USS Stark steams quietly near Bahrain on May 17, 1987. The Tanker Wars between Iran and Iraq are ongoing, and the United States is trying to keep commerce flowing in the Persian Gulf. Unbeknownst to the crew, it is the final hour of life for 37 sailors onboard the Stark.
2000 hours: USS Coontz gains contact on an Iraqi F-1 Mirage attack aircraft and sends the Stark the track.
2005: Range: 200 nautical miles. Stark’s commanding officer is informed.
2055: The commanding officer asks why the Mirage isn’t on radar. The operator increases range scales. Target acquired. Range: 70 nautical miles.
2102: The Mirage illuminates Stark with her radar, locking on. “Request permission to warn the incoming bogey,” the radar operator asks the tactical action officer — the person responsible for the defense of the ship. He replies, “No, wait.”
2105: Range: 32.5 nautical miles. The Mirage turns to intercept Stark. The team in the Combat Information Center misses the turn.
2107: Range: 22.5 nautical miles. The Iraqi pilot launches the first Exocet anti-ship cruise missile. Two minutes to impact. The forward lookout on Stark sees the launch but misidentifies it as a distant surface contact.
2108: The Stark begins hailing the Mirage on the radio. The Iraqi pilot does not respond; he is busy launching his second missile. Stark’s systems detect another radar lock on their ship. The tactical action officer gives permission to arm the counter-measure launchers and place the Phalanx close-in weapons system, which would mount the final defense for the ship, on standby. One minute to impact.
2109: Stark locks onto the Mirage with her radar. The lookout reports an inbound missile, but the report is not relayed to the tactical action officer. Seconds later, the first Exocet slams into Stark but fails to explode. Lt. Art Conklin, the damage control assistant, recalls, I heard the horrible sound of grinding metal, and my first thought was that we had collided with another ship. I immediately opened my stateroom door and headed for Damage Control Central. Within a fraction of a second I knew we were in trouble. I smelled missile exhaust and heard over the 1MC [ship’s main announcing circuit], “Inbound missile, port side! All hands brace for shock!”
2109:30: The second Exocet rips into Stark and explodes. In less than a minute, nearly a fifth of the crew is killed and many more wounded as the fuel from the unexploded first missile continues to burn. Heroic efforts throughout the night are all that keep Stark from sinking.
The effects of the Exocet, the most lethal anti-ship cruise missile at the time the above story took place, were devastating for a ship not expected to survive even a single missile hit by the Navy’s ship survivability standards. But the Navy will not be so lucky in the future. As the team on the Stark demonstrated, even a capable crew can fail to adequately respond to the speed of naval combat in the missile age. An hour of confusing and threatening behavior by the Iraqi aircraft was followed by two short minutes to realize that Stark was under attack.
The inability to keep up with the increasing speed of naval combat is only going to get worse. The advent of hypersonic weapons, particularly anti-ship cruise missiles, represents a grave threat to U.S. surface forces. Today, the Kalibr missile system, used by Russia, China, and many other countries, accelerates to Mach 3 in the minutes prior to impact. It took almost two minutes for the subsonic Exocets to cover the 20 nautical miles to Stark; the Kalibr missile could cover that same distance in a mere 45 seconds. Russia announced initial operational capability for the Mach 8 Zircon cruise missile in mid-2017 as well.
Hypersonic missiles, which travel at speeds greater than Mach 5, shorten John Boyd’s famous observe-orient-decide-act loop, making it nearly impossible for human minds and teams to even comprehend the information, let alone defend against a short-range attack. These new weapons represent a paradigm shift in naval combat. When discussing salvos of anti-ship cruise missile attacks against U.S. warships, some brush off the threat, intimating that various anti-missile missiles and counter-measures will protect us. But a closer examination of U.S. defensive systems and adversaries’ offensive missiles calls such sanguine assessments into question. Combat leaders should strive to more deeply understand the nature of the threat — specifically the science behind naval combat and missile attacks. The mathematics shows that in a missile war, mere survival is difficult, let alone success. It only gets worse as the number of missiles inbound increases. This problem should prompt the Navy to take a hard look at how it will defend against this threat and how it should train its sailors — examinations that are underway but not adequate to address the magnitude of the current threat. A review of the salvo equations, another historical case study, and a logical next technological step are in order.
Doing the Math
Salvo equations provide a mathematical model of combat losses in missile warfare by relating the number of missiles fired and the probabilities of a miss, of shooting missiles down, and of damage resulting in a mission or catastrophic kill. The models are simple and can be implemented on a spreadsheet, making it easy to toy with “what if?” scenarios to see how each element of the equation can affect survival.
