The following story is the next installment of a series on micronaval warfare. Read Part One of Admiral Lacy’s oral history on the emergence of micronaval warfare, “The Battle of Locust Point.”
By David R. Strachan
The following classified interview is being conducted per the joint NHHC/USNI Oral History Project on Autonomous Warfare.
Admiral Jeremy B. Lacy, USN (Ret.)
November 19, 2033
Interviewer: Lt. Cmdr. Hailey J. Dowd, USN
We are joined again today by Admiral Jeremy B. Lacy, widely considered the father of autonomous undersea conflict, or what has come to be known as micronaval warfare. Admiral Lacy spearheaded the Atom-class microsubmarine program, eventually going on to establish Strikepod Group 1 (COMPODGRU 1), and serving as Commander, Strikepod Forces, Atlantic (COMPODLANT). He is currently the Corbin A. McNeill Endowed Chair in Naval Engineering at the U.S. Naval Academy.
This is the second installment of a planned eight-part classified oral history focusing on Admiral Lacy’s distinguished naval career, and his profound impact on modern naval warfare. In Part I, we learned about the genesis of the Atom-class microsubmarine and its operational construct, the Strikepod, and the series of events leading up to the first combat engagement of the micronaval era, the Battle of Locust Point, where a Strikepod of prototype Atoms tracked and engaged several Russian Istina-class microsubmarines prowling the upper Chesapeake Bay during Baltimore Fleet Week, 2016. Today our discussion will focus on the aftermath of Locust Point, the continued development of the Atom-class microsubmarine, as well as the early days of autonomous undersea conflict, including the first major confrontation of the micronaval era, the Nanxun Jiao Crisis.
The period following Locust Point was a time of high tension and uncertainty. With the United States now facing an unprecedented undersea threat to both the homeland and forces abroad, the president, through a series of classified executive orders, mobilized a full range of U.S. defensive capabilities. This included authorizing a broad expansion of the Atom-class microsubmarine program via the Joint Undersea Initiatives Group (known colloquially as FathomWorks), a consortium of leading defense contractors and specialized units of the United States Navy, Coast Guard, and Marine Corps charged with, among other things, developing a coastal undersea network to defend against foreign micronaval threats.
We joined Admiral Lacy again at his home in Annapolis, Maryland.
Let’s begin today with FathomWorks. Paint the picture, if you will, of those early weeks in the wake of Locust Point.
It was all about testing, evaluation, and improvement. The Atom had performed admirably, but for all intents and purposes, it was still a prototype, and there were serious gaps in coms and navigation that needed addressing, as well as kinks with Falken [the Atom’s artificial intelligence]. Our vision for the Atom-class was achievable, but in order to get there we had to enhance Falken dramatically. Strikepods would be operating – and fighting – in a complex, communication-denied environment for extended periods. This demanded a highly advanced form of artificial intelligence, as well as a more evolved approach to human-machine teaming.
We emerged from Locust Point feeling pretty validated, and confident that Strikepods could be more than just another club in the bag, that they promised much more than minehunting or intel gathering. We knew that this was a capable platform in and of itself that could be integrated into the fleet at the operational level.
But there was no time to dream and tinker. Our mandate was clear, and the sense of urgency was unmistakable. The Russians were on the move, and the threat they posed to our security was greater than at any time since the Cold War. Our analysts had taken one on the chin with Locust Point, but they’d been warning for months that Moscow was working on something highly advanced and very closely guarded, but there was no indication that anything was operational, much less poised to invade our inland waters. A lot of folks at the Pentagon laughed it off. But they weren’t laughing now, especially when we took a closer look at Poseidon.
So Locust Point influenced the Navy’s perception of the Poseidon program?
When we first heard of it, it seemed more implausible than Istina. That the Russians had the know-how or the resources to develop an autonomous microsubmarine was difficult enough to believe, but when we started receiving information on a nuclear-tipped autonomous torpedo, we were honestly beginning to wonder if we were the target of a wildly excessive disinformation campaign. Even when we received intelligence that it had been test fired from the Sarov, we just didn’t see it as a viable platform. But the Istina changed all that.
In what way?
If the Russians were capable of the miniaturization and AI integration – and audacity – we’d experienced with the Istinas, then we had to assume that Poseidon would eventually threaten our shores as well. We needed a viable micronaval defense.
And so the Atlantic Undersea Network was born?
