Building a domestic drone supply chain. KDE. CubePilot. ARK. MaxAmps.

For ten years the consumer-drone supply chain was effectively a single company. DJI's vertical integration across motors, ESCs, flight controllers, radio links, gimbals, and lithium-polymer cells set the price floor and the performance ceiling for the entire small-UAS market.¹ A non-Chinese OEM that wanted to build a Group 1 or Group 2 platform could either pay the DJI premium or build to a worse spec at a higher unit cost. Almost nobody in the West was going to win that argument on merit.

The arithmetic changed in 2024. The American Security Drone Act, §848 of the FY2024 NDAA, and the expansion of the §1260H list collectively closed the DJI procurement door for the U.S. federal customer and most state-and-local grant recipients. The vertical-integration advantage is now a vertical-integration liability. A new supplier mesh has been growing in its place — not centralised, not perfectly efficient, but real and increasingly producible. This piece names the five first-tier suppliers that, together, make a §848-compliant combat drone economically viable in 2026, and describes the trade-off each represents.

[01] · THE BOM PROBLEMWhat 'compliant' actually means at the component layer

A typical Group 1 or small Group 2 platform — quadrotor or fixed-wing, two to fifty pounds gross weight — has approximately 150 to 250 line items on its production bill of materials. Roughly fifteen of those line items account for most of the unit cost and almost all of the compliance risk:

Compliance fails component-by-component, not platform-by-platform. The integrator's task is to anchor each of these blocks to a supplier whose §889/§848/§1260H representations are documented and whose sub-tier supply chain can be audited. The five companies named in the title are the ones that, in 2026, do that work for the highest-volume Group 1 and small Group 2 design points.

[02] · MOTORS AND ESCsKDE Direct, T-Motor (US), and the brushless-motor question

KDE Direct (Daytona Beach, Florida) is a U.S.-owned and U.S.-manufactured brushless motor and ESC supplier whose product line spans the size envelope from compact Group 1 quadrotors to medium Group 2 fixed-wing. Their motors are engineered for industrial and military duty cycles — substantially longer life and lower thermal failure rate under sustained high-current operation than the consumer-grade alternatives — and the company has been an explicit supplier into Blue UAS Cleared List airframes for several years.²

The trade-off is unit cost. A KDE motor-and-ESC pair runs roughly two-and-a-half to four times the unit cost of an equivalent T-Motor or Brotherhobby part sourced from the open Chinese supply chain. For a quadrotor counter-UAS interceptor at the $20,000 unit-cost target we described in the cost-per-kill analysis, the motor-and-ESC BOM contribution is the single largest controllable cost line in 2026, and KDE is the dominant variable in it.

The plausible second source is the U.S. manufacturing line that T-Motor has stood up — sometimes marketed as T-Motor US or under partner brands — which is intended to provide a Blue-aligned motor option at a lower price point. The compliance posture of that line is a function of which production phase the parts come from and whose §848 representation accompanies the shipment; integrators should not assume parity with KDE without a signed BOM representation.

[03] · FLIGHT CONTROLLER / AUTOPILOTCubePilot and ARK Electronics

The autopilot is the most architecturally consequential component on the platform. It carries the navigation filter, the control loops, the failsafe state machine, and increasingly the gateway to a separate companion computer running the autonomy stack. Two suppliers anchor the §848-compliant autopilot space in 2026.

CubePilot (Australia, Hex/ProfiCNC) ships the Cube Orange, Cube Black, and the export-restricted Cube Blue variants, which run on the Pixhawk standard reference design and support both ArduPilot and PX4 firmware.³ Cube Blue is the U.S.-defense-market SKU with a Blue UAS posture and BOM disclosure designed for §889 audit. CubePilot is not U.S.-domiciled — Australia is a Five Eyes partner, and §848 treats allied origin permissively where the statute permits — but the integrator should confirm the specific representation in the shipment's compliance documentation rather than relying on the family name.

ARK Electronics (Madison, Wisconsin) is the more recent U.S.-domiciled entrant in the Pixhawk-compatible flight-controller market. ARK has shipped open-architecture flight controllers (the ARKV6X and ARKV6X-RT families), GNSS receivers, jetson-class carriers, and ancillary modules sized to compete directly with CubePilot on performance while offering a fully U.S.-pedigree compliance posture.⁴ ARK is the cleanest single-vendor answer to the "where is the autopilot from" question on a §848 audit.

