Shield AI has unveiled X-BAT, a stealthy jet-powered ‘autonomous fighter’ designed to take off vertically and land the same way, tail first, after completing its mission. The company is best known for its Hivemind autonomy software and its much less complex, but combat-proven V-BAT vertical takeoff and landing drone. Now it wants to have a very disruptive impact on the growing marketplace for Collaborative Combat Aircraft (CCA) and Unmanned Combat Air Vehicle (UCAV) type drones with a design that aims to offer a totally different level of mission flexibility from launch and recovery points on land or at sea.
Ahead of today’s reveal, TWZ had a candid and in-depth conversation about the vertical takeoff and landing (VTOL) X-BAT, its genesis, features, benefits, and the potential roadblocks to making it a reality, with Shield AI’s Armor Harris, Senior Vice President and General Manager of the company’s aircraft division.
Prior to joining Shield AI in 2024, Harris had occupied several senior positions at SpaceX, including senior director of its Starshield government-sales-focused business unit and head engineer for the Starlink satellite internet constellation. Perhaps most importantly, he had a key role in the development of the vertical landing capability for the Falcon 9 reusable space launch rocket.
To provide some basic details first, Shield AI says the fully runway-independent X-BAT, with its ‘cranked kite’ planform, will be 26 feet long, have a wingspan of 39 feet, and be 4.7 feet tall. The drone, powered by a single afterburning jet engine, will have a maximum range of 2,000 nautical miles and a service ceiling of around 50,000 feet. It has a highly modular design with a focus on open mission system architecture to make it easier to integrate new and improved capabilities and functionality down the line.
What Shield AI is attempting to pull off with X-BAT is no small task and there will certainly be skeptics along the way, and maybe justifiably so, as the company tries to break into the higher-end air combat aircraft space.
With all that being said, let’s get into the interview, with Harris first laying out the basic case for X-BAT.
Armor Harris: So, we’ve been approaching trying to solve the biggest problem that the United States has today, which is how do we counter the rise of our peer adversaries, i.e. China, and their ability to outspend and out-produce us in the military technology of the 21st Century? And the way to do that is with America’s fundamental advantage, which is innovation. So we’ve developed an innovative aircraft, that is not just an aircraft, but is a weapon system that fundamentally changes the balance of power in the Pacific.
What it does is it basically brings the capability of something like an F-35 or a comparable fifth-gen [fighter], puts it in a vertical takeoff and landing package, and then delivers it at a tenth of the cost, life-cycle[-wise] of a fifth-gen. So, breaking the cost curve. It enables us to counter China’s fifth and sixth-gen aircraft with something that comes at a fraction of the price.
It’s really fundamentally three things. So, one, it’s vertical takeoff and landing. And the reason why vertical takeoff and landing is so important is that in all of our wargames, we lose more aircraft on the ground than we do in the air. And over the last 20 or 30 years, the United States has spent a considerable amount of money making aircraft incredibly survivable in the air. More and more low observable, more and more advanced technologies, to be survivable in the air. What China has done is said, okay, cool, you guys keep going down that road, we’re just going to take them all out on the ground before they even get there. Or we’re going to deny the ability for those aircraft to even get into the fight, because tactical aircraft are heavily dependent on tankers, to cover the ranges in the Pacific. So, we’re going to prevent you from basing them too close, which means you’re going to rely on the tankers, and we’re going to develop a lot of ways to hold those tankers at risk so that you can’t even get them into the theater.
VTOL sidesteps both of those challenges because it enables survivability on the ground. So if you can be survivable on the ground, in addition to being survivable in the air, then it prevents them from being able to target you on the ground by just cratering your runways. And then it enables you to base them close enough that you don’t need the tanker support, so you’re able to generate that air power without that tanker train. So the VTOL is pretty game-changing in that regard.