Yes. The Atom was incredibly flexible by design, and could be configured to fulfill a multitude of roles. We’d been tossing around concepts for mine warfare, including a variant optimized for the seabed, something akin to a mobile, local area SOSUS, with the capability to detect and engage both surface and subsurface targets. The idea, conceptually, was to have Strikepods fanned out across the seabed near certain approaches to the eastern seaboard – Bangor, Boston, New York, the Chesapeake, Norfolk, Kings Bay. These would serve as the ears for either roving hunter-killer Strikepods, or Atoms housed in undersea microsubmarine batteries. In this way, AUDEN would essentially be an integrated minefield.
The signature feature of the seabed-optimized Atom was the Advanced Seabed Warfare module which, in addition to providing a suite of highly advanced communication and sensor technologies, also housed the SEASTAR [Seabed Static Array] – a passive microsonar array that would deploy from the module and extend about 150 feet upward into the water column. With their SEASTARs deployed, the networked seabed Atoms could act as a large array, identifying and classifying targets and passing that information along to the Strikepods – either roving, or turret-based.
AUDEN was clearly prompted by the immediacy of the Russian micronaval threat, but were you also troubled by other adversary programs, such as the Chinese Shāyú-class microsubmarine?
We were keenly aware of adversary developments, particularly with the PLAN. The intelligence we’d been receiving on the Shāyú was spotty, but given China’s public successes with Haiyi gliders and deep diving vehicles like the Hailong III, as well as their commitment to an Undersea Great Wall, we were fairly confident that not only did the program exist, but that it was in all likelihood operational. And it wasn’t long before our suspicions proved correct.
In the South China Sea?
About six months after Locust Point, late in the spring of 2017, I got a call from Seventh Fleet inquiring about our progress on the improved Atom, and whether we’d be up for an overseas deployment. At that point we’d pretty much improved the coms issues, and Falken’s training and testing was nearly complete. After checking with the engineers, we thought we were up for the challenge, and on April 1, 2017, half a dozen 5-ship Strikepods were deployed to the Spratlys in what was called Operation Eminent Shadow. Each pod’s initial configuration was one rogue, two remoras, and two relays, and their mission was to carry out general ISR and ASW operations, focusing primarily on the waters off of Subi, Mischief, and Fiery Cross, and to escort both surface ships and submarines performing FONOPS throughout the central Spratlys.
So there was harassment under the sea as well?
Oh, absolutely. Of course, none of these encounters ever made headlines. It simply wasn’t in anyone’s interest to publicize them, and anyway there was more than enough high-profile harassment happening on the surface to keep the public and media buzzing.
The concerning thing was that Chinese subs were showing up unexpectedly, at times and places that suggested they knew where we were. Their Y-8s were dropping buoys practically right on top of our boats, and destroyers and frigates would show up suddenly to shoo us away. It was reminiscent of what we were experiencing with our boomers right before the Istinas showed up at our front door. Needless to say, COMPSUBPAC’s hackle was up. So the Strikepods were there to help figure out what the hell was going on.
Were they under the control of COMSUBPAC?
Actually, they were under the control of COMPODGRU 1, out of Norfolk, Virginia. There’d been a great deal of debate over this – whether Strikepods were a platform or payload, whether there should be a centralized Strikepod command versus a more decentralized payload approach. But we pointed at the Air Force piloting Reapers and Predators out of the southwest, and said look, there’s no reason why Strikepods can’t be run out of Norfolk, particularly when we’re talking about providing a persistent forward presence like any high-value asset.
So we’d airlift the Atoms to Guam, load them onto waiting Virginias, who’d then deploy them – from the torpedo tubes, initially, until we were able to fully integrate a launch and recovery system with the Virginia Payload Module. And then we’d get to work, reporting contacts, keeping tabs on the PLAN. We were even able to test the Remora Module in a live environment, tracking a Yuan-class submarine for ten straight days all over the Central Spratlys. It was all going pretty smoothly, with high fives all around. But the Chinese were still showing up unannounced.
What was it like when you finally encountered the Shāyú?
It was another day of Eminent Shadow, nothing particularly unusual. It’s Caitlyn’s thirteenth birthday, and I’ve got chaperone duty at 1400 for ice-skating and a fondue dinner, so I skip out early, but then at around 1545 I get an “urgent” from the watch: Something’s happened.
When I get back, the place is going bananas, and a watch officer briefs me. One of our Strikepods – Delta – on patrol off Subi had detected three small contacts closing at about fifteen knots. Initially they were classified as biologicals, given their signatures and behavior, and sensing a collision, the rogue ordered evasive action, accelerating and diving, while ordering a relay to break off and come shallow, probably to ensure communications. A second later, one of the contacts breaks formation and heads toward the relay.
