HII spreads amphibious shipbuilding beyond Pascagoula

HII’s Ingalls Shipbuilding division is extending distributed construction into the San Antonio-class amphibious transport dock programme, beginning with Philadelphia, LPD 32.

Eight structural units have been allocated to two partner shipyards during early production. Completed work will be transported to Ingalls in Pascagoula, Mississippi, where the ship will undergo larger-scale assembly, outfitting, systems integration, testing, and delivery.

The approach builds upon distributed work already used for Arleigh Burke-class destroyers. Structural units for Thad Cochran, DDG 135, were completed before the ship’s October 2025 keel-laying milestone, allowing Ingalls to bring externally manufactured work into the principal construction sequence.

HII doubled its distributed shipbuilding activity during the previous year and expects outsourced hours to increase by a further 30% during 2026. Partner capacity will handle selected structural work, while Ingalls preserves skilled labour, covered space, and integration resources for tasks that are more difficult to distribute.

Fleet demand continues to exceed the rate at which the existing US shipyard workforce and supplier network can deliver complex vessels. Constructing another complete naval yard would require substantial investment and many years, whereas qualified regional fabricators can absorb defined modules sooner.

Distributed construction shifts rather than removes risk

A steel unit can be built away from the final yard only when every participant works from the same controlled data. Plate dimensions, stiffeners, penetrations, foundations, weld sequences, and reference points must align closely enough for the module to fit its surrounding structure without extensive rework.

Dimensional control becomes increasingly important as units grow because small errors in early fabrication can accumulate across panels and decks. Laser measurement, digital surveying, controlled fixtures, and agreed acceptance tolerances are required before a module leaves the partner yard.

Transport requirements also influence fabrication. Units must fit barges or other transport systems, withstand movement and lifting, and arrive without distortion or corrosion. Temporary reinforcement may be required, while lifting points need to be incorporated during design rather than improvised after construction.

Outfitting modules before delivery can remove work from the congested main yard, but it increases integration risk. Pipes, cableways, ventilation, foundations, and equipment must use the correct materials, cleanliness standards, identification, and inspection procedures. A misplaced support may interfere with equipment installed several months later.

Quality records must travel with the hardware. Weld procedures, welder qualifications, material certificates, non-destructive inspection results, deviations, repairs, and dimensional surveys all need to enter HII’s configuration system. A physically complete module with incomplete records can still delay acceptance.

Partner yards can expand the workforce available to naval construction, recruiting and training welders, fitters, inspectors, planners, and supervisors outside Pascagoula. Ingalls can then direct scarce experienced labour towards combat-system spaces, machinery integration, final erection, trials, and delivery.

Distributed production also adds management overhead because engineering queries must be resolved across organisations, production changes communicated quickly, and supplier performance monitored remotely. Late design revisions become particularly expensive when a unit is already being fabricated hundreds of miles away.

Domestic module production sits alongside US efforts to assess Japanese and South Korean shipyard capacity. Overseas construction introduces greater legal, security, and political complexity, while HII’s model attempts to widen capacity without moving controlled shipbuilding work outside the United States.

LPD 32 is a demanding test because amphibious ships contain extensive accommodation, aviation, medical, vehicle, well-deck, and command facilities. Their internal systems are dense, and late rework can consume large numbers of production hours. Units allocated externally need interfaces stable enough to separate from the main build.

Partner-yard economics require a steady flow of work. Businesses may invest in halls, cranes, tooling, digital systems, and workforce for a limited package, yet a stop-start module schedule can leave them carrying fixed costs without sufficient throughput. Long-term planning is essential if distributed construction is to become a permanent capacity layer.

Ingalls must also prevent the main yard becoming a storage and rework facility. Modules need to arrive in the order required for erection; early delivery consumes waterfront and laydown space, while late delivery leaves skilled teams waiting.

Cybersecurity and data segmentation add another layer because external yards require detailed design information without necessarily needing access to the whole ship. Controlled digital environments must give suppliers enough information to manufacture accurately while protecting sensitive combat-system and configuration data.

The industrial model works best where mature, repetitive structural work can be separated cleanly, leaving Ingalls in control of the most complex integration and testing. It becomes less effective when designs remain unstable or modules contain systems requiring frequent intervention from the principal engineering team.

Philadelphia will show whether destroyer experience transfers efficiently to amphibious construction. The most useful measure will be total schedule and labour after transport, inspection, fit-up, and rework, rather than the number of hours moved away from Pascagoula.

American naval shipbuilding needs more capacity, but capacity distributed across several yards must still function as one production system. HII’s task is to make physical distance invisible when the units meet on the final assembly line.