Improving Naval Ship Acquisition

This is a policy proposal written by Austin Vernon and myself for improving US naval shipbuilding. It was produced for Rebuilding.tech, a techno-industrial policy playbook curated by IFP, FAI, and American Compass.

USS Iowa, one of the many WWII-era ships designed in-house by the Navy.

US Naval vessels today regularly take far longer to build than scheduled, and greatly exceed their already-high cost estimates. A 2018 GAO report found that more than 80% of both lead ships (the first ship built of a series) and follow-on ships (subsequent ships of the series) were over budget, sometimes dramatically so. The first Zumwalt-class destroyer was over budget by 38%, and the first two Littoral Combat Ships were over budget by 150%.

Similarly, of eight lead ships reviewed by the GAO, every one was delivered behind schedule, and five of the eight ships were delivered two years late or more. Delays and cost overruns are wasteful and hamper the effectiveness of the Navy: reducing them would mean getting more ships, faster.

Delays and cost overruns are in part driven by the Navy’s ship design and acquisition process. Broadly speaking, the Navy creates high-level requirements for complex, multi-role ships, and then outsources the design of these ships to third-party contractors. Once a design is selected, it is turned into production drawings (so-called “detail designs”), which are used to produce the ships. In an effort to reduce the time it takes to deliver a ship, ship construction is often started before ship design is complete. However, this strategy frequently backfires: as design work is completed, changes to under-construction ships are often required, resulting in costly and time-consuming rework.

We recommend several changes to this process to reduce cost and schedule growth, and to reliably deliver ships faster and for less cost.

1. Instead of complex, multi-role ships which have expensive and often unnecessary features, the Navy should focus on simpler ships with narrower use cases.

2. Rather than outsourcing design to third parties, ship design should be brought in-house, and NAVSEA should expand its staff of Naval Architects from around 300 to closer to 1200.

3. Production on ships should not begin until design is substantially completed.

Simple ships

Many of the US Navy’s recent ship designs are large, complex multi-role ships. The Navy expects the same ship to hunt pirates, counter ballistic missiles, track submarines, and more. Also common are high-end features unnecessary for a ship’s mission. These complex ships have many negative consequences on the ability to design ships, increase production throughput, and meet budget and schedule targets. A return to simpler ships would help the Navy deliver more ships on time and within budget.

A few examples of scope bloat:

1. Ford-class carriers have high-end radars with similar capabilities as the radars of guided-missile destroyers. Yet they do not have long range surface-to-air missiles, they are surrounded by escorts with similar capabilities, and the emissions from these radars make them easier to detect, track, and target. The cost of the systems is substantial (likely more than $1 billion after including support systems), and they add design complexity and construction time.

2. Recent Burke-class destroyers have extensive helicopter facilities, despite their main role being vehicles for offensive and defensive missiles. The helicopters add significant cost, weight, and crew to the ship. The Flight III ships, first launched in 2023, also have a much larger radar jammed into the same hull.

3. The Littoral Combat Ship (LCS) was supposed to be a jack of all trades, filling roles like patrol or minesweeping by swapping out modules. But LCS ships are undergoing rapid early retirement because the attempt to fill so many roles compromised many attributes, and designing modules proved unworkable. Even worse, many specialist ship roles, like minesweeping, are now being retired without replacements because the LCS failed to fill the role.

4. Frigates are typically smaller and less heavily armed than more capable and more expensive destroyers. But the new Constellation-class frigate will cost nearly as much as the Burke-class destroyer, because of its extensive radar and missile capacity. The bloated design missed an opportunity to complement the Burke by being strong where the Burke is weaker, like anti-submarine warfare.

The extra features raise cost, hurt producibility, and can even compromise ship effectiveness.

A likely cause for this scope bloat is the adherence to the “Distributed Maritime Operations” doctrine, or DMO. The logic is that a traditional battlegroup is too vulnerable to massed missile attack, so the Navy prioritizes multi-function ships which can be widely distributed (making them harder to target), but still coordinate their attacks with advanced networking capabilities

Simpler ships would still fit within DMO as 2 to 3-ship task groups. And reduced design overhead and smaller, more narrowly focused ship designs would also allow for more ships to be built. Building large numbers of simpler ships would, in turn, help reduce US ship cost by taking advantage of learning curve effects.

