Join War on the Rocks and gain access to content trusted by policymakers, military leaders, and strategic thinkers worldwide.
Under the Arizona sun, mothballed aircraft glisten in perfect rows at Davis-Monthan Air Force Base. In the brackish calm of the James River, reserve ships drift, their gray hulls streaked with rust. Both rest in quiet suspension, reminders that hardware decays without care. Future force design will share this truth: Hardware degrades if it is not sustained. Procurement is the flashy step, sustainment the sticky one.
Replicator is the Pentagon’s catalyst to deliver new capability at speed and scale. The Defense Department and steward, Defense Innovation Unit, have broken down Replicator into multiple lines of effort. Replicator is fielding thousands of affordable, autonomous systems for asymmetric offensive power. Replicator-2 is focused on a credible defensive layer as threats evolve to a similar force concept. Replicator-3 should be the next step, building the sustainment network that keeps the future force combat-credible. It should shepherd the design of a new sustainment architecture that resources and forward deploys the infrastructure, manpower, and foreign partnerships to keep pace with production.
A Production Revolution Without a Sustainment One
In response to China’s military expansion, the Pentagon is racing to field vast numbers of unmanned systems across the Pacific – a strategy that risks building an arsenal faster than it can be sustained. This is not a call to slow procurement; production is the essential first step. Replicator has focused rightly on speed, scale, and affordability, but less on how thousands of systems will be kept operational and tactically effective once fielded. The issue is not that the Department of Defense ignores sustainment, it’s that sustainment isn’t resourced or structured properly for the type of systems or speed we’re buying them.
Audits by the Government Accountability Office show the historical reality. From 2011 to 2021, Arleigh Burke-class destroyers averaged 25 days of maintenance delays per ship and 23 severe casualty reports, a complete loss of a mission area, per year. An additional Government Accountability Office report attributes these shortfalls to corrosion, component reliability, and continued shortfalls in depot capacity and workforce skills. The Defense Department spends billions on sustainment but lacks focus and structure. If manned fleets with onboard maintainers and mature infrastructure struggle to stay ready, an unmanned surge built in haste will be operationally hollow.
America’s deterrent is only as strong as the network that sustains it. We are fielding thousands of unmanned systems without a credible plan to store, service, or power them once deployed on the notion that systems are attritable. A Replicator-3 should fuse the industrial and logistical revolution, planning for sustainment as much as production.
The False Premise of a Smaller Tail
A persistent myth in defense debates holds that unmanned systems carry a smaller sustainment tail. They do not – at least not for the long-range, maritime, sensor-dense systems the Indo-Pacific will require. Whether scaling SM-6 shooter capable hulls or deploying hundreds of unmanned all-domain systems across the battlespace, sustainment demands multiply as firepower does. Adoption of condition-based maintenance practices, modular components, and attritable designs for smaller systems can reduce depot work. However smaller more numerous systems will increase organizational and intermediate level maintenance as they scale. Without the forward sustainment infrastructure, the United States will not capture the capability through capacity promised by modern technology.
Small unmanned systems may simplify some repairs, but the technicians, tooling, and forward infrastructure to execute them remain under-resourced and lack structure. While a focus on commercial components and software-defined design will reduce supply chain strain, platforms face the same mechanical decay realities as civilian aircraft and boats – their batteries die, spark plugs foul, lubricants degrade, and components corrode. An unmanned augmented force is strategically sound, but its sustainment architecture remains underdeveloped and unfit for scale.
The services recognize this even if public discussion lags. The Navy has formed squadrons for surface and sub-surface drones and a career field for robotic warfare specialists. The Air Force is establishing collaborative combat aircraft squadrons. Those organizational moves quietly admit what defense discourse rarely includes: unmanned does not mean unsustained.
The Indo-Pacific Is Not Ukraine
Ukraine shows that unmanned systems can be produced and expended in huge quantities. While small quadcopters are the next evolutionary step of the recoilless rifle, a future Indo-Pacific conflict presents a different geopolitical and physical environment – an away game dependent on allied access; defined by vast distances, salt, humidity, and maritime logistics; and a region not yet in active conflict.
The systems fit for the Pacific are not disposable quadcopters. They are large, power-hungry, sensor-laden machines, modernized and modularized extensions of today’s warships and aircraft. Shedding hotel loads and life-support systems reduces unit cost, but not complexity in aggregate. That complexity drives unique design and sustainment demands that diverge sharply from Ukraine’s attrition model and drawing direct corollaries is a mistake.
Unmanned systems that aren’t hardened against corrosion and capable of long idle endurance will not be successful in the Indo-Pacific. They should also be able to maintain secure long-range control, use modular architectures, and be built on open sustainment rights for government upkeep. And let’s not forget manpower and forward-deployed infrastructure. Manning requirements for unmanned systems are significant, which might seem counterintuitive to some observers. Without those features, attritable mass becomes disposable fragility.
