other words, if I say, okay, while the drone's in flight, okay, you know, you found a target, you're looking at X.

You go into another mode in the drone that enables it to automatically navigate around just so it can keep eyes on that target. That's a level of autonomy. What we're endeavoring to, we're pushing toward is being able to do things like swarm automatically. We haven't demonstrated that yet, although we're proceeding forward with development.

Fantastic. So it's, um, was it fully autonomous from takeoff? Well, autonomous from takeoff to landing, or you need an operator to get to an area, then it takes over itself.

So you need an operator to launch it. Once the operator launches it, the drone automatically climbs to altitude and it'll do circles until you send it off on its mission. So you can send it off on multiple mission types, pre-programmed route. You can manually fly it or you can, as I mentioned, you can direct it toward a particular target and it can automatically navigate. Interesting. It's kind of the inverse of a real pilot. Yeah.

Isn't it? Kind of, yeah. Fascinating. Wonder how advanced it will get in the future. Okay. There's been a lot of discussion around drones integrating with broader smart city and defense ecosystems. Do you see a future where drones become a standard part of urban infrastructure? Absolutely, yes. I think you're going to see as time goes on, you're going to see drones with an auto launch capability from building tops, auto recovery,

Um, I do believe you're going to continue to see mobile systems that can launch drones in different environments, including urban environments. Now, to be clear, most of the drones that are used in the urban environment, um, are, are VTOL, right? Vertical takeoff and landing. Uh, our particular drone at the current time is not a VTOL system. It's, it's hand launch. It climbs to altitude and then it's recovered, uh, by coming in for a short landing period.

Okay. What's this flight? Obviously battery powered. How long does the battery last? I can longer to be airborne. We advertise 90 minutes. We've actually done significantly longer than 90 minutes in certain scenarios. So yeah. Okay. How does it, it's obviously on VTOL. So kind of what skid lands or something. It's skid lands. Yeah. It's a very short skid landing. And, uh, and it's interesting when you watch it on video, uh,

It's really, you know, people would be concerned about, you know, potential damage to the aircraft or it running into something. But it's very, very accurate. When you, when you tell it where to land, it'll come in with a, you know, a relative, you know, decline that gives it the right, you know, trajectory coming into the land. And once it touches the ground, unless there's ice or something, it, it, it skids maybe about six feet.

But we've been able to land it within, you know, two meters of the designated landing spot. So it's very accurate. And in fact, something I'll tell you about the way the system was designed, they were very concerned about the ruggedness of the system, given the way that it lands. The materials that it's made out of, however, it's very light. It doesn't come in with a heavy load.

Um, when we test it, we've tested it before they ever released and launched the EB vision. We had a, a launch system that would actually propel it as if it were taking a hard landing on gravel and they tested it. They flew it like, I forget the exact number, but something like 385 times just to make sure that the drone was, uh,

you know, adequately rugged and, and it's, it survives many, many, you know, even crashes. Uh, so it's a very robust system. Fantastic. Is it modular? Uh, if there is any damage, quite simple to repair. Yeah. You can simply replace the wings. Uh, you can replace components in it, you know, battery sensors, if need be. Hmm. Excellent. Can we think of that old world two rocket plane? The chipmunk's head. I forget the name of it. Comet was it?

I hope not to be related to that, but anyway. No, I know. It's a vision that came into my mind. So how do you envision, pardon me, envision drone technology evolving over the next decade? Are we looking at complete autonomy, swarm intelligence, or even closer integration with AI-driven robotics? I think you're going to see swarming become, you know, a relatively automated function. When you talk about complete autonomy, I don't see that coming in the next 10 years. The reason I say is, you know,

Airborne flight is inherently risky, and I cannot see humans being comfortable with fully automated artificial intelligence being able to fly these systems without some human input or human oversight. I also don't believe that the FAA or EASA is going to be able to support that in a one-decade timeframe. Over two, three decades, we may see the technology advance where everybody's comfortable with that, but I doubt it in the next 10 years.

Yeah. I can see it's getting used to it, isn't it? Plus I wonder how many other, um, like say unions, I don't, for example, for argument's sake with, um, delivery drones. I don't know if they have any unions, if they get involved and are there, you have to keep a human in the loop to save our jobs. Interesting. You said that because when the, when we were starting to push regulations with the FAA, right, we actually had support from the American pilots association, um,

which was a little bit unexpected because we expected resistance given that we're, you know, the industry was pushing for unmanned systems to be used for commercial purposes inside U.S. skies. We expected pushback to come from pilots associations, et cetera. And we actually saw a very cooperative, you know, environment where they got involved in a regulation setting and working with the FAA to approve use.

So it actually turned out to being very surprising and positive. Well, that's interesting. Very interesting. As you say, you would expect resistance, right? Yeah. And I think the biggest resistance you would see on cargo drones would be from the truckers, to be frank, over time. And I've seen no evidence of that to this point. So with regards to...