The second thing is that the aircraft is multi-role. So, it’s capable of doing air-to-air, air-to-surface, air-to-ground, and electronic warfare missions, in addition to ISR [intelligence, surveillance, and reconnaissance]. And fundamentally, the Department of War and the United States can only have so much money. It’s hard for us to continue to invest in bespoke capabilities for specific mission sets, because over, say, the 20-year lifespan, or more, of a system, the threats evolve, right? We went from fighting terrorists in the Middle East to countering the rise of China in the Pacific and, oh, by the way, we’re also dealing with adversarial Russia in Eastern Europe, which are basically three different problem sets. And if we’re trying to solve that with single-mission weapon systems, that’s a pretty challenging thing to do. So, the ability to be multi-role and adapt to the evolving threat landscape we hold as being pretty important.
And then the third thing is that it’s really the first weapon system that is being ground-up designed around autonomy in the air. So you’re probably broadly familiar with the US Air Force’s Increment 1 CCAs. Those are fundamentally tied to the manned ‘quarterback’ to be not only a C2 [command and control] node, but also to do some other things. X-BAT is designed to have the size, weight, and power needed to carry all the sensors and gear needed to be truly standalone, and also leveraging the autonomy that shield has built for the big jet programs that we can’t talk about too much due to classifications. But we’re building on top of all of that in order to bring the autonomy to the table needed for the aircraft to be able to be truly standalone.
It can be collaborative with other assets in the event it needs to be collaborative with other assets, but it’s the world’s first true standalone autonomous fighter jet.
Tyler Rogoway: First, how did you guys go from what you’re doing now to this? What was the thing that said to you, hey, we need to get into a VTOL low-observable (stealthy), fairly high-performance uncrewed airframe, and that’s where we want to take the company as a flagship product?
Armor Harris: Shield AI has really been the preeminent provider of fast-jet autonomy, and that’s really what the world knows us as. But in order to really demonstrate the power of what a fully autonomous system is able to do, we have to bring our own platform to the table that’s really built around and built for the autonomy, and then we’re trying to solve the warfighter’s biggest problems.
That’s really the core tenet of what Shield AI is as a company, is we are extremely mission-focused on trying to solve the most challenging problems in the world today.
One of those was affordable mass of drones and solving a targeting loop with V-BAT. V-BAT is really about solving targeting problems. We have a lot of effectors. We don’t have an ability to target them. But then with X-BAT, it’s about all the challenges that you’re familiar with that not only our own Air Force and Navy are facing today, but also air forces and navies around the world are facing today.
Tyler Rogoway: So, the vertical takeoff, what I saw in your promotional videos was, it goes into reheat [afterburner] and lifts off like a rocket. Is that fully unassisted, there’s no boosters, there’s enough thrust-to-weight ratio in that design with a payload to get that off the cradle and up into forward horizontal flight with just the engine?
Armor Harris: Yep, correct. The fundamental premise of this architecture is that other VTOL systems, like the F-35B, need a complicated like lift fan system, to achieve it. Osprey needs the tiltrotors. This is a straight-up tail sitter, simple F-16 propulsion train, single engine down the middle, afterburning engine, and the afterburner gives you a thrust-to-weight needed to take off.
Many fighter jets today actually have a thrust-to-weight ratio greater than one anyway, in many configurations. And the way that we achieve the VTOL control is with a thrust vectoring nozzle that was originally developed for an F-15 program in the late ‘90s. It was called ACTIVE [Advanced Control Technology for Integrated Vehicles]. There were a bunch of them [experimental thrust vectoring efforts in that timeframe]. There’s an F-16 one, too. This is that three-dimensional nozzle.
Tyler Rogoway: And you’re saying F100/F110-class engine for this?
Armor Harris: Yep.
Tyler Rogoway: Okay, now the hard question, how does it land? So you come back with a fuel load high enough to go into full reheat all the way down, but the aircraft will also be in a very light state? What’s the concept of operations to recover it?
Armor Harris: Yeah. Great question. Are you familiar with the X-13? So, X-13 was an aircraft in early 1950s that demonstrated launch and recovery from an articulating trailer, similar to what we’re talking about here. Recovery is a little bit different. It’s not catching on a wire, it’s engaging in a mechanism, but broadly speaking, it’s very similar.
