IS1 [Redacted]: The first Chinese missiles fell short of the mainland and exploded harmlessly in the Pacific Ocean. They didn’t even hit Hawaii. Not even close. It was bizarre. All of our intelligence re- ports indicated that they had a very reliable ballistic missile system. At least the blasts weren’t nuclear. I’m so confused. Something just doesn’t add up.
ETR3 [Redacted]: The war was barely an hour old and none of our submarines had checked into the command net after the flash mes- sage we sent out. Not even a slot buoy. Not that we expected them to, given the lack of reliable satellite communications, but no one had come up on any net. After a few hours, a weak “beast” buoy signal started coming across, but it was too weak to localize because we had too many satellites down.
LTJG [Redacted]: I don’t know how the surface fleet escaped dam- age from this attack. The carnage in Pearl Harbor was appalling. After things calmed down a bit from the initial hysteria of the attack, reports started trickling in from the west coast that numerous tankers had run aground or were sinking in the shipping channels at Long Beach, San Francisco, and the Strait of Juan de Fuca. The Coast Guard sectors had declared those ports closed indefinitely until the wrecks could be moved.
The Board of Inquiry into the opening salvoes of the Sino-Ameri- can War would take more than a year to complete. The data was flung across many organizations, difficult to transfer, sometimes corrupted or unreadable, and many of the key pieces of information that would have indicated impending maritime hostilities were missed. Many of the of- ficers and Sailors said that the volume of data coming into the Maritime Intelligence Operations Centers was more than they could cope with most days, especially when there were a lot of signals or alerts that need- ed further investigation.
The undersea surveillance sites were hit particularly hard. The Chi- nese maritime militia, fishing fleets, and some Chinese Coast Guard units moved further east than they ever had before. Many ships even pushed well east of Guam. Sailors at the Naval Ocean Processing Facil- ity (NOPF) on Whidbey Island, Washington, told investigators that in many cases they simply couldn’t separate all of the contacts out to de- termine what was going on. Despite augmenting their watch bills as the activity picked up in the week before the war broke out, the leadership stated flat out that they could not cope with the deluge and that many sectors not looking to the west were simply not searched. Arrays started going down sporadically, taking more sailors off the search to trouble- shoot. The Sailors managed to track a few submarines, but it would be another year or two before post-event analysis figured out that the auton- omous underwater vehicles (AUV) that had played a crucial role in the attacks on commercial shipping had preceded even the submarine surge by more than a week.
The missile boats that managed to evade detection reached launch positions near Hawaii and the west coast before the intelligence com- munity had realized that they were missing. Strangely, none of our ships were directly targeted in the opening attacks. Waves of hypersonic weap- ons slammed into dry dock caissons, fuel depots, and graving docks.
The Chinese submarines had learned the art of steganography: hid- ing in plain sight, in the noise. The Board of Inquiry, similar to the reck- oning after Pearl Harbor, went further than most to look at not just how American forces responded to the attacks or what was missed in the tac- tical signals, but where the Navy veered off course in the years leading up to the war in failing to develop the systems necessary to prevent such wholesale surprise. The President of the Board found, in light of the pronouncements of renewed great power competition almost a decade before, that the Navy had not sufficiently prioritized recapitalization and improvements of the undersea surveillance systems, specifically in the areas of expanding research into oceanography to better exploit the environment for both offense and defense, expanding the coverage of the system, leveraging industry gains in artificial intelligence and ma- chine learning to improve accuracy, and developing better technologies to sense and track both submarines and AUVs. Had we kept pace or overtaken Chinese investments in these areas, the President of the Board stated, we might have prevented this devastating attack.
Before the War
Both China and Russia have made tremendous gains in recent years in the areas of submarine operations, artificial intelligence, and hyper- sonic weapons. Both powers appear to be dedicated to overthrowing the current world order by out-running the United States technologically. China, in particular, has a penchant for taking a much longer and more strategic view of events than the myopic United States. Their stated de- velopments, if left unchecked, will topple the international order by fait accompli—armed hostilities with the United States only if necessary. Preserving the position as a world leader since World War II will require the United States to be able to credibly deter Chinese or Russian designs. Strategic developments, particularly in the Western Pacific, have left the undersea as the last domain where we enjoy an advantage. Widening that advantage through the use of better intelligence and exploitation of the environment will keep that credibility intact and forestall the war of which the Secretary of Defense, Chief of Naval Operations, and other senior leaders speak today.
The Cruel Sea
Traditionally, unlike war on land, the sea is utterly barren, neutral terrain. It is the great equalizer between enemies. It used to provide little additional value to those on the attack or defense. But as science ad- vanced, we learned more about the oceans and slowly learned that we can exploit them for tactical advantage. Starting early in the Cold War, the United States started conducting more research and installed the Sound Surveillance System (SOSUS) to provide broad ocean coverage. Intelligence collection in this domain evolved to not just knowing the enemy, his operating patterns, technology, and capabilities, but under- standing how the environment connects them and us. That connection, observed by SOSUS, is how we held a decisive advantage over the So- viet Union during the Cold War, an advantage that continues to this day, albeit one that is deteriorating steadily.
