Best of LinkedIn: Defense Tech CW 04/ 05

00:00:00: provided by Thomas Allgeier and Frenness.

00:00:02: based on the most relevant posts on LinkedIn about defense tech and CWO four and O five.

00:00:07: Frenes is a B to B market research company that eclips product and strategy teams with market and competitive intelligence across the defense industry.

00:00:15: So it's the start of February twenty twenty six.

00:00:17: And looking at the landscape right now it feels like the defense tech industry is kind of holding its breath.

00:00:24: Yeah.

00:00:24: We aren't just seeing new gadgets anymore.

00:00:26: We're seeing the foundational doctrines of warfare well shift in real

00:00:30: time.

00:00:30: It definitely feels.

00:00:31: You know, we've moved past the look at this cool drone phase, right?

00:00:35: And we're straight into the how do we actually build a military around this stuff phase, right?

00:00:39: The novelty is wearing off and the let's say reality of deployment is setting in

00:00:44: exactly.

00:00:45: and Looking at the stack of sources we have for this week, calendar weeks four and five, a mission for this deep dive is pre-specific.

00:00:51: We need to parse the signal from the noise.

00:00:54: There's just so much out there, flashy demos, press releases, startup hype, but you know, you and I, we represent the listener here.

00:01:02: We need to figure out what is actually deployment ready.

00:01:05: We

00:01:05: need to find the capabilities that survive contact with reality.

00:01:09: Looking at the research, three big themes stood out to me.

00:01:12: First, we have the evolution of drones and counter UAS towards like true robot ecosystems.

00:01:19: that's getting messy and crowded on the ground.

00:01:21: And then I guess the software layer, AI and autonomy are finally penetrating high-end warfare.

00:01:26: And we're talking ASW and air combat, not just chatbots.

00:01:29: And finally, the strategic layer, the massive shift towards data-driven warfare and, well, the industrial funding boom that hit in twenty twenty five.

00:01:38: It turns out writing the code is the easy part.

00:01:41: Building the factory and finding the talent, that's the hard part.

00:01:44: That's it.

00:01:44: So let's get right into the mud.

00:01:45: Theme one, drones and counter UAS, the kinetic frontline.

00:01:50: I want to start with a piece of intel that sounds like science fiction, but it's just gritty tactical reality from Ukraine.

00:01:57: You're looking at the robot logistics insight.

00:01:58: Yeah.

00:01:59: From Tim DeZitter.

00:02:00: That's the one.

00:02:01: He highlighted this doctrinal signal from the Pokrovsk front.

00:02:06: Here's the scenario.

00:02:07: A ground robot, a UGV malfunctions during a mission.

00:02:12: Now, standard procedure for the last hundred years says you send a couple of soldiers to go get the gear.

00:02:17: Which exposes them to artillery.

00:02:19: Snipers, you know those ubiquitous FPV drones.

00:02:22: Right, so instead the unit dispatched a second robot to go retrieve the broken one, a robot towing a robot.

00:02:28: And

00:02:28: what stuck out to me in Desider's analysis wasn't just that they did it, it was how long it took.

00:02:32: Fifteen

00:02:33: hours.

00:02:33: Yeah, a fifteen hour operation.

00:02:35: Which,

00:02:35: I mean, that sounds incredibly inefficient to me.

00:02:38: If I'm a commander, do I really want to tie up resources for fifteen hours just to drag a busted chassis back to base?

00:02:44: Wouldn't it be faster to just send a team?

00:02:47: Faster.

00:02:47: Yes.

00:02:48: But safer.

00:02:49: Absolutely not.

00:02:50: And that's the critical takeaway here.

00:02:51: This is about force preservation.

00:02:54: Fifteen hours for a robot is cheap compared to fifteen minutes of exposure for a human squad.

00:02:59: These interpoints out that we're moving from single gadgets to a ground robotics ecosystem.

00:03:05: It's unmanned sustainment.

00:03:07: So the robot isn't just the spear tip.

00:03:09: It's the ambulance in the tow truck too.

00:03:11: Exactly.

00:03:12: And once you start building an ecosystem like that, you run into engineering problems.

00:03:16: I mean, if we have robots rescuing robots, these machines are getting complex.

00:03:20: For sure.

00:03:21: And we had a really sharp critique this week from Scott Davis about how we're building them.

00:03:25: I

00:03:25: saw that.

00:03:26: He's going after the industries.