Building for Production, Not Readiness
Defense unicorns like Anduril and Saronic are investing heavily in production infrastructure stateside, reindustrialization vital to future defense challenges. However, that capital is overwhelmingly tied to manufacture, not sustainment. The forward support network of storage, equipment, tooling, and technicians required to keep distributed fleets operational in theater is not being co-invested at the necessary scale.
Advocates often invoke liberty ships and B-24 bombers as analogs for mass production of modern unmanned craft. The comparison flatters our nostalgia. Those programs surged after heavy wartime attrition; we have not yet reached that point. The systems we build today we will have to sustain tomorrow. If conflict comes later than expected, many could sit idle for years, demanding disciplined upkeep to remain combat-ready.
Some argue virtual training can preserve readiness, but trust between operator and machine is forged only in real operations. Faults in heterogeneous hardware/software integration and basic administrative friction appears only through use, not simulation. These systems will require practiced precision by their operators and maintainers. If crews never fly or sail these systems in peacetime, they will not trust them in crisis and history proves some will not work at all in conflict.
What Should Change
The trend towards cheaper, smarter, and numerous unmanned systems has the necessary momentum to succeed. However, leadership within the Defense Department, Congress, and industry need to focus on the sustainment tail.
The future sustainment network should not resemble a conventional supply base nor current maintenance infrastructure. It should be adaptive, dispersed, hardened, semi-autonomous, built to operate under persistent threat, and limited resupply. Nodes across the region should support storage, forward maintenance, and rapid iteration. Infrastructure cannot rely solely on access to established bases or Military Sealift Command’s aging fleet. The Defense Department should invest in small modular reactors such as those developed by Radiant to power regional maintenance facilities and additive manufacturing nodes like those offered by Rangeview to produce replacement parts in theater. Further forward, systems like Edge’s Moonshine Hydrogen, which produces hydrogen energy from aluminum and seawater, offer off-grid energy for small units and can sustain power, connectivity, and conditioning requirements for distributed containerized systems and munitions.
Automation doesn’t erase labor – it redistributes it. Even Waymo’s autonomous vehicle fleet employs nearly 1,700 engineers and technicians to sustain roughly 1,500 cars. The military needs its own version of that model: small, expeditionary teams that link logistics to combat power by swapping components, servicing batteries, refueling, patching software, and launching systems from austere environments. The Navy’s new Robotic Warfare Specialist is a start, but each service will need its own distributed maintainer force.
Replicator-3 should close the gap between production speed and forward sustainment capacity. That requires three parallel efforts. First, quantify the sustainment floor by tying a minimum sustainment ratio to procurement budgets so forward infrastructure and manpower are funded from the start. Second, tie industry incentives to resilience, not output, rewarding contractors for uptime, dual-source supply chains, and verified sustainment data. Third, use foreign military financing and sales to build regional sustainment capacity as a collective defense function, funding partner-led nodes and shared system pools to anchor distributed capability – one that lets smaller nations visibly contribute without firing a shot.
The Strategic Cost of Inaction
A numerically impressive but poorly sustained unmanned force will weaken deterrence. Replicator-3 should take on the harder work: build the sustainment foundation that keeps the future force enduring.
Otherwise, commanders will face the same wrenching choices Government Accountability Office research describes: cannibalize, wait for repairs, or ration capability. Pacific systems will never be cheap enough to discard en masse; they should be ready and reliable. Modernization succeeds not only by buying inventory but by guaranteeing readiness.
Sustainment problems are solvable if we design, fund, and structure their requirements now. Procurement will always be the headline, but sustainment should be the organ. If the Defense Department is serious about unmanned mass as a core pillar of deterrence in the Indo-Pacific, Replicator-3 should be the sustainment revolution – a deliberate effort to design, resource, and forward-deploy the infrastructure, manpower, and partnerships to match the pace of production. Otherwise, we will have a spectacular arsenal on paper and a fragile deterrent in reality.
Nick Johnson is an officer in the U.S. Navy currently serving as the fires and effects officer for Naval Special Warfare Group ONE. He is a TOPGUN graduate, former strike fighter tactics instructor, and F/A-18 Super Hornet pilot. He is a member of the Dark Trident Research Group, which advises the Office of the Chief of Naval Operations. He previously served as a defense ventures fellow with Anduril Industries, where he focused on autonomous system behavior and mission autonomy.
The views in this article are those of the author’s and do not represent the policies or positions of the Dark Trident Research Group, the U.S. Navy, the Department of Defense, or any part of the U.S. government.
**Please note, as a matter of house style War on the Rocks will not use a different name for the U.S. Department of Defense unless and until the name is changed by statute by the U.S. Congress.
Image: Petty Officer 1st Class Jasmine Mieszala via DVIDS