So with truckers, it's more going to be, they're more at risk from autonomous lorries rather than airborne drones, or are we going to get to a stage where you have enormous drones? I think either or, right? If you're taking more of the cargo airborne or more of the cargo through automated ground vehicles, certainly you're going to see pushback from the truckers unions in terms of job loss. But it's like anything else, over time, the technology is going to demand at some point that,

It's just how it is. And it's going to have to be accepted at some point, my opinion. Well, that's fair enough. Would you envisage then for long distance freight, um, some kind of, I don't know if they exist, autonomous blimps or airships. Do they exist now? I'm not sure.

Um, there may be, uh, I'm not that aware of autonomous airships at this point in time, but it's certainly not beyond imagination. That'd be cool. That would be cool. Um, okay. Uh, with the rise of drone surveillance and military applications, uh, there are concerns about privacy, which we kind of touched on earlier, security and ethical use. How does AgEagle approach these challenges?

Uh, we, we definitely look at government regulations and we also consider internally. There are a number of things that we've looked at in terms of ethical use of drones where we've internally regulated ourselves not to do certain things. Um, so, but I think the, the, the biggest regulator is going to be the government ultimately, uh, that puts rules around where you can fly, what purposes you can fly for and how you use the system.

Would that be flight exclusion zones or limits on height, number of drones in the air? It could be either or. I don't want to imagine what the government is going to demand, but ultimately anything that would allow somebody to, let's say, spy on a neighbor or spy on somebody inside their own country, whatever.

the government is going to put regulations in place that prevent that at some level. Yeah, that makes sense. Rightly so. Yeah. Especially the media. If the news still around in the future, they're more likely to use drones on a, for instead of helicopters for footage and stuff. And another thing that's popped into my head, I've always promised flying cars. I've been for the last, I don't know, a hundred years. Um, that's going to be an interesting, uh,

uh, kind of battle for the airspace, isn't it? If you've got flying cars and drones, you're gonna have to have designated spaces lanes as it works. It's already complicated, right? You've got different corridors in the airspace that are, that are segregated for different things. Once we have flying cars and delivery drones and other types of drones and manned aircraft, all in the same space, all sharing space.

I will say two things. Number one, the regulations need to be in place and our air traffic control and monitoring systems need to be much more advanced than they are today. Absolutely. I guess that'll be a place for artificial intelligence, I presume, for the actual muck. Most definitely. Yeah, it'd be too complicated here. And...

So last question, what advice would you give to startups and entrepreneurs looking to enter the drawing industry? What are the biggest pitfalls and opportunities they should be aware of? The biggest advice I would say right now is I say I can foresee the industry consolidating over the next 10 years. You know, it's already been through one round of consolidation. Once the tech bubble kind of started or the techs, you know,

Tech fire was lit under drones, you know, say 10 years ago or so. There was already one round of consolidation where a lot of companies had stood up, figured out that they couldn't support themselves. They either went out of business or merged with somebody else. I see another round of that coming because the, the, the number of drone companies out there has grown pretty rapidly and it's not hard to start one of these companies. It is hard to sustain it. If you don't have, you know, big volume sales, uh,

Um, so I can see definitely another round of consolidation coming. And what I'd like to do is make sure that we take advantage of that in terms of improving our portfolio. Absolutely. So who do you think would be the winners in the end? It'd be a big defense contractor is going to buy everyone out or?

Um, they're not going to buy everyone out, but I do think, you know, that that's clearly been a trend in the industry, uh, where the big defense guys do buy other companies. I think you may see some medium companies merge, right. Or acquire one another to become big companies. I think of, you know, just three, four years ago, L3 and Harris merge, right. And then L3, uh, went on a, another buying spree and integrated several companies into it. So, um,

You may see some of that. Yeah. I mean, if and when the Ukraine war is finished, finally over, interesting to see the ecosystem there, how it develops. They're going to have one monopoly. It's going to monopolize there as well, presumably, but how an international company is going to get involved.

It's a really good, really, really good question and something to think about. You know, if you look at all the use of drones inside Ukraine today, I think clearly what you're going to see with the war coming to an end is that demand shrinking very rapidly. So companies that have been in high production mode to support Ukraine are all of a sudden going to be left with drones.

nothing in that particular project area. So it's going to put pressure on the companies. Again, that may help with consolidation in some sense. Yeah, be adapt or die, right? Correct. The technology can be converted for civilian use relatively easily just without the bonds. That's right. Yeah. Yeah.

Uh, okay. Uh, very succinct answers. Very, that's pretty quick. Is there anything else, um, you'd like to mention we haven't discussed? You feel is important. The good thing I would say is keep an eye on AgEagle cause we've got some big things coming up and, um, we're going to make some announcements a little bit later in the year, which I think will keep everybody interested. And I really appreciate you having me on. Not a problem. Uh, our pleasure. Um, well, Bill, with that then, uh, thank you for your time. That was very, very interesting.

Enjoy your day. Thanks a lot for having me. Thank you. Also, don't forget to subscribe to IE Plus for premium insights and exclusive content.