The challenge with X-13 was twofold. One, it was early 1950s-era jet engine technology, so while it actually did dozens of launches and recoveries, it couldn’t carry a useful payload or [have] useful range. And then, second, very unnatural thing for a pilot to do, landing on their back like that. We solved that with leveraging tail-sitting GNC – guidance, navigation and control – that we developed in V-BAT, and are bringing it over from that. And then, crucially, we use reheat for takeoff, but then when you’re coming back in, you’ve used your fuel, so you now have a thrust-to-weight such that you can land on dry power [without afterburner], which is important for not melting things on the ground when you’re coming back.
Tyler Rogoway: Are you seeing any sort of weapons recovery ability at that gross weight coming back?
Armor Harris: Yeah, you can bring back your internal weapons.
Tyler Rogoway: And you’re achieving that weight via advanced carbon fiber materials?
Armor Harris: So, first, scale. So, F100/F110 engine, we’re working with both Pratt[ & Whitney] and GE [General Electric] on multiple engines for this platform. Two, it’s about a third of the size of an F/A-18 in terms of size scale. So you have a big engine in a small plane. And, so, your thrust-to-weight is a lot greater than what you might think. And then, second, we’re actually not doing anything magic on the design. It’s actually a very conventional design. It’s composite outer surfaces, [but] a lot of metallic internals, bulkhead stringers, etc. We are aiming for speed-to-fleet here. We are not trying to set records on most efficient lightweight structure design ever. So actually, the structural design of the aircraft is very conservative.
Tyler Rogoway: Do you see more margin in the future to gain more recovery capability as you evolve the design to reduce weight?
Armor Harris: Yeah, exactly. So, we’re actually pretty heavy to start, and then, as we go along, you can always pull mass out of it so that we get there.
Tyler Rogoway: Now we’re going to get into the next big one, which is range. 2,000 nautical miles, and that was full range. So, it’d be 1,000-nautical-mile combat radius, roughly. How do you achieve a 1,000-nautical-mile combat radius with an F100 or F110 engine in an aircraft that’s about a third of the size of a Hornet?
Armor Harris: Yeah, great question, and that [range] includes weapons load, as well. So that’s fully missionized. That’s not empty ferry range. The way that we achieve it is the L-over-D [lift-to-drag ratio] of the design is significantly different than a comparable [fighter] like [an] F-16 [or] F-18 type design. This is not a 9G, hyper-maneuverable platform. This is a long-range high L-over-D, also highly LO [low-observable; stealthy] design. Second is that if you put a big engine in a small plane, then the performance envelope for that is actually pretty exceptional. So you cruise pretty high, pretty fast, and because of that, you’re able to go a long way.
Tyler Rogoway: Just back to the landing real quick. Is there an issue with airflow when it’s coming back down? Is there an issue with air volume into the inlet when you’re in that phase, even at high military thrust without the inlet facing away from the airflow?
Armor Harris: That’s actually one of the trickiest things with the design. So the video is kind of idealized in the way that it comes down. The actual trajectory is, it comes in and does a cobra maneuver to go from horizontal to vertical, and then it translates down… And then the inlet is highly engineered to be able to accommodate that transition, and then also work well in vertical orientation, additional to the horizontal orientation. But the trajectory is shaped to ensure you’re never like ‘blanking the inlet’ [where the engine suffocates due to airflow blocked by the airframe], so you always have airflow.
Tyler Rogoway: I know technology has changed when it comes to developing very advanced inlets capable of things we never saw in the past.
Armor Harris: It’s heavy computer modeling, but we’ve actually already been through the wind tunnel with it – I don’t know if we showed any of the graphics there – to validate all the CFD [computational fluid dynamics], and show that it can do the transition from 90-degree AOA [angle of attack] to zero-degree AOA, and back.
Tyler Rogoway: Okay, now LO [low observable; stealth]. So, I’m looking at this, and it’s looking to me like this is like an actual LO platform, not LO features, reduction in certain bands. It looks like this is engineered to be pretty survivable as far as radar goes. And then there’s the infrared (IR) part of it, sticking an afterburning engine in there, nozzle, etc. There’s trade off I’m sure. Can you just talk a little bit about what your vision was for the LO survivability side, both in the radio frequency and IR spectrums?
