If you were worried that the United States military hadn’t picked up on the importance of drone warfare in the Russo-Ukrainian War, it appears that someone in the Pentagon was indeed paying attention.
The Pentagon committed on Monday to fielding thousands of attritable, autonomous systems across multiple domains within the next two years as part of a new initiative to better compete with China.
The program, dubbed Replicator, was announced by Deputy Defense Secretary Kathleen Hicks, speaking at the National Defense Industrial Association’s Emerging Technologies conference here.
“Replicator will galvanize progress in the too-slow shift of U.S. military innovation to leverage platforms that are small, smart, cheap and many,” Hicks said.
Hicks and Vice Chairman of the Joint Chiefs of Staff Adm. Christopher Grady will oversee the program, with support from Doug Beck, director of the Defense Innovation Unit. Further details, Hicks said, will be released in the coming weeks.
Replicator rests on two assumptions. The first is that China’s core advantage is mass — “more ships, more missiles, more people,” as Hicks said — and that the United States’ best response is to innovate, rather than match that pound for pound.
The second is that attritable, autonomous systems are the right form of innovation. Hicks pointed to the war in Ukraine, in which cheap, often commercial drones have proven indispensable on the battlefield for reconaissance, targeting, and attacks. Russia too, she said, appeared to have a similar mass before launching its invasion last February.
However, this program is squarely focused on China. Hicks called this moment a “generational challenge to American society.”
”We’ll counter the [People’s Liberation Army’s] mass with mass of our own, but ours will be harder to plan for, harder to hit, and harder to beat,” she said.
Even so, Hicks noted the Pentagon will remain focused on its core systems. “America still benefits from platforms that are large, exquisite, expensive, and few,” she said. Instead, she said, Replicator is particularly focused on accelerating DoD’s recent investments in autonomous systems.
Replicator’s goal of fielding small drones in high numbers and on a rapid timeline echoes calls from former DIU director Mike Brown for the Pentagon to better leverage commercial innovation to deliver capability at scale — an approach he called a “hedge strategy.”
House appropriators have backed that idea in their fiscal 2025 defense spending bill. The legislation would allocate $1 billion toward establishing a DIU-managed hedge portfolio made up of low-cost drones, agile communication and computing modes and AI capabilities.
The Department of Defense requested $1.8 billion for artificial intelligence for fiscal 2024 and was overseeing more than 685 related projects as of 2021. Replicator is intended to pull those investments together and further scale production, Hicks said.
Insert your own hedge funds and Skynet jokes here.
The strategy makes a good deal of sense…up to a point. The fast and cheap portion makes a lot of sense, given Ukraine’s use of dirt cheap flatpack cardboard drones we talked about earlier this week.
It’s the out of control/autonomous portion of description, combined with the aggressive timeline, that I question. As far as I can tell, all of Ukraine’s drones have been human guided rather than autonomous.
Lots of work on AI has been done over the last few years, and its entirely possible that AI drone tech is farther along than we know, but having been involved in numerous large software projects for multiple companies, I can tell you things always seem to take longer than they should even when the federal government isn’t involved. Long term, having autonomous or semi-autonomous drone will give you a lot of extra capabilities, but I’m very skeptical about that two year timeline.
Also, unless we plan to launch those drones from Taiwan itself, I’m skeptical that we’ll have suitable naval launch platforms ready. Flying a few drones off the deck of destroyer is easy, flying thousands for a real drone swarm is probably impossible. You don’t want to try running drone and manned planes off fleet carriers at the same time.
Can you run them off an amphibious assault ship? Probably, as a temporary expedient, but that’s going to limit your helicopter and F-35B takeoff and landing windows. Longer term, you’re probably going to need to construct ships designed with specialized launchers to send a whole lot of drones in a short space of time.
I’ve been talking about the inevitability of drone swarms in combat for some time. The goal is entirely feasible, I just question the “two years to fight China” timeline.
I sure hope the Pentagon powers that be have a manned drone swarm program backup on hand…
Tags: AI, China, drone, Military, Russo-Ukrainian War, swarming, Taiwan, technology
Question from a non-military person:
Could swarms of drones be launched from drone-ships or drone-subs, designed to be very stealthy and comparatively cheap? Why couldn’t we build a bunch of small, unmanned ships or subs that would be launched from either Taiwan or the Philippines or even Japan, if not from a U.S. ship? They get close to whatever the target is, and open up and let their flocks fly.
You could do it that way, but that’s at least an order of complexity higher, as any additional drone-launching platform is going to require it’s own complex set of software. Also, stealth for large craft tends not to be cheap.