So now Falken is faced with something it’s only experienced in the lab. We’d been developing some rudimentary combat tactics and introducing them into the training regimen, running scenarios where Falken would encounter something hostile. But in almost every case it chose to flee rather than fight. There were just too many variables, too much ambiguity for it to go lethal without a human on the loop giving the order.
The first contact reaches the relay and their signals promptly disappear. With the two remaining contacts closing on Delta, Falken orders Flee, and the Atoms scatter. The contacts split up and lock on to the two remoras, and thirty seconds later they also disappear. The rogue scans for additional threats, and finding none, orders relay-2 to head to the surface to phone it in.
Fortunately Falken had ordered all ships to fire their onboard imaging systems, so from the moment Delta went evasive there was running video. Of course 99 percent of it was just black water and bubbles, but there were five screen grabs that were very compelling.
You had a visual?
It was another one of those moments – like Cape Charles when we first laid eyes on an Istina. I was honestly half expecting to see an eye, some flukes, or a long jaw and teeth. But there it was – a short hull, a propulsor. We knew it then – we’d just encountered the Shāyú-class microsubmarine.
Was there continued harassment by the Shāyús during Eminent Shadow?
Almost daily, and as such there was a real sense of urgency at FathomWorks to get Falken in a place to adequately defend itself. We were losing Atoms at a rate of nearly five per week. They needed to fight back.
Were there no attempts at communication? No back channel diplomatic overtures by either side?
It’s important to understand – this was the dawn of autonomous undersea conflict, a time when there were very few environments left for sovereign governments to carry out covert operations without risk of exposure. The Chinese tactics were risky, to be sure, but they believed – correctly – that we had a shared interest in keeping this sort of thing quiet. The undersea community, regardless of nationality, has always been characterized by a cult-like devotion to secrecy, and in the unmanned era, it would be no different. In fact, in some ways there would be an intensification of that silence, if you will, given the willingness to take greater risks with systems that were hidden from public scrutiny, and posed no risk to human life.
But the risk, of course, was that this new type of conflict could spill over into the manned, visible world and precipitate a more serious, potentially bloody crisis. I’m referring, of course, to the Decatur, and the Nanxun Jiao Crisis.
Yes, of course. Just because the conflict is unseen and unmanned doesn’t insulate it from the overarching strategic reality. It’s woven into that reality, and its effects can indeed break the surface and escalate, as it did at Nanxun Jiao.
Can you tell us about the Decatur incident, and the events leading up to the strike on the Nanxun Jiao installation?
So, after Subi we could confirm the Shāyú’s existence, and that the Chinese were serious about it as an ASW platform. It was an eye-opener for sure, but beyond that we knew very little – its capabilities, performance characteristics, or Chinese microsubmarine tactics, or doctrine. But then we had a HUMINT breakthrough.
In early 2018, we get a call from the CIA station chief in Manila. Apparently the embassy received something in the mail that might be of great interest to us – a letter with a simple handwritten sketch of what looks like a missile turret, but on closer inspection the missiles are actually small submarines. Below that sketch are a series of random dots that are actually a fairly accurate representation of the Spratlys. One of the dots has a circle drawn around it – the northern reef of the Gaven Reefs, what the Chinese call Nanxun Jiao.
Needless to say, we were intrigued, but what was particularly intriguing was how the sketch bore a striking resemblance to our working concept for AUDEN. So the CI folks immediately open a file, and we’re left pondering the possibility that the Chinese have deployed a battery of microsubmarines, and that it could be based on a design stolen from a highly classified U.S. Navy program.
What was the response?
Well, we were alert to the possibility of disinformation, but to what end? To draw us in to Nanxun Jiao? A lesser-militarized island, one that appeared to be used primarily for logistics and resupply?
We weighed the options carefully, and ultimately decided we needed to take a look. Rather than divert resources from Eminent Shadow, we shipped a new 5-ship Strikepod via SH-60 to the nearest destroyer, USS Decatur, and about six hours after delivery, the FONOP is underway, with the Strikepod sweeping the reef. At first it’s pretty routine, nothing unusual, but then at the five minute mark, we get a flash: seabed contact. Falken positions the Strikepod for a closer look, and thirty seconds later, another flash. Shāyús in the water. Six of them. So we order an immediate withdrawal to Decatur, and the Shāyús give chase.
By now the PLAN destroyer Lanzhou has made her appearance, and has closed to around three miles. The Strikepod, with the Shāyús in pursuit, is at flank, but won’t reach Decatur before it can be safely recovered. So we have a decision to make. Since the encounter at Subi, the engineers at FathomWorks had been working nearly nonstop on combat scenarios, and even worked with DARPA to develop a special wargame for Falken to help it anticipate conflict and learn how to fight. But we were still uncomfortable with it making the call. So we dipped a SUMO [Shipboard Undersea Modem] and hoped it would get the message.