Armor Harris: I’ll just say it is significantly more survivable than other CCAs that are out there to date.
Tyler Rogoway: And on the IR side, the trade with the nozzle and going with a larger engine like that, have you found ways to mitigate that? Obviously, you can’t get into details on it.
Armor Harris: Yeah, can’t get into details. Obviously, there are trade-offs associated with VTOL, right? So, we have to have a thrust vectoring nozzle on the back in order to do the VTOL control. So, you’re not going to be doing the B-2-style exhaust shaping. But there are ways to be very mission effective, and still protect, for trade-offs.
Tyler Rogoway: And this would be a high-subsonic cruise type platform, I’d imagine. Is that what you’re targeting?
Armor Harris: Yeah, high-subsonic cruise.
Tyler Rogoway: So, will it be able to exceed the current performance of CCA offerings that are in the USAF’s Increment 1?
Armor Harris: Yeah, it’s much faster.
Tyler Rogoway: On just maneuverability, I get it’s not a high-G load aircraft, but when you put thrust factoring engine on, are there any maneuvering advantages that you’re seeing with this design in certain parts of the envelope? What about at high altitudes?
Armor Harris: Yeah, there is, especially at high altitude. So having a thrust vectoring engine makes up for a lot of the inherent challenges associated with tailless design.
Tyler Rogoway: Okay, payload and weapons. I saw a couple of AIM-120 AMRAAMs, even a single AIM-174 as options in the video you just showed us. There’s also smaller weapons that are coming, smaller missiles. Especially with a true LO platform, you have some opportunity there to get closer to the target. Can you give us an idea, a little bit about your vision for weapons and also sensor payloads?
Armor Harris: So, there’s two internal weapons bays and there’s also external weapons hard points. So you can carry the big sticks externally, and you can carry other things internally, but multi-role is a key tenet of the design. So, think air-to-air, surface strike, and air-to-ground weapons all being critical things.
We’re pretty tight-lipped about the mission systems right now. I think what we’ll say there is that if you’re trying to be multi-role you also have to have the sensors needed for it to do multi-role targeting, both active and passive. And that’s a key part of the standalone system, is that, yes, it can work in a network of systems, but if that network is not available for whatever reason, it can also hunt on its own.
Tyler Rogoway: Getting to a larger question. Right now, there’s a stated requirement to keep a man-in-the-loop for kinetic decisions, and some non-kinetic, too, from what I understand. It’s getting to the point where it’s as simple as a yes or no. It’s like a binary thing. Are you guys preparing this and kind of building this for the reality that we might move away from those sorts of demands for communications architecture and command and control of these things, and let them work a little bit more on their own, of course, based on programming by humans? It sounds like this is a much more independent concept than I’ve seen with the USAF’s CCA program. Can you talk about that?
Armor Harris: Let me answer that in a couple of ways. So, one, fundamentally [at] Shield AI, we believe that a human should be on-the-loop for an offensive kill decision, but there are a lot of different scenarios. I think it’s helpful to outline a couple different scenarios.
There are defensive autonomous weapon systems that exist today, like, think the CIWS [Mk 15 Phalanx Close-In Weapon System] system on Navy ships. I think everybody is on board with a scenario in which you’re assigning the aircraft a box of the sky that it’s patrolling. And if anything crosses this box, because if it’s in this box it’s about to be overhead of, say, a carrier strike group or something like that, it’s going to take the shot for anything that comes in that box, right? And has the tech onboard to do IFF [identification, friend or foe] and things like that. So that’s a scenario, it’s pretty defensive, that, I think, is pretty uncontroversial.
You can then give it tasking for offensive scenarios where, hey, you’re going to look for this target in this area, and if you can positively ID it, with two phenomenologies, whether that’s onboard or off-board, then it’s got the authorization to go prosecute that target. There are also scenarios in which this is not the shooter, and this is playing a support role, doing electronic warfare things, or acting as a sensor, so basically providing a menu of options to the end user, according to the rules of engagement, and the doctrine of employment that the end user has.