Plus the timeline for designing, developing and building these new drone-launching craft is going to be little to not at all sooner shorter than designing a manned drone-launching ship.
Great platform for domestic control, too.
“Hicks pointed to the war in Ukraine, in which cheap, often commercial drones have proven indispensable on the battlefield for reconaissance, targeting, and attacks. Russia too, she said.”
What Hicks pointed out misses the point, to wit: “[U]nless we plan to launch those drones from Taiwan itself, I’m skeptical that we’ll have suitable naval launch platforms ready.“
The logistical situation is not at all similar. Given the asymmetry between many cheap Chinese drones and a blue water naval craft that can bring US drones to the South China Sea, the advantage goes to the defender, i.e., China.
Unless long-ranger drones could reach the Taiwan Straits from the Philippines, the drone fleet would have to be based in Taiwan. I think this would provoke a first strike.
We need a relatively dumb drone-bus in the JASSM form factor so six can be packed into each Rapid Dragon cargo-drop launcher. A flight of C-17s drops its Rapid Dragons, six each JASSM flying drone pez dispensers deploy and fly towards the target, and then a thousand angry drones dispense to conduct mayhem.
Maybe cardboard JASSMs. Cardboard is the new stealth, after all.
The AI issue is basically related to the swarm tactics: The drones pretty much have to coordinate, and have to be in aggregate fairly clever to avoid each other while effectively swarming the target. No human operator can do that for them. There’s been some interesting research on swarming behavior using fairly low level computing, so maybe the swarming can be autonomous but the target gets a human touch.
But that’s why there’s money for what they are calling “AI” in this.
I can’t help but work how drone swarms will deal with weather conditions in western Pacific. Halsey lost ships to typhoons, air ops ceased, and at least one carrier ended up back in the shipyard due to damage.
As for using unmanned lauching ships, they’ll never go for it. Gotta have command berths or 0-6’s can’t be promoted to flag.
“A flight of C-17s drops its Rapid Dragons,..”
This would seem preferable to ship-based platforms, especially if the aircraft has stealth capabilities. One downside is that a state of readiness would require lengthy holding patterns from a stand-off position. Alternatively, a ground fleet in the Philippines would offer a rapid strike potential and surface-to-air missiles would safeguard against sneak attacks.
The Subic Bay debacle makes me wonder if we can rely on the Philippines to stay the course, though.
“two years to fight China”
“two weeks to flatten the curve”
Repurposed MIRVs on SRBMs, ICBMs or orbital drop?
I’m just impressed someone in the Biden admin is actually talking about China as a threat. That’s a big improvement from that partner nonsense.
“’m just impressed someone in the Biden admin is actually talking about China as a threat.”
Absolutely true, but there’s no excellent idea that cannot be bastardized into uselessness by bureaucracy.
I’m worried about why they think they need a two year program for a war with China.
I would think that Xi would have taken pause from seeing how the Russia-Ukraine war is playing out. Russia’s weapons are worse than expected, Russia is not going to be much help if China gets into a war, especially in the near-term; in fact, Russia’s internal problems may be worse than anyone realizes. And China should see that the U.S. isn’t sending its best stuff to Ukraine, and the 2nd best stuff is beating Russia’s stuff.
Not to mention China not having much of any oil of its own, and (from what I’ve read) it’s not easy for Russia’s oil to get to China, because the pipelines mostly go elsewhere.
But all that assumes a coherent and reasonably accurate flow of information to the top. Maybe Xi is being told all sorts of nonsense about Chinese capabilities.
Everyone managing a project with a computing component needs this reminder often, and usually more often than you’d expect.
Anyone who is familiar with Captain Basil Hart’s theory of the “indirect approach” will be fascinated by the implications of drone warfare.
https://johnrobb.substack.com/p/replicator
Drones, in and of themselves, will not prove to be decisive. There must still be Boots on the Ground.
Anyone who is familiar with Captain Basil Hart’s theory of the “indirect approach” will be fascinated by the implications of drone warfare.
https://johnrobb.substack.com/p/replicator
Drones, in and of themselves, will not prove to be decisive. There must still be Boots on the Ground.
I apologize for wasting band width without the double post.
“Great platform for domestic control, too.”
If you read the John Rob link, he argues that AI drones will end participatory governance. In short, your AR-15 will not protect you against a despotic government possessed of next-generation drones.
[…] the month of August, and Confused in Tijuana about The Berlin Wall and more stories BattleSwarm: Fast, Cheap and Out of Control, Israel’s War Against Color TV, and The Cardboard Drone Dilemma Behind The Black: India […]