What followed was seven minutes of sheer chaos. The Strikepod goes hot, and now it’s a furball. All that training seems to have paid off, and we were looking at a much different outcome than Subi. But what we quickly realize is that the Strikepod isn’t the Shāyús’ objective. Three of them disengage and resume course toward Decatur, and by the time the captain orders the ATT [Anti-Torpedo Torpedo System] to engage, it’s too late, and they slam into Decatur’s hull just aft of the sonar dome. Meanwhile, the remaining Shāyús and the Strikepod fight it out until the end. About a minute later the Lanzhou makes its aggressive pass across Decatur’s bow.
Was there any damage to the Decatur?
Nothing of any consequence. Some scratched paint. The Shāyús were inert. It was a warning, and a damn stern one at that.
What was the reaction in Norfolk?
Shock, on a host of levels. We’d just experienced the most aggressive harassment yet. Chinese employment of microsubmarines was unsettling enough, but they were based on the seabed near disputed areas, and were tasked with threatening our manned warships and kinetically engaging our unmanned systems.
But more unsettling still was what the imagery revealed. As CYAN’s letter suggested, the Chinese system was nearly identical to AUDEN. The turret, the network of sensors on the seafloor, complete with SEASTAR-like tentacles.
They’d gotten there first?
AUDEN was still in testing, so it certainly appeared so. Maybe they’d already made a lot of headway, and the stolen design helped push them over the edge. We knew the Chinese were making huge strides in AI and seabed warfare, but this was too much too soon. The signs were clearly there. They’d either hacked us, or they’d had help.
We knew this was coming, though. We knew that microsubmarines were the future of offensive mining operations, and we were well on our way with Strikepods. But we weren’t expecting our adversaries to be quite so far along.
What was the reaction in Washington?
We’d been patched into the Situation Room from the very beginning, so all the principals were well aware of what we’d found. Needless to say, they shared our concerns, but were also particularly concerned with the strategic implications – that China had moved beyond island building to leveraging the seabed for weapons emplacements. And of course, if it was happening at Nanxun Jiao, surely it was happening at Subi, Mischief, Fiery Cross, and others.
It was a complex situation to say the least, and there were some fairly heated discussions on how to proceed. Many believed we should strike immediately, to send a message that this type of illegal installation would not be tolerated, particularly as it has been used to destroy sovereign U.S. property and threaten a U.S. warship. Others called for expanding Eminent Shadow to include all of the waters off disputed outposts in order to build a diplomatic case and compel the Chinese to dismantle the sites.
And where did you fall on the matter?
I felt the crisis called for a blended approach, that neither one alone would have done the trick. Even though a show of force risked emboldening the Chinese, we needed to send a strong signal. Diplomacy could come later, but behind closed doors. A Cuban Missile Crisis-style U.N. confrontation would have made for riveting diplomatic theater, but what if the Chinese didn’t blink? We were talking about what they believed was their territorial sea. So they don’t blink – they refuse to dismantle their microsubmarine batteries. What then? U.S. credibility would be on the line, and we would have no choice but to take action – what would be, at that point, very public action. And suddenly we’re at war in the South China Sea.
And ultimately your approach won the day?
Yes, it did. The president ordered an immediate expansion of Eminent Shadow. We were deploying Atoms by the dozen to the South China Sea, and all told, spent about eight weeks gathering evidence.
As you might imagine, we encountered quite a bit of resistance. The Chinese were onto us now, and were expecting us to go sniffing around. The PLAN was on high alert, stepping up air and surface activities, and Shāyús were everywhere.
The Atom’s LENR [Low Energy Nuclear Reactor] afforded excellent standoff capability, so we were able to launch all of the sorties from well offshore. The Strikepods would arrive on station, form up, and wait for the go order. They’d fight their way in, and, with any luck, a few Atoms would penetrate far enough to capture some imagery, beam it to the others, and then race to the surface to relay the data. The plan worked about 75 percent of the time, gathering more than enough evidence to satisfy the policymakers. We lost nearly 120 Atoms out of the 150 participating in the expanded Eminent Shadow. Expensive, yes, but they weren’t human casualties. This new form of conflict was evolving right before our eyes.
And what did you find?
Ultimately we found operational systems off of Subi, Mischief, and Fiery Cross, with partial construction off Hughes, Johnson, and a few others.