Tyler Rogoway: I’m thinking about what could you do creatively with this aircraft and its VTOL abilities. I’m thinking about ships, I’m thinking about seeing a carrier out there with all that goes into an aircraft carrier. You could actually give CCA capability and not even put it on the carrier. And that, to me, solves a lot of problems, potentially. What do you see on the naval side of this, and how this can fit into what the Navy and the Marine Corps are trying to achieve, and the challenges they have? I’m thinking F-35s off of an LHA or LHD [amphibious assault ships]. What a force multiplier if could be if they didn’t have to build another F-35 that’s unmanned. So, what are you seeing here? And also what about containerized and other non-traditional naval deployment options?
Armor Harris: So, first off, the carrier is the most valuable real estate that we have in the American arsenal. And as Admiral [Samuel] Paparo [head of U.S. Indo-Pacific Command; INDOPACOM] characterizes it, it’s the most survivable runway that we can have in theater. And you’ve got highly advanced F-35s, hopefully sometime soon, things like F/A-XX coming off of that. So the idea of basically putting CCA-class things on a carrier, on the one hand, is a useful thing, on the other hand, there’s a trade-off there, because there’s only so much real estate on those flight decks.
The video below from Collins Aerospace includes depictions of CCA-type drones launching from a U.S. Navy Ford class aircraft carrier.
So, if you give the Navy the ability to put air power on other ships, that enables them to come off of those other ships and turn those ships into aircraft carriers, as well. And that opens up a whole menu of possibilities. You can actually fit 60 X-BATs on an LHA. And you think about turning the 10 of those [ships] that we have in the inventory into an air wing carrier of 60 strike fighters that can go out and control the sea in 1,000 nautical miles in any direction, that’s a pretty powerful capability.
You think about being able to put these on ships as small as Littoral Combat Ships, which actually have pretty big flight decks and aviation facilities, and the ability to put it on things that small. Things like USVs, unmanned surface vessels, being able to go in deep inside of the enemy’s weapons engagement zone, conduct fully autonomous operations there, pop-up threats that are hard to predict, that sort of thing.
And then, you also don’t have a ton of requirements you’re imposing on the host vessel to actually be able to carry this. So obviously, the Navy has big shipbuilding challenges today. I think this opens up a lot of possibility to take non-traditional hull like cargo ships, and take containers off, put a deck on, and then, all of a sudden, you’re in business carrying an air wing, right? So, that’s a simplification, obviously, but it kind of shows the universe of possibilities that you open up.
Tyler Rogoway: I know it’s your program, but let’s just say you have a great success in this, it proves to be something forces want, and you get orders at scale. So, as this continues, maybe there’s a huge demand, unlike you’ve had for anything you’ve done before. So, who builds it? How do you see that working? Do you see partners coming on board to help you with this? What’s the vision to get it on the ramp at scale? Or I guess, on the trailer at scale?
Armor Harris: Building the – we call them launch and recovery vehicles – at scale is also a challenge, because we’re gearing up to build multiple launch and recovery vehicles for every one aircraft. So, you’re going out, and you’re pre-positioning those all over the theater, wherever you can operate. So, we make those fairly cheap, make a ton of them, have them all over the place. And so that’s a production challenge, in and of itself.
Over the coming months, you’ll see us announcing some partnerships. You know, as a company, we try to work with the best of the industrial base, both domestically and internationally, and partner with a number of other companies. We obviously are gearing up to build a pretty sizable factory to be able to build these at rate, because, really, the value proposition becomes extremely compelling when you have a large number of these flying all over the place in theater. It enables you to scale your airpower at a fraction of the price of large, expensive manned systems. So yes, production and capability is top of mind, and you’ll see a lot on that in the coming months.
Tyler Rogoway: Comms and C2 interface. Can you give us an idea of sort of what you’re looking at for this to solve beyond-line-of-sight control? Especially being as that it’s going to be fairly independent, it can do its own thing, more so than just being ‘tethered’ from the start to a manned platform, and then building in some independence later on. So, what’s your vision of how it would work, if you can just give us a scenario?