The Navy has had more success in the past with more focused ship design. Two small examples are:

1. The Perry-class frigate was an affordable ship designed for escort duties, primarily providing anti-submarine and local anti-air protection, eschewing other features. The ship proved to be popular: it could successfully perform its escort duties, it was tough, and the low acquisition and operational cost allowed a large number of them to be acquired. The Constellation-class could have filled this role, but was packed with too many features.

2. The T-AGOS-class is a highly specialized surveillance ship to track enemy submarines from long distances. The defining feature of the ship is its small waterplane area twin hull that provides stability for the sensors. The more stable hull significantly increased mission capability in rough seas.

A more conservative, executable plan for the Navy might look like:

• A modern, pared down escort frigate without the bells and whistles — Air defense is very different from when the US was last designing frigates, making previous designs obsolete. Part of the issue with Constellation-class is that the Navy specified a very capable AEGIS radar system and a large number of larger missile launch tubes, increasing cost while duplicating capability available from other ship classes. An alternative more appropriate for an escort would be a modest radar with less missile tubes and the use of the Evolved Sea Sparrow air defense missiles, of which four can fit in a vertical launch tube. Since radars need 16x more power to double their range, the radar appropriate for ESSM and local air protection would be a fraction of the size and ship power in comparison. Such a ship should be given a robust anti-submarine warfare capability similar to what is planned for the Constellation, especially as diesel electric submarines become more capable and China increases its nuclear attack submarine fleet.

• A guided missile destroyer without helicopters or upsized ballistic missile capable radars and only limited sonar — In some sense, the Burke-class has become a victim of its own success. The initial designs were very successful as a guided-missile destroyer. The inability of the Navy to develop new ship classes has pushed more and more roles onto the Burkes. That leaves the opportunity for a refactored, narrowly focused guided-missile destroyer that uses modern radars matched to SM-2 and SM-6 surface to air missile range, carries more missiles by carrying the ESSM as part of its loadout, and deletes helicopter aviation. The result would be a ship that is much more capable than the original Burke-class in a smaller form factor and could take advantage of other advances in technology like diesel-electric power systems, allowing a less complex power train.

• Surface warfare focused destroyer — A more speculative option is a destroyer focused on surface warfare such as attacking ships and land targets. Navy surface ships have become defensive with offensive power moving to aircraft and submarines. A destroyer with an 8 inch gun, offensive missiles, long-range torpedoes, etc. could add some punch, but this isn’t a role the Navy currently fills after the Zumwalt-class failure.

• Ford-class carriers pared down to focus on flight operations, or a conventionally powered (non-nuclear), stripped down carrier — To some extent the Navy is already trying to simplify follow-on ships of the Ford-class, but much more can be done, such as deleting much of the centralized information processing capabilities. Another target might be excessive automation. Many of the features have high development costs and could be pared back with slight increases in crew to simplify the design. And while nuclear power provides additional carrier endurance and capabilities, conventional fuel ships are ultimately much easier to design and build than nuclear ones. The Navy operated oil-fueled carriers until the mid 2000s. A simpler to build, oil-powered carrier that focuses purely on flight operations rather than radars, flag accommodations, or hypothetical future electricity use cases such as directed energy weapons would allow the Navy to field and maintain more carriers.

• A specialized ballistic missile defense platform based on a commercial ship hull — The US military has historically preferred to intercept ballistic missiles outside the atmosphere. The advantage is that one missile defense battery can cover a very wide area. A specialized ballistic missile defense ship could be kept farther back from more forward groups, protecting them without giving away its position with easily detectable radar emissions.

• A drone carrier based on a commercial ship hull — Drones are evolving rapidly, but don’t integrate well with the Navy’s current ships. A smaller, relatively simple and flexible ship to act as a drone carrier could fulfill the role of a modern escort carrier.

• A drone-based minesweeper platform — Minesweeping is extremely dangerous and the ships are quite complex to reduce their vulnerability. Modern mines are exceptionally challenging to deal with because they have more sophisticated trigger algorithms. One possible solution is to use a small sea drone that carries equipment that can imitate the sound and magnetic signal of specific ships likely to be in the mine’s database. Without humans, the size and cost of the ship can be much smaller, obviating the need for fancy hull design and construction.