Capt. (Ret.) Wayne Hughes, the father of modern naval tactics, analyzed the historical record of missile attacks against various ships — both merchant shipping vessels and warships. He found that warships that failed to provide a defense when attacked faced a 68 percent probability of being hit. Mounting a credible defense reduced the chance of being hit to 26 percent. The salvo equations can be further analyzed to show the effects of poor human performance, improper use of radar and detection systems, and the inability to eliminate all incoming missiles in the salvo. Hughes sums up the conclusions succinctly: Attacking effectively first is paramount. Failing that, if the ship expects to survive and avoid taking hits, the human team has to mount a timely and effective defense, which is increasingly challenging and will only get more difficult as hypersonic weapons drive faster engagement speeds and observe-orient-decide-act loops.
Naval Combat in the Missile Age
S.L.A. Marshall, the eminent Army historian, asserted in his landmark 1947 study Men Against Fire that, even in the best infantry companies, only 25 percent of soldiers actually fired their weapons in combat. The statistic seems bizarre and counterintuitive, but Marshall provides solid evidence from World War II. Even in some of the grimmest, close-quarters combat in the Pacific islands and during the landings at Normandy, only an average of 15 percent of infantrymen actually fired their weapons. The underlying reasons most soldiers did not engage, according to Marshall, were a lack of a definite targets, worries about fratricide, unwillingness to reveal their position, and the fog of war.
Do Marshall’s conclusions hold true onboard a modern warship? Stark proved that a ready, well-trained Navy combat team could similarly fail to fire its weapons despite sufficient indications of impending attack. The HMS Sheffield (D80), sunk by the Argentinians during the Falklands War in May 1982, is another example of warships failing to engage in combat during wartime.
In both cases, watch teams reacted with human precision. Psychological biases, team dynamics, and personalities all play a role. A timid watch officer may hesitate to call the captain, even though he knows he should, let alone launch weapons or counter-measures to defend the ship. As the Navy moves into the age of hypersonic weapons, the lack of time available to observe, orient, decide, and act on the battlefield will overwhelm even the best watch teams. Humans simply cannot cope with the speed of future naval combat.
The Vincennes: Decision-Making in the Fog of Naval Warfare
The next war, fought at an accelerating cognitive tempo thanks to hypersonic weapons, cyber effects, and an even more saturated information environment, will require even more of human watch teams. The accidental downing of Iran Air Flight 655 on July 3, 1988 by the USS Vincennes demonstrates the dangers of a more chaotic combat environment and the increased cognitive demand that it requires.
0330: USS Elmer Montgomery detects 13 Iranian gunboats nearby. They split into three groups, and one takes station off Montgomery’s port quarter.
0411: Montgomery reports 5-7 explosions to the north near merchant shipping. Vincennes is ordered to assist. She orders her helicopter, OCEAN LORD 25, on ahead.
0615: OCEAN LORD 25 is attacked by rockets and small arms fire. Vincennes sets general quarters and all hands man their battle stations.
0620: Vincennes takes tactical control of Montgomery.
0639: Vincennes calls her operational commander and requests to engage. Permission is granted.
0643: Vincennes opens fire with her main 5 inch guns. She comes under small arms fire from the gunboats, now inside 8,000 yards.
0647: The radar operator gains a new air contact taking off from Bandar Abbas, 47 nautical miles away and heading for the ship. Classification unknown.
0648: USS Sides also detects the aircraft, now designated Track 4131, and locks on with a missile on her forward launcher.
0649: Vincennes begins challenging the aircraft continuously on international and military distress frequencies.
0650: Someone reports Track 4131 as an Iranian F-14, despite it transmitting a civilian transponder code. The forward gun suffers a casualty and can no longer fire. The tactical action officer orders the continuous challenging of Track 4131 over the radio.
0651: Vincennes reports her intention to engage the F-14 at 20 nautical miles to her operational commander. Vincennes takes tactical control of Sides. Vincennes begins maneuvering radically, using maximum rudder and speed in an attempt to keep the aft gun mount engaged on the Iranian gunboats. Books, loose equipment, and other items begin falling from the shelves around the ship.
0652: Several watchstanders incorrectly report Track 4131 descending in altitude.
0653: Track 4131 closes to 12 nautical miles, still on an intercept course with Vincennes. Vincennes requests to engage — granted.
0654: Vincennes launches two missiles.
One minute later, both missiles destroy Iran Air Flight 655. The commanding officer of the Sides sees the explosions and the debris falling. It would be several hours, long after Flight 655 is reported overdue, that the mistake is realized.
In the inquiry that followed this real incident, the chairman of the Joint Chiefs of Staff concluded that, given the “pressure-filled environment” on board the Vincennes, the outcome was a “reasonable performance under the circumstances” and that “it is imperative to have an emotional and intellectual feel for that picture.” At the time, Vincennes had been battling several groups of Iranian fast-attack craft for three hours — each capable of damaging the ship and killing personnel, tracking an Iranian P-3 maritime patrol and reconnaissance aircraft flying a classic targeting profile to provide targeting information for Iranian attack aircraft, fighting with a gun mount out of commission, maneuvering violently to keep the aft gun shooting, and being tactically responsible for two additional warships (the Sides and the Montgomery).
