The Drone War is Won with the Support of Replicator-3

Under the Arizona sun, the aircraft stacked at Davis–Monthan and the decommissioned ships on the James River remind us of a simple truth: hardware, without care, decays. The future of military force will be no exception: buying systems is the visible part, keeping them ready is the decisive one.

The Replicator program of the United States Department of Defense aims to deploy thousands of low-cost drones and autonomous systems very quickly.

• Replicator-1 mainly aims to have more offensive power: many economical systems that can attack the enemy in an unpredictable way.
• Replicator-2, on the other hand, is designed for defense: creating a credible “shield” against adversaries who will use similar drones and autonomous systems.

The next step should be Replicator-3: a revolution in sustainment, the design of a support network that keeps these forces truly deployable.

Today the priority is production, in response to the Chinese military growth, especially in the Indo-Pacific theater. But an “unmanned wave” risks being operationally empty if there is no plan for how to store, repair, power, and operate it for years. Data from the Government Accountability Office shows that even manned fleets struggle to remain ready. Thinking that thousands of complex drones, built quickly, can do so without a different structure is illusory.

The Limits of the Unmanned Myth and the Indo-Pacific Challenge

A widespread myth holds that unmanned systems have a shorter logistical tail. For long-range maritime platforms, rich in sensors, this is false. Maintainability can improve with modular components, predictive maintenance, and “attritable” design, but the multiplication of systems increases maintenance work at the operational and advanced levels. Technicians, equipment, and forward infrastructure are needed, which are currently undersized.

In Ukraine, small, inexpensive drones are seen, produced and lost in huge quantities. The Indo-Pacific is different:

• huge distances and hostile marine environment;
• dependence on allies and distant bases;
• large, energy-consuming, and complex systems, more similar to modern ships and aircraft.

These systems will need to withstand corrosion and long periods of inactivity, ensure secure long-range control, be modular, and based on clear maintenance rights for the government. Moreover, they will require much more manpower than often believed.

Private production – from companies like Anduril or Saronic – today focuses on assembly lines, not on the support network in the theater. But the systems built now could remain inactive for years before a conflict, and without constant care, they will not be ready. Trust between operators and platforms arises from real use, not just from the simulator: if not operated in peacetime, many systems will not work as expected in war.

What Replicator-3 Should Do and the Cost of Inaction

Replicator-3 should bridge the gap between production speed and sustainment capability, acting on three fronts:

1. Link sustainment to procurement
   • Define a minimum budget share for infrastructure, personnel, stocks, and advanced maintenance for each acquired capability.
   • Support funding included from the start, not added later.
2. Reward resilience, not just volumes
   • Incentives for operational availability, dual-source supply chains, quality of maintenance data.
   • Contracts that evaluate “uptime,” not just quantities produced.
3. Build a shared regional support network
   • Use funding and foreign military sales to create maintenance and storage nodes managed by partners in the Indo-Pacific.
   • Shared pools of systems that allow allied countries to contribute to deterrence without having to fight on the front line.

The sustainment network will need to be dispersed, resilient, partially automated, capable of operating with irregular supplies. Small modular reactors, off-grid energy generation, and additive manufacturing in the theater can power and supply distributed fleets without relying solely on large vulnerable bases.

Finally, dedicated expeditionary teams will be needed, military equivalents of the maintenance model of large commercial autonomous vehicle operators: small groups that, from austere bases, manage batteries, modules, software, and operational launches.

If this sustainment revolution does not occur, commanders will find themselves cannibalizing platforms, waiting months for spare parts, or drastically limiting employment. The result: an impressive arsenal on paper but fragile deterrence in reality. Replicator-3 must therefore be, above all, the logistical revolution that makes the era of unmanned systems credible.