• Pared-down LSTs and troop ships — Roughly 40 years ago, the Marines changed doctrine to primarily assault via helicopters and hovercraft rather than traditional landing ships. Amphibious assault ships became mini-aircraft carriers, greatly increasing their cost and complexity. Rotor-lift is now extremely vulnerable to air defenses, and hovercraft aren’t much better. The viability of >$50 billion worth of assets supporting this mission is questionable. Traditional troop ships and landing ship tanks are a fraction of the cost, more reliable, and scale better — rotor-lift based transport can move a relatively small number of troops, and would thus have limited use in a great power struggle. The current rotor-lift-based amphibious force can be seen as a boutique capability that can access more places compared to traditional troop ships.

• More Logistics Ships — The Navy has struggled to take advantage of how cheap modern ships can be to build a robust and reliable fleet of logistics ships for supplying fuel, ammunition, and other supplies at low cost, especially when these types of vessels would be critical for any extended action in East Asia.

Each of these ships would be more affordable, could be produced in larger flights, and would still maintain all the capabilities the fleet requires. Obviously these can’t all be built overnight, but each one is a potential step towards the Navy adopting simpler, more focused ship designs. They would be easier, faster, and cheaper to design and build. The hull count and overall capability of the Navy could rise without a massive increase in resources.

In-house design

Historically, the Navy performed most of its ship design in-house. Up until the 1960s, the designs for naval ships were produced by various Naval departments such as the Bureau of Ships. Ship design was briefly outsourced in the late 1960s and early 1970s under Robert McNamara’s Total Package Procurement (TPP) process, only to be brought back in-house (now under the auspices of Naval Sea Systems Command, or NAVSEA) in the 1970s due to dissatisfaction with TPP.

However, following the end of the Cold War, Naval ship design capabilities were hollowed out following so-called “Acquisition Reform.” During the 1990s the number of naval architects employed by NAVSEA fell from around 1,200 to less than 300, even as ships became more and more complex. Rather than doing in-house design, design is now done by third-party contractors, based on high-level requirements provided by the Navy. For example, the Littoral Combat Ship was designed by Lockheed Martin and General Dynamics, and the Constellation-class frigate was designed by Fincantieri Marinette Marine.

Outsourcing the design of ships creates several problems. Shipbuilders often can’t afford to keep large design teams employed full time due to the lack of a substantial US shipbuilding market, the naturally peaky nature of demand for naval ship design services, and the uncertainty as to whether a shipbuilder will win a given contract. During TPP, for instance, shipbuilders that didn’t win shipbuilding contracts quickly dismantled their design teams, and the shipbuilders that did let theirs atrophy once design was complete and the ships had moved into construction.

Similar concerns were at work with the Littoral Combat Ship. The Navy originally planned to build one ship based on Lockheed’s design, and another based on General Dynamics. These “Flight 0” ships would be tested and a winner selected, which would continue production as “Flight 1” ships. But the contractors complained that it would be too expensive to maintain their design teams and production lines between initial ship construction and choosing a winner, and so two additional Flight 0 ships were authorized to help bridge the gap.

Effective ship design requires a sufficient staff of skilled and experienced marine engineers and naval architects to design the ships. When these teams aren’t maintained, knowledge and expertise is lost and can be difficult to recover. This likely contributes to long design times and often inadequate designs produced for US naval vessels. The nature of naval ship design makes it difficult for these teams to be maintained across several different shipbuilders, and having the Navy perform this work is an obvious solution.

It's also common for ship requirements to evolve as design proceeds, as the cost impacts and tradeoffs of various capabilities become better understood. Naval vessels are complex, and their performance often resists simple quantification: design rarely proceeds smoothly from a simple set of specifications. When the Navy has a robust design department, performance and cost tradeoffs can be more easily evaluated and requirements modified as design proceeds. When a contractual barrier separates the designer and the requirements setter, this sort of modification is more difficult. More generally, having a large team of design experts makes the Navy a better informed customer, and less susceptible to principal–agent problems where the ship designer and builder acts in its best interests, rather than the interests of the Navy.

For example, the Littoral Combat Ship initial requirements specified a high maximum speed and cruising range, requiring a complex propulsion system which used both diesel and turbine engines joined by a combining gear to let them operate simultaneously. This propulsion system was difficult, expensive, and time-consuming to develop, and ultimately didn’t meet the specified speed and range requirements. These speed requirements were eventually determined to be unnecessarily high: modifying the requirements early on as the tradeoffs of speed, cost, and complexity became apparent could have avoided this problem.

