The Unplanned Costs of an Unmanned Fleet
Two subjects are nearly inescapable in commentary about the U.S. Navy today. The first is the much-maligned, 15-year saga of the littoral combat ship (LCS), which has provided an unfortunate case study for interest group capture, misalignment of ends and means, cost overruns, and engineering failures.
The second subject is more hopeful: proposals for unmanned surface vessels that will deliver cost savings and increase the size of the fleet. As China leap-frogs the United States in raw numbers of ships built and deployed, this subject has acquired great urgency. It includes concepts such as human-machine teaming and autonomous swarming as proposed cost-effective solutions to the numerical asymmetry at sea.
Very little commentary, however, explicitly connects the two subjects. This is unfortunate because, while the LCS is not unmanned, it is further on the unmanned spectrum than any other U.S. Navy vessel in operational use, making it the closest real-world test case for future surface fleet architecture. In fact, the central argument for zero-manning — that removing sailors from ships will save money, allowing the Navy to purchase and field large quantities of vessels for distributed maritime operations — is exactly what the LCS program promised.
In the case of the LCS, this promise was a fallacy for two reasons. First, replacing sailors with technology reduced maintenance at the operator level, but increased it at the regional maintenance center and original equipment manufacturer levels. This raised costs overall, meaning fewer platforms could be purchased. Second, minimal manning made platforms less resilient. Fewer sailors meant fewer problems spotted, and less capacity to fix them while underway. Hence, if fielded in anything approximating combat conditions, the LCS would not remain effective for long. We argue that these two challenges are as — if not more — likely to occur on unmanned ships as they did on minimally manned ones.
All the Problems of the LCS and Then Some
Former Deputy Secretary of Defense Robert O. Work has shown that the LCS was far from a “ship without a mission,” as critics have sometimes contended. The program had a clear plan, but it required vast technical leaps to build a ship with a core crew of only 40 sailors, compared to the 176 on the Oliver Hazard Perry-class frigates that the LCS was intended to replace.
The LCS was designed for what the Navy calls “distributed maritime operations” (then called “Global Concept of Operations”), its principal operational concept for achieving sea control in a high-end fight. This entails dispersing numerous vessels — large and small, manned and unmanned — across an operating area to act as sensors, defensive pickets, and strike platforms. Networked together, these vessels can exploit common information streams and synchronize maneuver and fires. The LCS’ chief attributes — speed, small size, low cost, modularity, and minimal manning — were designed to deliver the quantity, geographic range, and flexibility in roles that distributed maritime operations demanded.
These attributes were supposed to be complementary. Minimal manning would reduce life-cycle costs, since there would be fewer living spaces and life-support systems to maintain, fewer personnel to pay, and fewer underway replenishments. The LCS’ swappable mission-modules for anti-surface, anti-submarine, and anti-mine warfare would also reduce costs, since disparate missions would no longer need purpose-designed platforms like minesweepers, or large ones with multi-mission capability. The ship’s low production and life-cycle costs would enable the Navy to buy more of them, and quantity would mean greater geographic reach for the fleet. It would be a virtuous cycle.
The linchpin of the virtuous cycle was cost savings, which the Navy expected to derive from the LCS’ minimal-manning construct. Cost savings have become a perennial concern as the Navy faces pressures to expand the fleet while trying to make up for repeated cost overruns in programs including the Ford-class supercarrier, Zumwalt-class destroyer, and many lesser-known examples. Moreover, the Navy faces broader budget pressures from outside the organization, including from partisan politics and inflation.
The Navy reduced its planned acquisition of LCS hulls from 55 to 32 in 2014, and five years later Congress barred the Navy from purchasing any more. However, implementing distributed maritime operations remains the Navy’s ultimate goal. It requires that the Navy acquire a lot of ships — and soon. As the Navy’s march to its previous target of 355 ships has stalled, planners are now banking on the theory that cost savings from unmanned vessels will make up the difference between current force levels and what distributed maritime operations will require. They made the same argument for the LCS.
As Crew Numbers Go Down, Maintenance Costs Go Up
Many of the early documents about and analysis of the LCS program stressed affordability, driven by optimal manning, rotational crew-deployment models, and reliance on shore support to reduce the overall life-cycle costs of the platform. There is a simple logic to this: Fewer personnel equals fewer habitability requirements — and hence greater room for combat systems — as well as reduced pay and fewer replenishments at sea.
But removing personnel from ships also means something else: more complex technical systems. Any unmanned piece of equipment is usually more complex than a manned machine assigned to do an equivalent task. This can make routine and corrective maintenance more costly, and also carries hidden costs by way of the high-level training required to conduct maintenance. The LCS was a case in point, as minimal manning drove procurement of systems that required technical expertise from the original equipment manufacturers.