00:03:28: obsession with software.

00:03:30: His argument is, we're over-engineering code to fix, well, bad hardware design.

00:03:36: It's a classic trap.

00:03:37: Take the drone dilemma.

00:03:38: You want to fly high to cover a large area, but you need to fly low to see details or drop a payload.

00:03:44: Right, and doing both just eats up battery.

00:03:46: It eats up battery and requires incredibly complex flight logic, complex autonomy.

00:03:52: David says, stop trying to code your way out of physics.

00:03:54: So what's the alternative then?

00:03:56: Build better hardware.

00:03:58: He points to the hybrid hopper concept.

00:04:00: It flies to get to the location, but then it lands on a roof or a hill to observe.

00:04:04: Uh, so you're grounding the sensor.

00:04:06: Exactly.

00:04:07: By landing, you stop fighting gravity.

00:04:09: You stop draining the battery on hovering.

00:04:11: And suddenly, your mission logic gets simple.

00:04:14: You don't need an AI supercomputer to manage flight stability if the thing is just sitting on a brick wall.

00:04:20: It's a good reality check.

00:04:21: Sometimes mechanical innovation beats algorithmic complexity.

00:04:25: But let's flip the script.

00:04:27: If the enemy is using these hoppers and rescue bots, how do we stop them?

00:04:31: The counter UAS, the COS market is just exploding right now.

00:04:35: And we saw two completely opposing philosophies this week on how to handle that problem.

00:04:40: Okay,

00:04:40: so on one side we have the brute force method.

00:04:44: Philippe Albino shared details on Rhinemiddles new approach.

00:04:47: And subtle is not the word I would use.

00:04:49: No,

00:04:49: kinetic solution to a kinetic problem is the phrasing, I think.

00:04:52: They put two miniguns on a ground robot.

00:04:55: We are talking about firing three thousand rounds per minute.

00:04:58: It sounds like something from an action movie, but Albina argues the math actually works.

00:05:02: And it comes down to the economics of the intercept.

00:05:05: That is the single biggest problem in air defense right now.

00:05:07: Right, because if you shoot a hundred thousand dollar missile at a twenty thousand dollar Shaheed drone, you might win the engagement, but you lose the war.

00:05:15: You go bankrupt.

00:05:16: Precisely.

00:05:17: But if you spray it with, what, two hundred dollars worth of bullets, you win.

00:05:21: Abino describes it as creating a cone of projectiles.

00:05:25: A wall of lead.

00:05:26: A wall of lead, yeah.

00:05:27: You aren't sniping the drone, you're creating a barrier that the drone has to fly through.

00:05:33: Highly effective for swarms.

00:05:34: But you know, bullets fall down and they don't discriminate.

00:05:38: Is brute force the only option?

00:05:39: No, and that's where Raven Decker and Scott Fowler come in.

00:05:42: They highlighted a partnership between Drone Shield and Nutellic, who used to be Avalor AI.

00:05:47: This is the opposite of brute force.

00:05:48: This is about modularity.

00:05:49: This

00:05:50: is the software brain for the counter UAS system.

00:05:53: Think of it as an operating system for defense.

00:05:55: Their argument is that threats evolve too fast for any single hardware solution.

00:06:00: Today, it's a quadcopter.

00:06:01: Tomorrow, it's a jet-powered drone.

00:06:03: You need a system that is effector agnostic.

00:06:06: OK, just to clarify for everyone, effector being the thing that actually stops the drone.

00:06:10: so jamming lasers, nets, bullets.

00:06:13: Correct.

00:06:14: The drone shield and intelligent approach is to build an open architecture.

00:06:17: You might need jamming today, a laser tomorrow, and that Rheinmetall minigun next week.

00:06:23: The software stays the same, the effector just swaps out.

00:06:26: That feels much more future-proof.

00:06:28: And speaking of future-proof, we have to talk about the high-end fight.

00:06:31: We aren't just fighting cheap quadcopters.

00:06:34: Catherine Knee from Lockheed Martin shared something about the JGM missile that caught my

00:06:38: eye.

00:06:38: Yeah, this is significant because it bridges the gap.

00:06:41: The JGM is a serious piece of hardware.

00:06:44: and Lockheed demonstrated a true vertical launch capability with it against a UAS threat.

00:06:50: Why

00:06:50: does vertical launch matter specifically here?

00:06:52: It's all about geometry.