By contrast, the FFG 7 Oliver Perry-class frigate, designed in-house by the Navy, was much more successful in navigating the complex tradeoffs inherent in Naval vessel design, as Robert Keane, former Chief Naval Architect of NAVSEA, has noted:

The weight constraint was directed in an attempt to control acquisition cost, and it forced compromises to be made in major ship characteristics. This included having only a single propulsion shaft which limited speed and had survivability implications. The 76-mm main gun was too small to provide naval gunfire support to troops ashore, but this was not a requirement. In order to accommodate the LAMPS III helicopter, only a rather mediocre hull mounted sonar could be installed (the towed array sonar compensated). A highly capable — for its time — AAW system was provided. Active control fins were installed which resulted in seakeeping performance equal to that of much larger ships. The fleet’s initial negative reactions were soon overcome, and the ships proved very popular.

Keane goes on to say that “the tradeoffs which were made during the FFG 7 early design stages would have been difficult, if not impossible, in an industry-led design or during a design competition.”

We see the importance of having substantial in-house design expertise for large, unique or semi-unique government funded projects in other domains. The Transit Costs Project diagnoses the exceptionally high costs of American mass transit construction as partially due to a lack of in-house expertise to design and manage the projects. By contrast, Spain has some of the lowest transit construction costs in the world, and all its transit design is done in-house by the government. A similar lack of in-house expertise and project management capacity has been blamed for California’s High-Speed Rail debacle.

Conversely, the success of the Apollo Program in the 1960s can be partly attributed to NASA’s robust in-house aerospace expertise (at the peak of the Apollo Program, NASA employed around 34,000 people). The initial designs for the Saturn V rocket and the Apollo spacecraft were both produced by NASA. NASA closely monitored its subcontractors performance and intervened when necessary.

Project schedule

With any large, complex project, a key method of cost control is to have the design as defined as possible before construction begins. When the design is still on paper, it’s relatively easy to make changes. Once a paper design gets turned into steel, equipment, and machinery, making changes is much harder and more expensive. The previous GAO report notes that achieving substantial design completion before beginning ship construction is a standard best practice in commercial shipbuilding. And we see the risks of making design changes during a construction process in other domains. Constant design changes driven by an evolving regulatory landscape are part of the reason for the precipitous rise in nuclear power plant construction costs in the US. More broadly, in his book on how to manage large projects effectively, megaproject expert Bent Flyvbjerg notes a key rule is to have the design and planning as complete as possible before beginning, to reduce the likelihood that problems and delays will occur.

Abraham Lincoln is reputed to have said that if he had five minutes to chop down a tree, he'd spend the first three sharpening the ax. That’s exactly the right approach for big projects: Put enormous care and effort into planning to ensure that delivery is smooth and swift.

Think slow, act fast: That’s the secret of success…

On project after project, rushed, superficial planning is followed by a quick start that makes everybody happy because shovels are in the ground. But inevitably, the project crashes into problems that were overlooked or not seriously analyzed and dealt with in planning. People run around trying to fix things. More stuff breaks. There is more running around. I call this the “break-fix cycle.” A project that enters it is like a mammoth stuck in a tar pit.

US naval shipbuilders routinely violate this tenet. The GAO report mentioned above noted that for most lead ships, construction was in many cases concurrent with design. In an unfortunate number of cases, technology development, which should be prior to design work, took place concurrently as well. The ostensible purpose of this concurrence is to get ships built more quickly, but it often results in severe schedule delays when redesigns and problems inevitably arise.

By not beginning construction until design has been substantially completed, production can be completed more smoothly, redesigns and problems reduced, and schedules and costs ultimately reduced.

Conclusion

These recommendations reinforce one another. Adopting simpler ship designs will reduce design time and make it easier to reach substantial completion before construction begins. Simpler ships that can be designed and built quicker are less likely to need revisions during construction due to changing technology or mission scope. A robust in-house design team will make it easier to properly weigh the importance of different features, balance cost, schedule, and capabilities, and generally make more informed decisions on how a ship should be designed to meet cost and schedule goals. It will also make it easier to produce designs quickly (because of quick feedback and knowledge of actual requirements) and interface with shipyards to address design problems.

By making these design and acquisition changes, the US can build more ships, more quickly, and for cheaper.