With time, the projected life-cycle cost savings slowly began to diminish as core crew numbers increased by 25 percent, and the number of required shore-support personnel tripled. A similar dynamic is already playing out with early-stage planning for unmanned vessels. But even assuming that these are temporary speed bumps on the way to more mature technology, we can envision other areas where the deficiencies highlighted in the LCS’ minimal-manning model are unlikely to be alleviated by the zero-manning model.
In concept, the LCS manning model envisioned the transition of sailors, and engineers in particular, from a “maintainer-operator” role to that of “operator only.” However, the early LCS hulls faced many equipment casualties at sea, demonstrating the continued need for underway maintenance. Breaking the operator-only mindset proved difficult, as the original equipment manufacturers preserved their monopoly on expertise by keeping a close hold on the intellectual property — in the form of technical manuals, parts, and tools — necessary for maintenance and training. In other words, reducing maintenance at the operator level simply increased it at the intermediate and original manufacturer levels. Manning was not reduced, just transposed.
With time, an informal alternative to reliance on contractor support has emerged. Through direct experience operating their equipment while underway, LCS sailors have developed “tribal knowledge” of their systems. They have also acquired onsite knowledge by observing contractors and regional maintenance center engineers. As sailors transition to shore tours at regional maintenance facilities and training groups, designing programs to train the next generation of LCS sailors, the Navy achieves some self-sufficiency, an experiential economy of scale that can help recoup the costs of overreliance on original equipment manufacturers and contractors.
Yet it is difficult to see how this optimistic scenario could occur with fully unmanned platforms. First, with no sailors aboard, the underway experimentation and practice that produced tribal knowledge in the LCS case can’t happen. Nor will sailors be present to observe and learn from contractors who repair equipment. Without the economy of scale that began developing in the LCS case, maintenance costs will remain beholden to third-party contractors.
Second, while contractors can fly out to a manned platform that is underway, they cannot do so for an unmanned vessel. Without accommodations and life-support systems, unmanned vessels will have to return to port for repairs, or else be sustained at sea and in theater by amphibious ships, submarines, or expeditionary sea bases. The Navy already has some experience using manned platforms, including the LCS, to operate and maintain unmanned ones, such as the MQ-8C “Fire Scout” helicopter.
But small unmanned platforms controlled by sailors for brief periods of time, and then returned to a manned ship, are not entirely analogous to larger unmanned surface vessels. One difference arises from scaling: The more unmanned platforms in the fleet, the greater the maintenance burden that manned vessels will bear, drawing them away from other tasks. Another is the autonomy and multi-mission capability expected of future unmanned vessels. The Fire Scout is designed to accomplish a narrow mission and return “home” after a few hours, but most future unmanned vessels won’t be. Therefore, the Navy may have to tolerate some level of “minor” equipment casualties on underway unmanned vessels. The drawback is that, left unattended, minor casualties often become large ones.
Unmanned Ships Are Less Resilient
The higher production and life-cycle costs of the LCS led to fewer platforms purchased, and therefore insufficient quantity for distributed maritime operations. But assuming the Navy solves those problems and fields as many unmanned vessels as it thinks necessary, the next question is whether that quantity could be sustained in non-routine — i.e., combat — scenarios.
The current plan for the employment of unmanned vessels calls for them to team with manned units, thus expanding the fleet’s reach. The fleet, by design, will be a hybrid. Unmanned units will track or engage an enemy before it can harm high-value manned units such as aircraft carriers. This places a higher premium on the survivability of unmanned vessels than is often assumed. Unless some portion of the unmanned pickets survive an enemy’s first salvo, only nominal benefits will accrue from this hybrid fleet model.
The LCS, for its part, has a famously poor record of resilience, an ongoing problem attributable in part to minimal manning. With fewer sailors onboard to inspect engineering spaces, for example, one hull type suffered repeated engine casualties that simply went unnoticed. Another particularly salient example was the USS Fort Worth’s catastrophic engine failure in 2016, which an investigation attributed to insufficient oversight over watchstanders, reliance on personnel to perform a task they were not officially assigned to and briefed on (such as starting equipment), and the ship’s leadership’s absence from supporting roles due to their focus on a separate engineering casualty. In endorsing the investigation, the commander of the Pacific fleet specifically highlighted the impact that the smaller crew numbers played in the casualty. This was three years after the Navy had already adjusted LCS manning requirements due to crew fatigue and watchstanding shortfalls.
The primary cause of the Fort Worth casualty was operator error, but the Navy has attributed most other LCS engineering failures to design flaws or installation mistakes. In these cases, the problems could not be fixed at sea by design: the LCS operational construct explicitly backstopped minimal-manning with the promise of shore-based contractor support for both preventive and corrective maintenance. As the late Sen. John McCain pointed out, casualties from poor design or installation would disrupt operations far less if the sailors could fix the problem while underway. And the LCS’ survivability testing, which measures a ship and her crew’s ability to “avoid, withstand, and recover from” damage sustained by rough seas or combat, suggests that its minimal crew size would inhibit damage control efforts.