So you have the evidence you need to make a compelling case. Now you move on Nanxun Jiao?
Yes, the planning for the Nanxun Jiao strike – or what was now being called Operation Roundhouse – was complete.
Tell us about Operation Roundhouse.
During the expanded Eminent Shadow, we were also keeping a close eye on Gaven Reefs, with particular focus on any human comings and goings at the Nanxun Jiao undersea installation. The last thing we wanted was casualties, and from what we could tell, the schedule was routine. Every Wednesday at 10am local, a dive tender would head out over the spot and they’d deploy, and we’d listen as they performed routine maintenance, and after about two hours or so they’d surface and head in.
After several weeks of recon, the date was set – a Saturday at 0300 to avoid any possibility of human activity or unwanted attention from satellites or passing aircraft. (The turret was at a depth of fifty meters, so there would surely be a disturbance on the surface.)
The package was sixty Atoms strong, organized into three waves of Strikepods that would make a staggered approach. The Shāyús had a tendency to be all in at the first sign of trouble with almost no reserve, so we hoped the first wave would draw them out, exposing the turret’s flanks and allowing waves two and three to penetrate and carry out the mission.
The package launched at 0215 for the cruise to staging, approximately 10 miles west of Nanxun Jiao. The go order went out at 0245, and then we waited, watching a real-time feed via a surfaced relay trailing approximately 500 yards behind.
At three miles out we hear the Shāyús pinging away, and at two miles they engage the first wave. Resistance is light, which, in hindsight, should have been cause for concern. We only lose about fifteen Atoms during the first wave, and waves two and three meet with almost no resistance at all. Falken had been trained to race toward the target and detonate at the last possible moment, but with such little resistance we were able to perform a static demolition to ensure more thorough destruction.
So we pull back, and we’re watching the split screen with feeds from various perspectives, including the kill vehicles as they continue to close, and as the turret comes into view…
[Admiral Lacy pauses here.]
That’s when you saw them?
I called immediately for confirmation. The pilots switched frantically between feeds, and, yes – quite distinctly. Dive suits. Four of them.
The room erupted, and we immediately fired off a flash to abort. The relay confirmed, but it was already too late, and half a second later the screens went blank. BDA later confirmed that the mission was a success. The turret was completely destroyed, and many of the surrounding network sensors appeared badly damaged or disabled. We searched in vain for three hours, but there was no trace of any divers.
The next day we received the news.
From Chinese state television?
The lead story was an explosion at a CNOOC facility that claimed the lives of four offshore divers. Of course, it wasn’t uncommon for Beijing to cover up accidents of any kind, particularly when they were security related or politically sensitive. But then we received word from CIA, who’d been running CYAN from the beginning. They’d been trying to track him down, to recruit him for more, and in the course of their research they’d determined that the letter had likely come from a member of the Nanxun Jiao dive team, and that one of the four divers reportedly killed in the CNOOC accident – a Mr. Xin Li – was in fact CYAN.
The Chinese had been onto him?
Perhaps CIA had been careless while poking around and attracted the attention of Chinese counterintelligence, or…
Or they’d been tipped?
A very real possibility as well.
And they’d known you were coming?
The fact that Mr. Li had been exposed, and that this had likely led to his death – his execution – along with those of three others, was of course problematic enough. But the fact that there were divers down as Roundhouse was underway, and that one of those divers was most likely the individual who led us there in the first place – the implications were unimaginably grave.
Some questioned whether they’d been killed elsewhere, and were never down there at all. But I’m fairly certain they died there that night. The Chinese wanted to send a message to would-be traitors, and especially to us. It was textbook psychological warfare. They wanted us to know we’d pulled the trigger.
What I can tell you is that there were people in the room that day who were never the same again.
I know I wasn’t.
And the aftermath?
Nanxun Jiao was spun as a stalemate, but it was really a Chinese victory. Four casualties and the perils of autonomous undersea conflict were too much for the politicians to stomach, and so, much as Chinese island building had gone unchecked for years, so too would their undersea buildup continue.
In just three years we were an order of magnitude beyond Locust Point, and the brave new world of autonomous undersea conflict was coming into focus. The underwater realm was more secretive, more complex, and more dangerous than ever before, but one thing in particular was becoming inescapably clear:
Nothing less than total undersea dominance was at stake.
[End Part II]
David R. Strachan is a naval analyst and writer living in Silver Spring, MD. His website, Strikepod Systems, explores the emergence of unmanned undersea warfare via real-time speculative fiction. Contact him at firstname.lastname@example.org.
Featured Image: Project 885 Yasen Class Submarine by Isra Tan.