The trade-off here is breadth versus depth. CubePilot has the larger installed base, a deeper third-party ecosystem, and longer-running firmware integration. ARK is U.S.-domiciled, ships against an aggressive product cadence, and is the better answer when the procurement file needs a single-paragraph supply-chain story. A serious integrator may design to support both, switching by SKU.

[04] · BATTERIESMaxAmps, Bonka, and the energy-density gap

MaxAmps (Spokane, Washington) is the largest U.S.-manufactured lithium-polymer cell and pack supplier for industrial and defense UAS.⁵ Their cell chemistries are tuned for the discharge profiles a small UAS requires — sustained high-C draw on hover, brief peak surges on aggressive maneuver — and their packs ship with the documentation required for a §889 and §848 representation.

The trade-off, again, is cost and energy density. Comparable-spec packs from Chinese suppliers — Bonka, Tattu, and others — retail at thirty to fifty percent of the MaxAmps unit price, and the high end of the PRC supplier base now ships cells with slightly higher gravimetric energy density (Wh/kg) than is available domestically at the same cycle-life rating. For a counter-UAS interceptor whose mission is six to eight minutes total flight time, the cycle-life and energy-density gap is small and the cost gap is well within the budget envelope. For a long-endurance ISR platform whose mission is six to twelve hours, the gap becomes a design driver and the integrator may be forced to specify a more exotic Li-ion or LiPo-hybrid chemistry — which narrows the supplier set further.

The single biggest assumption an integrator makes is "how long does this thing have to fly." That assumption determines whether the U.S. battery supplier list is one company or zero.

[05] · STRUCTURE AND COMPOSITESDomestic carbon-fiber and the airframe layer

The airframe is the part of the BOM that gets the least attention in compliance conversations and the most attention in cost optimization. Carbon-fiber composite airframes for Group 1 and small Group 2 platforms can be sourced from domestic shops with substantial cost predictability — Rock West Composites, Composites One, and a handful of regional layup houses are the names that come up most often — and the propeller layer has long been served by Master Airscrew (U.S.) and APC Propellers (U.S.) for the most-common rotor diameters.⁶

The trade-off in the structural layer is producibility rather than compliance. A bespoke carbon-fiber airframe with hand layup runs $800 to $2,500 per platform in low volumes. The same airframe designed for injection-molded reinforced polymer or SMC compression-molded composite can fall below $250 per platform at volumes above 10,000 units annually. The decision is upstream of supplier selection — it is a design-for-manufacturability decision — and it is the single most important cost lever on a sub-$25K combat-UAS BOM.

[06] · WHAT THE FIVE SUPPLIERS LOOK LIKE TOGETHERThe systems-integration thesis

A 2026 small-UAS integrator who anchors to KDE Direct (motors/ESCs), ARK Electronics or CubePilot Blue (autopilot), MaxAmps (energy), and a domestic composites stack (structure) has a §848-compliant platform with a defensible procurement file. The unit-cost premium relative to a DJI-equivalent platform is real — at present, somewhere between 1.8× and 3× for an apples-to-apples performance spec — and is shrinking by roughly fifteen to twenty percent annually as production volumes catch up to demand and as second-source suppliers mature.⁷

The procurement implication is straightforward. A buyer who specifies §848 compliance plus a unit-cost target needs to commit to a production quantity that justifies the supplier base. A 50-unit purchase will pay full premium across the BOM. A 5,000-unit purchase from a single integrator can collapse the premium to a level where the §848-compliant platform is unit-cost competitive with the non-compliant alternative on a fully-loaded TCO basis. The DoD's Replicator and Replicator 2.0 initiatives — which name attritable UAS specifically as a target capability — are the policy vehicle to push volumes into that regime.⁸

[07] · CLOSINGThe integrator's job is to choose the trade-off

There is no domestic supplier mesh that gives an integrator full performance parity with the DJI consumer line at the DJI consumer price. The mesh that exists today, anchored by the five suppliers above, gives a defensible §848 and §889 posture, a documented bill of materials, an audit-survivable procurement file, and a unit-cost premium that is closing on its own schedule. That is enough to win the procurement fight — once the procurement officer is asking the right questions.

Vertex Autonomy builds the X-7 Talon counter-UAS interceptor and the X-4 Raptor loitering munition against this supplier mesh, with KDE Direct motors, ARK Electronics flight controllers, MaxAmps cells, and a domestic-composites airframe. Every BOM line is documented at sub-tier and represented in writing. See the platforms. Open a procurement channel if you want the supplier-by-supplier BOM walk-through.