00:06:54: If you are on a ship or a truck in an urban canyon, you can't always point a launcher at the horizon.

00:07:00: Vertical launch means you can shoot straight up.

00:07:01: turn and engage a threat coming from any angle.

00:07:04: It's about flexibility in tight spaces.

00:07:06: Exactly.

00:07:07: And adding to that high-end mix, Jack Chung noted that Lockheed's Helios laser system was praised for neutralizing drones at sea.

00:07:14: So lasers are finally moving from PowerPoint to deck of a ship.

00:07:18: They are.

00:07:19: And again, it comes back to that economic equation.

00:07:21: If a laser, which costs pennies per shot in electricity, can fry a drone, you save your multi-million dollar SM-II missiles for the anti-ship cruise missiles.

00:07:32: His

00:07:33: resource management.

00:07:34: You don't use a sledgehammer to kill a fly, especially when you need that sledgehammer for the tank coming up behind it.

00:07:39: Precisely.

00:07:40: But before we leave the hardware world, there is one factor that no amount of software or funding can fix.

00:07:46: The weather.

00:07:46: The weather.

00:07:47: The Arctic test.

00:07:48: Mark C.L.

00:07:49: and Adam Painter raise a really cold point about testing in Latvia.

00:07:53: minus fifteen degrees Celsius.

00:07:55: That is where the rubber meets the frozen road.

00:07:57: It's easy to look good in a desert test range.

00:08:00: But Adam Painter's takeaway was that when technology meets extreme cold, simplicity is what saves lives.

00:08:06: Totally.

00:08:07: Batteries fail faster, sensors ice over, touch screens don't work with thick gloves.

00:08:12: Painter's point is that arctic proof on a spec sheet is often, well, a lie.

00:08:17: If your system is fragile, it's useless on the Eastern flank in February.

00:08:20: It's a stark reminder that geography still dictates technology.

00:08:23: Okay, so that's the hardware reality, frozen batteries, miniguns, and rescue robots.

00:08:29: Let's shift gears to theme two, AI and autonomy.

00:08:34: This is what they're calling software-defined defense.

00:08:37: And interestingly, the most advanced stuff we saw this week isn't in the sky, it's underwater.

00:08:42: Yes, the anti-submarine warfare domain.

00:08:45: or ASW.

00:08:46: Jennifer McArdle shared some insights on Helsing's Lura AI.

00:08:50: Now for the listener who isn't a naval expert, ASW is notoriously difficult.

00:08:55: You're basically listening to a noisy ocean and trying to hear a specific metal tube moving miles away.

00:09:00: It is the ultimate signal to noise problem.

00:09:02: Whales, waves, shipping traffic, it's all just noise.

00:09:05: What Helsing is doing is deploying this AI On gliders, these are low-power, long-endurance underwater drones.

00:09:11: But

00:09:11: here's the key part.

00:09:12: The AI processes the data at the edge, right, on the glider itself.

00:09:16: That is the game changer.

00:09:18: Usually you record sound and send it back for analysis, but sending data underwater requires loud sonar pings or surfacing to use radio, which immediately reveals your position.

00:09:27: So edge processing means the glider listens, the AI identifies the SUG locally, and it stays silent.

00:09:32: Exactly.

00:09:33: It only transmits when it has a hit.

00:09:35: It allows for a passive, silent dragnet.

00:09:38: It turns the ocean transparent for you while you stay invisible to the enemy.

00:09:42: That's a huge tactical leap.

00:09:43: That is incredibly stealthy.

00:09:45: Now taking it from the sea to the sky, Harold Copey shared a collaboration between GROB aircraft and Helsing, and I love the tagline here.

00:09:53: The aircraft is the platform, the software is the weapon.

00:09:56: This is the K-One Europa concept.

00:09:59: They took a standard trainer aircraft, the G-One-Twenty-TP, prop plane and turned it into a test bed for autonomous combat.

00:10:06: This isn't just a drone pilot on the ground with a joystick, though.

00:10:09: Oh,

00:10:09: far from it.

00:10:10: This is about an aircraft that understands intent.

00:10:12: It collaborates with manned aircraft.

00:10:14: It makes decisions when communication is cut.

00:10:16: It's a shift towards the aircraft being a computing node that just happens to fly.

00:10:20: But the speed of decision making, it's getting a little scary, isn't it?

00:10:24: Ricardo Grande brought up the UK's baby raptor interceptor.