Getting information from this environment and exploiting the ocean in real time continues to be challenging at best. Low bandwidth, asyn- chronous communications are the norm. Classifying detections proves eBually difficult and is more reliant on corporate knowledge from pre-vious encounters than well-encoded processes to identify what a Sailor is looking at. Maintaining an acoustic contact can be as frustrating as the environment changes from moment to moment, region to region, or as other noises interfere. The availability of detailed, real time acoustic modeling at the tactical edge is virtually non-existent. The shore-based SOSUS monitoring stations, like NOPF Whidbey Island, have evolved technologically, but are still heavily reliant on trained Sailors to detect, track, and classify signals across the expanse of an entire ocean. They do an exceptional job, but the maritime environment is slowly overwhelm- ing their capacity.
The Growing Threats
Russia and China continue to present an evolving threat to the sta- bility of the international political and economic order. Submarine ac- tivity from both countries has now returned to Cold War levels.1 Both countries have announced programs to field hypersonic weapons, un- manned underwater systems, new sensors, and artificial intelligence to augment military forces.2 Russia continues to show intense interest in the undersea cables that carry 97 percent of global communications and more than $1E trillion in daily financial transactions. Satellites cannot reconstitute even a fraction of this capability.3 Russian submarines have been found by other nations near these cables—their locations are pub- licly available online.4 Russian submarine operations in 2017 prompted the Navy to re-establish the U.S. Second Fleet in May 2018 to respond to their resurgence in the Atlantic.5 The Cold War has started anew.
China stated openly their desire to augment submarine commanders with artificial intelligence. They are investing heavily in Buantum tech- nologies and are rumored to have made a breakthrough in a quantum detector that can be used to detect submarines.6 If true and deployed, it could effectively shut out the United States from the undersea, the last domain where we still enjoy the advantage. Chinese hypersonic technol- ogy appears to be well ahead of the Russians.
The development of submarine launched hypersonic weapons cou- pled with disruptions to the global communications and financial net- works presents a near-existential threat for the United States. Admiral James Foggo’s postulation that the Fourth Battle of the Atlantic is begin-ning should be broadened to encompass the Chinese and Russian threat in the Pacific as well.7 The Cold War has expanded to a second front. Unlike the last Battle of the Atlantic, the United States may be unable to carry the war to the enemy if we cannot retain the advantage in operating without detection underwater and have exceptional abilities to detect and track adversary forces. We must part the seas.
Despite the meteoric rise of China and their philosophy on research, development, and theft as a means to innovate, the United States still retains the advantage in science and technology. To keep our advantage undersea, the intelligence community in particular must double down on investments in advanced sensors, artificial intelligence, and developing the human-machine team at all levels—from tactical to strategic. The costs will be high, but they must be paid.
Data as a Strategic Asset
We must recapitalize the existing infrastructure and networks with new sensors, cables, systems, and platforms designed to produce quality data that supports the training and development of artificial intelligence and machine learning algorithms. Without properly curated data, the al- gorithms become only as good as our human operators, or worse. While many senior leaders, especially in the intelligence community, will state emphatically that the United States has mountains of data, the vast ma- jority of that data is not in the correct format or quality for use in ma- chine learning applications, nor could it be made readily available for al- gorithmic training. The intelligence community must work to pay down the organizational debt of data that has accumulated over the decades, but in parallel, the community must also partner with other communities to field the new physical systems that will produce the data necessary to do our jobs better in the future. Artificial intelligence will not reinvent naval intelligence overnight but laying the foundations correctly now for future applications will ensure that the United States remains supreme in the undersea domain.
The artificial intelligence community is remarkably open source. Few algorithms or research are behind even something as simple as a pay wall, much less branded as intellectual property and sealed from public view. The performance of the algorithm thus is tied to the qual-ity of data used to train it. Data makes the capabilities emerge. Naval intelligence must make data a core business function of the community, strengthen relationships with the ‘collection’ communities, and realign investments and programs to ensure that the data produced will not just feed the analytical products crafted for decision makers, but improve algorithmic performance and support the warfighter better at the leading edge on a daily basis.
Human-Machine Team
Many fear that artificial intelligence will result in entire industries shedding their workforces; that machines will replace human minds. Re- cent advances in artificial intelligence and machine learning have shown that, in fact, the human-machine team is far superior to the machine or the human alone. Leveraging the uniBue abilities that artificial intelli- gence and machine learning algorithms bring to the functions of the in- telligence community will result in non-linear gains in intelligence and warfighting capabilities as a result.
At the most basic level, artificial intelligence and machine learning algorithms excel at unburdening human operators from repetitive tasks, such as conducting a sonar search for hours on end. Human operators fatigue, miss signals, cannot discern the signal from the display-ready data, or misclassify the contact. The challenge to our Sailors will only continue to grow as the seas become more congested with commercial traffic, China and Russia continue to deploy Buieter submarines, and more AUVs, which are significantly Buieter, roam the oceans. Algo- rithms, on the other hand, require no sleep and constantly search ev- ery azimuth with higher accuracy rates than their human operators. The human-machine team, in this case, would allow the human operator to use the advantage human intelligence brings in analyzing and higher cognitive functions to bear on signals the machine has identified for follow up. More advanced applications at the source of the data—the hydrophone—may reveal new information that was previously filtered out through signal processing, furthering our advantage in the undersea domain.