Armor Harris: So, there’s some beyond-line-of-sight comms gear on it. There’s also some line-of-sight comms gear on it. But the way you think about it working is that when those are available, you use those as your tasking interface. So you’re telling the aircraft, ‘hey, this is your zone of operations, we want you to do X-Y-Z in this area.’ And it goes out, it does its thing there, and then, over those links, it’s reporting back what it’s seeing, what it’s finding. But when those links are not there, that’s where the autonomy really kicks in, where there’s a mode of operation, where given its last set of instructions and the rules of engagement for what it’s allowed to do, what it’s not allowed to do, it’ll go and it’ll continue its mission autonomously when those comms links are not there.
And that’s really where the system is more advanced than anything else in the world to date.
Tyler Rogoway: So there’s going to be somebody on the ground or in the air, somewhere, that’s going to be watching probably a number of these, I’d imagine? And it’s going to be like a desktop interface, kind of, ‘hey, do this, do that, provide service here,’ and probably some profiles that you can easily and quickly choose for it to do? And then it will send back basically what it needs for the decision maker to make a decision? They give a yes or no, or whatever they need to do to make that critical decision? But if those links are severed, then it will default to a pre-programmed capability or its last setting? And then it can, obviously, I’d imagine, return home by itself and get back? Are you building in some sort of PNT [positioning, navigation, and timing for navigation without GPS] capability with this? What do you see on the PNT side, considering that it might be on its own and out there with nobody watching it for certain parts of its mission?
Armor Harris: Great point on the PNT stuff. That’s a core thing at Shield AI, that we believe that weapon systems need to be able to fully function without the use of GPS, because GPS is always going to be a contested thing. And that’s really been critical to the success that the Shield AI systems have had in Ukraine, because, for example, V-BAT does not rely on GPS, therefore [if its] GPS is jammed, it doesn’t care. This aircraft [X-BAT] is no different, where there’s no one-size-fits-all answer to how you replace GPS. It’s a combination of technologies doing different things and working together in concert. But the aircraft is fully functional without the use of GPS, all the way from launch, to out on the wing doing its mission, to recovery.
Tyler Rogoway: Aerial refueling. Any thoughts on where that stands, or if that’s something maybe you could add in the future?
Armor Harris: So, in the first-generation design, we have a space claim for a probe-and-drogue system. And there’s the interfaces for it and there’s the package for it. We’re probably not installing that right away, but that’s definitely an option for things that can be installed to further extend the reach.
Tyler Rogoway: What about cyber hardening and what can be done there? Because when you’re talking about an asset that’s disconnected and also this autonomous, that’s looking for a connection, what are you thinking for the cyber hardening? And then also the electronic warfare threat. How do you mitigate that?
Armor Harris: I believe that a future conflict is going to be won by who can update the software the fastest. The experience that I’ve personally had with multiple systems in Ukraine, ranging from Starlink to V-BAT, is that, on day one, your system doesn’t work great, no matter how hard you try, because there’s going to be some assumptions that are going to be wrong. And then the question is, how fast can you change those, right? So rapid software update capability is very critical to winning a future fight. If there was, hopefully there isn’t, but if there ever was a peer conflict, I think what a lot of people will find is that despite our best planning, we will have both sides who will have assumptions that were wrong and things that need to be changed. And it’ll be a question of which systems can update faster. So that’s on the one hand.
On the other hand, you need to protect against the ability for malicious actors, at really any stage of the game from development all the way to operation, to be able to have the ability to compromise the system. And it really is a ‘defense in depth’ strategy. And just like operations without GPS, there’s no one-size-fits-all answer to it. You have to have layers of defenses covering all aspects of the platform, the end users, networks, the company that builds it [their] networks, the software update process, etc. And that really gets baked in early in the software and networking architecture for not only the program itself, but also the C2, the customer, networks, etc.
So, the answer is really defense in depth.
Tyler Rogoway: Do you think there’s enough focus put on that overall – industry, DOD, etc – what are your feelings just on that, if you can share?
Armor Harris: Yeah, if I could soapbox for a second, I actually think that the vast majority of the time, the US Department of War is way too focused on making it difficult to change and update the software. And on the one hand, yes, you’re protecting against a category of cyber threats, but on the other hand, if it’s very difficult to software update a system, that system is going to be very challenged when the missiles start flying and you find that you need to make changes. So I’ve been trying to move folks within the various services more in the direction of, hey, we need a rapid software update capability in a system that we want to be relevant.