00:10:28: He did.

00:10:29: And he raised the philosophical and tactical question.

00:10:32: If AI interceptors react in milliseconds faster than humanly possible, what is the human's role?

00:10:39: If the machine observes, Oriens decides and acts before I can even blink, am I really in the loop?

00:10:45: Or am I just liability?

00:10:47: That is the tension we are going to face for the next decade.

00:10:50: And Torsten Stov's report on Endoral and D-Wave adds another layer of speed to this.

00:10:55: They are bringing quantum computing into the mix.

00:10:58: Quantum computing and air defense.

00:11:00: I thought quantum was still years away from practical use.

00:11:03: For encryption breaking, maybe.

00:11:05: But for optimization, it's here.

00:11:06: They use quantum annealing to figure out the best placement for air defense systems.

00:11:11: Explain that.

00:11:11: Why do you need a quantum computer to tell you where to put a missile launcher?

00:11:15: Because it's a math problem with infinite variables.

00:11:18: Terrain, incoming threat vectors, coverage overlap, battery ranges.

00:11:22: A classical computer takes hours to crunch that.

00:11:25: Stab reports that the quantum approach achieved a ten-x speedup.

00:11:28: But did it actually improve the defense, or did it just do the math faster?

00:11:32: Both.

00:11:33: In simulations with five hundred incoming threats, a massive saturation attack, the quantum optimized placement improved intercept rates by nine to twelve percent.

00:11:42: Twelve percent.

00:11:43: In a saturation attack that could involve nuclear or biological warheads, stopping twelve percent more missiles isn't a statistic.

00:11:50: That's saving a city.

00:11:51: Exactly.

00:11:52: When you are dealing with mass, optimization isn't just math.

00:11:56: It's survival.

00:11:57: And wrapping up the AI theme, we have Agentec AI.

00:12:01: Mario Delab gave us an update on DARPA's programs.

00:12:04: The AIR program.

00:12:05: They are training AI pilots for beyond visual range combat.

00:12:09: But I think the more critical one is AIECC autonomous cyber defense.

00:12:13: Why is that one more critical?

00:12:15: Because cyber moves at light speed.

00:12:17: If the enemy uses AI to hack your command and control, a human cyber defender is just going to be too slow to patch the whole.

00:12:24: You need AI to fight AI.

00:12:26: It's a machine-on-machine fight in the digital spectrum.

00:12:28: It is.

00:12:29: Which leads us perfectly into our third theme, strategy, industry, and command and control.

00:12:34: Because if the warfare is changing this fast machines fighting machines at light speed, the industry and the decision making structures have to change too.

00:12:42: Right.

00:12:42: And we're seeing a massive split in how different regions are handling this.

00:12:46: For sure.

00:12:46: Let's talk about data driven warfare.

00:12:49: Artem Rose shared some stats from the Army of Drones, twenty twenty five in Ukraine that are just staggering.

00:12:55: The numbers are intense.

00:12:57: Roughly eight hundred twenty thousand enemy targets hit.

00:13:00: Huge vehicle losses.

00:13:01: But the real story isn't the destruction.

00:13:04: It's the accountability.

00:13:05: What do you mean by accountability?

00:13:07: Morose points out that decisions in Ukraine are now based on verified video data, not intuition or field reports.

00:13:14: You can track performance, you know exactly which unit is effective, which pilot is missing, and which drone model is failing.

00:13:20: That sounds corporate, like KPIs for combat.

00:13:23: It's efficiency applied to lethality.

00:13:26: And Jonas Singer contrasts this sharply with Europe.

00:13:29: He says Ukraine has undergone a live-fire organizational revolution.

00:13:32: They have decentralized procurement.

00:13:34: They have the brave one platform to speed up innovation.

00:13:37: And Europe.

00:13:37: Singer argues Europe is still stuck in slow processes.

00:13:40: He calls for a live-fire organizational revolution in the West.

00:13:43: A drone PowerPoint versus a revolution.

00:13:47: Ouch.

00:13:47: It's a harsh critique, but his point is valid.

00:13:49: Innovation can't be a committee meeting when the adversary is iterating their software on a weekly basis.

00:13:54: Well,

00:13:54: the private sector is certainly trying to catch up.

00:13:57: Matt Atkins dropped a market report that shows, twenty twenty five was a breakout year for defense

00:14:02: tech funding.

00:14:02: It was a massive year.