Many applications of machine learning, specifically deep learning algorithms, show incredible promise to draw out connections from seem-ingly disparate data. Fusing open source intelligence with the acoustic collection systems will provide a significantly greater ability to detect, track, and classify contacts of interest and inform human operators when Russian submarines, for example, appear to be heading toward undersea cables. Similar algorithms can be used to explore historical intelligence data and look for new information, operating patterns, and previously undiscovered connections as well.
As computational hardware gets smaller, it enables greater process- ing of data and information into refined products closer to the source. This creates the opportunity for greater synergy between the intelligence community, the meteorology and oceanography community, naval re- search enterprise, and the warfare communities, which exploit their products for maximum advantage. Fostering closer relationships will allow the intelligence community to undertake collection and exploita- tion initiatives more rapidly, leveraging the latest gains in science and oceanography to give warfighters the greatest offensive and defensive advantage possible.
The fielding of hypersonic weapons and increased submarine ac- tivity from both Russia and China makes it imperative that the Navy takes substantial steps to maintain and grow our undersea advantage. The intelligence community should prioritize investments in new and advanced sensors to recapitalize the undersea networks and generate ex- ponential gains by applying artificial intelligence and machine learning algorithms to the data outputs. Rapid development and fielding of these algorithms, coupled with underpinning data as a core element of the broader naval intelligence community, from analysts to operators, will allow the United States to remain ahead of our peer competitors in the undersea domain. Our security as a nation depends on it.
Sino-American War–Redux
Let’s look at an alternate view of that future Sino-American War, depicted in the first segment of this article with the investments specified in the paragraph above initiated in 2018.
The United States had called the Chinese out for their aggressive approach to American targets. The undersea surveillance systems, en- hanced with artificial intelligence and machine learning algorithms, au-tomatically detected the AUVs and submarines, notifying the operators that the Chinese had moved further east than ever before. American sub- marines took up the trail, ready to deploy non-lethal means to deter and disrupt the Chinese undersea forces if they approached within missile range of Hawaii or the west coast. The first few missile submarines were forced to the surface and American cryptanalysts smiled at hearing the submarine commanders radio Beijing for a tug after their submarines were disabled. The AUVs, slow and unable to evade pursuers, were cap- tured wholesale by American warships, broadcast on television in real time for the world to see. Meanwhile, other American submarines, ex- ploiting the smallest margins in the environment, slipped undetected into the East and South China Seas to begin non-kinetic operations against the installations in the Paracel and Spratly Islands. New technologies to exploit the environment had allowed the American submarines to per- fect steganography and hide within the background clutter. The Chinese backed down, recalled their forces, and accepted the American offer to negotiate. The advances in the American naval intelligence systems had allowed the United States and China to escape Thucydides’ trap.8
Endnotes
1 Loulla-Mae Eleftheriou-Smith. “Russia’s Submarine Patrols Reach Levels Last Seen During Cold War.” Independent. April 3, 2017. https://www.independent.co.uk/news/world/europe/russia-submarines- patrols-highest-levels-cold-war-attack-putin-fleet-a766D8D1.html
2 Stephen Chen. “China’s Plan To Use Artificial Intelligence To Boost The Thinking Skills Of Nuclear Submarine Commanders.” South China Morning Post. February 6, 2018. https://www.scmp.com/ news/china/society/article/2131127/chinas-plan-use-artificial-intelli- gence-boost-thinking-skills
3 Rishi Sunak. “Undersea Cables: Indispensable, Insecure.” Policy Exchange. December 1, 2017. https://policyexchange.org.uk/wp-con- tent/uploads/2017/11/Undersea-Cables.pdf
4 David E. Sanger and Eric Schmitt. “Russian Ships Near Data Ca- bles Are Too Close For U.S. Comfort.” New York Times. October 25, 2015. https://www.nytimes.com/2015/10/26/world/europe/russian-pres- ence-near-undersea-cables-concerns-us.html
5 Sam LaGrone. “Navy Reestablished U.S. 2nd Fleet To Face Rus- sian Threat; Plan Calls For 250 Person Command In Norfolk.” USNI News. May 4, 2018. https://news.usni.org/2018/05/04/navy-reestablish- es-2nd-fleet-plan-calls-for-25E-person-command-in-norfolk
6 David Hambling. “China’s Quantum Submarine Detector Could Seal South China Sea.” New Scientist. August 22, 2017. https://www. newscientist.com/article/2144721-chinas-quantum-submarine-detec- tor-could-seal-south-china-sea/
7 Vice Admiral James Foggo and Alarik Fritz. “The Fourth Battle of The Atlantic.” USNI Proceedings. Volume 142, Issue 6, June 2016. https://www.usni.org/magazines/proceedings/2016-06/fourth-battle-at- lantic#footnotes
8 Graham Allison. Destined for War: Can the United States and China Escape Thucydides Trap? New York, NY: Houghton Mifflin Har- court, 2017.