Tyler Rogoway: This is also coming from your experience in Ukraine, and seeing what that’s like, and the vulnerability of not being able to get in there as fast as possible and generate new software for these systems?
Armor Harris: Yeah, and it’s also a situation where your best defense is also a strong offense, right? If you can update quickly, if there is a problem, you can also update your way out of the problem.
Tyler Rogoway: How would this aircraft be trained with? So, let’s say the DoD says, ‘I want 500 of them,’ and you guys are building them, and it’s working, and that’s all good. But how does, not just the people that are operating it, but also the joint force, get exposure to this and work with it? Some of the nice things about these uncrewed aircraft is you can literally simulate a lot of the training and park the thing and storehouse it, and all that flight cost is sort of out the door. And then you test, and the money is spent on testing and unique training opportunities. But it’s not like you have to fly it every day with a pilot to give them hours, like a crewed aircraft. It’s an autonomous vehicle. So how do you see that working in this new world of a CCA-minded USAF, and other services, as well?
Armor Harris: This is something that we actually do a lot of today. So today, at Shield AI, we do tons of flights with a variety of platforms from various other companies and customers around the world. And there’s many things you can do. So, when you have the manned aircraft in the air, you can have virtual, collaborative aircraft that it’s flying alongside of, because they’re autonomous, anyway. The thing simulated can be on the ground somewhere, and you can be passing that [data] over the network. Secondly, today, because of platform availability, we do a lot with surrogates. So think like flying the General Atomics MQ-20 in place of something that doesn’t exist yet, for example, or a target drone in place of something that doesn’t exist yet. This way you can train the autonomy ahead of the platform itself existing.
With X-BAT, we’re aiming for 10-times lower operating cost than a fifth-gen platform, which means that it’s pretty affordable from a flight hour standpoint. So, you do want to fly it a lot, but for the real-world exercises. But you’re also going to be leveraging the digital twins on the ground, and then fully on the ground, fully virtual. You have the model of the aircraft. You also have models of the manned systems. And you can run the entire force on the ground in a joint simulation environment. And this is something with the Navy in particular, [which] has really been innovating a lot on something that we think is very essential.
Tyler Rogoway: Is this being built to a standard, thousands of hours type of airframe life, or something less, more expendable?
Armor Harris: It’s 10 years. 10 years is the design life on the airframe. It’s a little bit of a balance. Domestic customers, generally, are fine with less. International customers generally want more. But from an engineering standpoint, there’s not really anything you do differently to go from five to 10. So, 10 is kind of the easy answer.
Tyler Rogoway: Cost, always a hard thing to answer, but is there a range that you’re targeting for this?
Armor Harris: Yeah, same price point of other CCAs.
Tyler Rogoway: So, would you put this similar in cost to Increment 1 [a unit cost range of roughly $20 to $30 million]? Because CCA is now a very broad term. Is there a dollar figure range?
Armor Harris: The reason why that’s hard to give is because there’s a little bit of a sliding scale for the mission systems installed on the aircraft. We’re pretty big proponents of the open-source mission systems and autonomy frameworks, and there’s a lot of flexibility in the specific mission systems you’re installing for specific use cases. But we’re very much in the middle of the CCA range. This is not an exquisite capability. This is very much focused on cost per effect.
Tyler Rogoway: Okay, so here’s the big question, state of development, where does it sit? How long has this been going on? And what is the vision for how fast to get it in a condition where you begin flight testing?
Armor Harris: So, it’s been in development for about a year and a half at Shield AI. We’re actually moving towards demonstrator flights doing the vertical takeoff and landing component of the mission profile in the second half of next year. And then all-up flights in ’28, and really prioritizing, again, speed to the joint force.
Tyler Rogoway: Any major hurdles that you see? The biggest hurdles, external, internal?
Armor Harris: You know, no program is real unless it has risks. The philosophy for this project has been identify the risks, which are numerous, early, and then do a lot of testing early and often to burn those risks down in the early stage of development.