00:14:04: Fifty billion dollars in VC funding for defense tech.

00:14:07: Fifty billion.

00:14:07: Yeah,

00:14:08: that is double the total from twenty twenty four.

00:14:10: And we saw ten new unicorns born.

00:14:13: So the cash is there, but I have to play devil's advocate here.

00:14:17: Is this a bubble?

00:14:18: Are we just throwing money at cool keywords?

00:14:21: That's the risk.

00:14:22: But Atkins makes a crucial distinction.

00:14:24: Twenty twenty five was the year of fundraising.

00:14:27: Twenty twenty six is the year of execution.

00:14:29: The check is cleared.

00:14:30: Now you have to build the factory.

00:14:32: And that brings us to the talent war.

00:14:34: Exactly.

00:14:35: You have fifty billion dollars.

00:14:36: But can you find the engineers who can get top level security clearances and know how to build autonomous guidance systems?

00:14:43: That talent pool is shallow.

00:14:44: That is going to be the bottleneck.

00:14:46: Speaking of bottlenecks and supply chains, Britta Jacobs shared some insights from a delegation trip to Taiwan.

00:14:52: This connects back to that idea of resilience.

00:14:54: She noted a focus on whole of society resilience.

00:14:59: It's not just about the military, it's about the civil infrastructure being ready for asymmetric warfare.

00:15:04: And Christoph Reithke connected this to the European startup scene.

00:15:08: He mentioned that dual use is becoming essential.

00:15:11: That's a commercial reality.

00:15:12: If you are a defense startup, relying solely on government contracts is risky and slow.

00:15:18: If your tech has a commercial application, dual use.

00:15:21: You can sell it to civilian markets to keep the lights on while you wait for the Ministry of Defense to sign the contract.

00:15:27: So if you make a drone battery, sell it to delivery companies and the army.

00:15:31: Exactly.

00:15:31: It's becoming a requirement for viability.

00:15:33: And finally, on the software side of things, we have a point about user experience.

00:15:38: Mikhael and Rogandino from the U.S.

00:15:39: Space Force and Dr.

00:15:41: Sinan Alaw Baidi from Barzan Holdings both touched on this.

00:15:45: This is the user-centered design argument.

00:15:47: For a long time, military software was terrible.

00:15:50: Clunky, hard to use, looked like Windows' ninety-five on a bad day.

00:15:54: I've seen some of those interfaces.

00:15:55: They look like you need a Ph.D.

00:15:56: just to open a menu.

00:15:57: Right.

00:15:58: But the inside here... is that bad design isn't just annoying, it's dangerous.

00:16:03: If the software isn't intuitive, if it doesn't work.

00:16:05: like the apps these twenty year old soldiers use on their phones, it fails the warfighter.

00:16:10: Cognitive load.

00:16:11: Yes.

00:16:12: In a high-stress situation, if you have to fight the user interface to send a command, you make errors.

00:16:17: Friendly fire, missed targets.

00:16:19: So making C to command and control systems intuitive is about operational effectiveness, not just aesthetics.

00:16:26: Okay, let's unpack this.

00:16:27: We've covered robot ecosystems that tow each other to safety, AI speags that are beating human reaction times, and a massive industrial pivot toward data and usability.

00:16:38: It's a lot to digest.

00:16:39: It is.

00:16:40: But if there's one thought to leave you with, I think Eva Zula summed it up perfectly.

00:16:44: I have that quote right here.

00:16:45: She warned that autonomy isn't a drone race, it's a capability race.

00:16:49: Precisely.

00:16:50: The winners in twenty-twenty-six won't be the ones with the flashiest demo video on LinkedIn.

00:16:54: The winners will be the ones whose architectures are resilient enough to survive electronic warfare.

00:16:59: It's about data pipelines, C-II integration, and systems that don't crash when the jamming starts.

00:17:04: Resilience over hype.

00:17:06: That is the boring, unsexy work that actually wins wars.

00:17:10: It is.

00:17:10: And that's where the industry needs to focus.

00:17:12: If you enjoyed this episode, new episodes drop every two weeks.

00:17:15: Also, check out our other editions on ICT and tech insights, health tech, cloud, digital products and services, artificial intelligence and sustainability in green ICT.

00:17:25: Thanks for listening.

00:17:26: Stay curious and keep looking for the signal and the noise.

00:17:29: Don't forget to subscribe.

00:17:30: See you in the next deep dive.