This was a complex, cognitively taxing scenario for all involved, especially the decision-makers. The fog of war was thick. Fatigue from hours of close combat wore down the officers and sailors. Rear Adm. William M. Fogarty, the senior investigating officer, realized he needed professional advice from Medical Corps personnel specializing in combat stress to help analyze the interview and physical data for this event — the effects were that strong. He found that the tactical information coordinator, an enlisted watchstander in the combat information center, emerged as a de facto leader based on his perception of a weaker supervisor; his recommendations were accepted by all, and he firmly believed he was tracking an inbound Iranian F-14 — the Anchoring Effect in full force.
Both case studies show that Navy human teams, even with a reasonable expectation of combat, can take several minutes to recognize the situation and act. Factoring in hypersonic weapons, watch teams must now process the threat even more rapidly, take defensive measures, and perhaps, succeed in knocking the one incoming missile from the sky, to say nothing of a salvo of missiles. Factor in any friction, such as whether the aircraft detected is hostile or not, as in the case of Vincennes, and the decision cycle will be even further lengthened, assuming action is ordered at all.
Searching for a Solution
The related fields of artificial intelligence, machine learning, and deep learning have unique applications for naval warfare in the hypersonic age. Deep-learning machines formed the core of the ATHENA ship self-defense system in Peter Singer and August Cole’s Ghost Fleet, responding to incoming attacks with a precision and speed unachievable by human operators, automatically identifying, tracking, and assigning the best weapon to counter the threat, and, if in automatic, launching weapons to defeat it. Moving away from fiction, machine-learning algorithms would help the Navy baseline its systems and spaces to aid in damage control, detect and track contacts (especially in highly uncertain areas like sonar), and much more. Artificial intelligence algorithms can be trained to pair with humans to complete tasks like mission planning and developed to provide greater information, not just data, across the decision cycle. In particular, deep learning shows a remarkable ability to distill patterns across many sets of data. In the case of Vincennes, an algorithm could have parsed airline flight schedules or historical air traffic patterns and dispassionately evaluated Iran Air 655 as a non-threat. Similar pattern analysis would have aided Stark in identifying the attack profile presented by the Iraqi Mirage.
The Navy has begun to invest in artificial intelligence, but the efforts have been lackluster. Writing in War on the Rocks recently, Connor McLemore and Hans Lauzen provided great advice for the Navy as it looks to invest heavily in artificial intelligence. Many other authors have already proposed applications of artificial intelligence in other maritime areas, but the service’s urgency to deploy algorithms has lagged significantly behind. The institution’s overall advance into artificial intelligence is everything you would expect from a major, cross-cutting new development effort in a military branch — unfocused and lacking vision. Any proposed “strategy” looks more like a shopping list than a coherent strategy linking a diagnosis of the problem with a guiding policy and set of coordinated actions to achieve it. Navy leaders have come out openly to discuss the pursuit of artificial intelligence, but those statements do not indicate that the Navy will fundamentally change the way that it develops and acquires software or delivers capability to sailors. Given the proposed timelines for fielding modernization packages to ships and aircraft, and of developing and maturing net technologies, the Navy’s pursuit of artificial intelligence will likely be insufficient to counter the threat. America’s adversaries are publicly disclosing the successes of their hypersonic weapons. Their new capabilities seem to be deploying with greater speed and agility than the Navy is capable of doing right now. Artificial intelligence algorithms are what we train them to be based on the data sets we have. Regardless of the Department of Defense’s recently released artificial intelligence strategy, the lack of a coherent strategy and vision for what the Navy’s research and development or acquisition enterprises should be developing will only hamper efforts to deploy credible capabilities in the eyes of the warfighters.
At the dawn of hypersonic weapons, the Navy is woefully unprepared. Nearly a decade of budgetary cuts and nearly two decades of operations in support of the global war on terror have exhausted the fleet and slowed technological development. American sailors will be involved in missile combat in the future, whether from a terrorist-launched Kalibr missile or a classic state-on-state war, and lives will be lost. Just how many lives will depend on the willingness of the Navy to use America’s technological and intellectual might to pair man and machine against the threats. The Navy cannot afford to wait any longer to develop and deploy algorithms for offense and defense. Its very existence as a service depends on it.
Lt. Cmdr. Ryan Hilger is a Navy engineering duty officer stationed in Washington, D.C. He writes frequently on the intersection of history, technology, and leadership. His views are his own and do not represent the Department of Defense.