So, the way that we’ve run this program, we’ve done a lot more testing already earlier than really anybody else does in aviation. There’s already been a lot of engine ground testing. There’s already been a lot of wind tunnel testing. There’s already been radar cross-section testing with a pole model. The first structural articles are already in build, are already in tests. There’s thermal testing that’s already being conducted. So, the name of the game is really reduce the risk as early as possible.
And then, from a design philosophy, we’re really not doing anything that hasn’t been done before. The beauty of X-BAT is that we’re bringing together a lot of highly mature technologies, because even the VTOL piece has been done before with the X-13. We’re just doing it with modern guidance, navigation, controls, and with a modern engine, right? A lot of the other pieces from the aircraft and the weapon system are in development in other programs, and we’re just bringing it all together. So that’s how you can innovate quickly.
I worked on landing Falcon 9 prior to being the head engineer for Starlink satellites. And the beauty of Falcon 9 is that it’s not actually the world’s most advanced rocket. The engines are gas generator cycle engines, which, if you’re a rocket nerd, is kind of like 1960s technology, but it’s very reliable, very simple, and was pretty quick to develop. It’s kerosene. It’s not like liquid hydrogen. And it’s a very similar approach to X-BAT, where, it’s doing a lot of novel things, but the underlying technology is actually very mature.
Tyler Rogoway: Any thoughts on the changing nature of this whole space, where companies like Shield AI can step in and become a major airframer? It used to be such an exclusive small club. Any thoughts as our parting question here of how you see the landscape changing, and what this opportunity means to bring in newer companies, and to sort of break that model that’s existed since really the Cold War? At least back then, there were a lot more prime contractors, but this is definitely a change, and you guys are a part of it.
Armor Harris: So, it’s really two things. So one is that the team that we’ve assembled over the last year and a half at Shield AI was really made up of a lot of the best folks from around the aerospace industry that have been the key leaders and engineers on the most game-changing things that have been developed and fielded, both unclassified and classified, over the last 10 years. So it really is a truly world-class team. Very excited and humbled to be around [them] every day. And it is hard to do this, no doubt about it. And part of it, though, is that we’ve just brought in the folks who are experts in doing it, to build the team.
But then, secondly, as you know, as computer-aided and assisted design has matured, the number of people that you need to do a program like this has dropped significantly.
I come from a space background, right? So, the Apollo program was, like, 250,000 engineers to build Saturn V, right? But then Falcon 9, we did that with, like, less than 2,000 engineers over the entirety of the program. And the benefit is computer aided design, automated schematics, all the software tools that we have today, [and] not even getting into all the AI-type tools that make software development a heck of a lot faster today than it was even like three years ago. So, you’re able to do these complex projects with a smaller and more affordable team.
And I think it’s also the collective lessons learned of the American aerospace industry, which has really experienced a renaissance in the last 10 to 15 years. Historically, an engineer would see one development effort over their entire career, maybe two. We were kind of in that world, like in the ‘80s, and ‘90s, and early 2000s. But, then, because the number of projects has really sped up over the last 15 years, now you have folks who in the last 15 years have done three or four whiteboard-to-flight type iteration cycles. And I like to say that engineering is a lot like any sport, it all comes down to reps. If you want to be a great basketball player, you better be pretty good at shooting hoops. How do you do that? You just got to shoot a lot of free throws. And engineering is no different. If you want to be good at it, you got to just get a lot of experience doing it. I think the aerospace industry has started to produce that experience in the last 15 years, and it’s created a kind of a class or a generation of very talented folks who are able to go make things like this happen in just a couple of years. So, I think it’s an exciting place to be as a country and as a world.
Tyler Rogoway: And you think this is all driving a lot of the ability for newer companies, non-traditional ones, that come in and compete with juggernauts?
Armor Harris: I think it’s an exciting time to be in the industry, and it’s an exciting time for America, and it’s really, I think, the best thing that we have going for us as a country right now.
Editor’s note: A big thanks again to Shield AI’s Armor Harris for walking us through the ins and outs of X-BAT.
Contact the author: Tyler@twz.com