Assuming that unmanned systems are not flawless or impermeable to combat damage, they will, like the LCS, employ equipment redundancies to make them more likely to survive routine equipment failures and combat damage. At first glance, the concept of having backup systems appears to work the same way on manned and unmanned vessels. The latter might simply automate the switch between primary and secondary systems, instead of a human operator engaging the backup.
But there are, in fact, two key differences in how redundancy works on manned versus unmanned ships. First, for simple tasks, a sailor represents the redundancy of last resort. He or she can manually accomplish the machine’s function after the final backup fails. After an automated fire suppression system fails, a sailor can use a fire extinguisher. If a crew-served weapon’s electronic controls are damaged, sailors can aim and fire it.
Second, sailors can repair equipment. A hypothetical perfectly designed system should be more resilient than a human-operated version, because machines do not make mistakes or get tired. But in the military world, an intentional actor (the enemy) can tamper with the equipment. Since the enemy has an incentive to try again each time a redundancy restores the target’s operability, the system will eventually run out of redundancies. Although many shipboard systems are too complex to repair even if fully manned, the human on-station can restore redundancies for some simple but highly critical ones. Something as simple as patching a leaky fire main, for instance, might represent repairable damage for a manned platform, but catastrophic loss for an unmanned one.
Can the Unmanned Fleet Be Saved?
The minimal-manning construct of the LCS undermined its utility for distributed maritime operations in two ways. First, removing humans from the ship placed higher demands on contractor support. This drove up production and life-cycle costs, driving down the quantity of platforms that could be purchased. Second, the platform’s minimal manning made it less resilient to routine wear and tear, and consequently, the Navy both decommissioned four LCS hulls early and had to withdraw others from routine operations repeatedly to conduct repairs. We conclude with three recommendations to help future unmanned surface vessels avoid a similar fate.
First, unmanned system development requires a different approach to project management than was used for the LCS. The LCS’s unique integration of software, network hardware, and mechanical systems — particularly in the engineering plant — resulted in many growing pains. Design flaws found their way into multiple hulls because the program relied on what’s known as a “waterfall” project management approach. This method is characterized by a detailed long-term project plan with a single timeline, which makes changes costly.
But unlike with the LCS, where adding personnel to the original manning concept helped resolve failed integration points, fully unmanned platforms will lack this backstop. As a result, there is an even higher premium on ensuring that the integration points of the ship’s networks and mechanical systems function properly before widespread fielding. Agile project management, a development style based on shorter timelines and multiple delivery dates, might help address the issue. The Navy’s program executive office, Integrated Warfare Systems, is currently working to incorporate agile continuous delivery processes. In this approach, the product timeline is less definitive, changes to the product are frequent and expected, and the end user helps guide each iteration. The shipbuilding version of this would include the use of land-based testing sites, as it will for the new Constellation-class frigate.
Second, even with perfect equipment, unmanned vessels will face attacks with a redundancy chain that is always one link shorter than it would be with sailors present. This does not mean that unmanned vessels are worthless. Indeed, some types of equipment may be too complicated for any sailor to either operate manually or fix on scene. For a vessel with a narrow role performed exclusively by complex equipment, manning might indeed incur greater costs than it is worth. But that itself is a lesson: With a distributed fleet architecture, the Navy should only use unmanned vessels for those mission areas where the ability to survive the first few salvos matters little to the extended fight.
Third, while purchasing and fielding a great number of vessels is necessary for distributed maritime operations, so is preventing them all from being sunk outright. Unmanned vessels should not be considered expendable if they are expected to provide quantity, so some proportion of them will have to be repaired in combat conditions. While in many cases (for instance, the loss of a rudder or radar array), a mission kill is inevitable whether a ship is manned or not, sailors can prevent a manned ship from sinking long enough for an ocean-going tug to save the hull and untouched equipment. On unmanned ships, the timeline for salvage will be much shorter. This suggests that, if future fleet architecture depends heavily on unmanned vessels, the Navy will eventually bear the costs of more manned support vessels as well.
The LCS program promised to do more and cost less. It ended up doing less and costing more, precisely because its operational premise did not appreciate the human element enough. Sailors are an organic part of a ship, not a cost burden. Until equipment can learn on its own, and repair itself, that will remain the case.
Jonathan Panter is a Ph.D. candidate in the Department of Political Science at Columbia University. His research examines the origin of naval organizational practices. Prior to attending Columbia, he served as a surface warfare officer in the U.S. Navy.
Johnathan Falcone is an active-duty surface warfare officer in the U.S. Navy, serving as chief engineer aboard a littoral combat ship. He is a graduate of Princeton University’s School of Public and International Affairs and Yale University.
The authors’ opinions are their own and do not reflect the official stance of the U.S. Navy.