Deep Signal: US drone manufacturing infrastructure buildout: Arsenal-1, SkyFoundry 10K/month, additive manufacturing pivot
US military executes three-tier domestic drone manufacturing buildout: Anduril's Arsenal-1 for Group 5 systems, Army's SkyFoundry targeting 10K+/month small UAS, and additive manufacturing for squad-level replenishment by end-2026.
- $1B Arsenal-1 committed capital Anduril, Pickaway County Ohio, ~1.7M sq ft
- 10,000+/month SkyFoundry small drone production target Army program, domestic manufacturers
- 120,000/year Annualized SkyFoundry output at target rate Compared to estimated 100,000+/month Russian FPV peak
- 500 units / $250M Roadrunner initial DoD contract Near-term Arsenal-1 revenue anchor
- Date
- 2026-05-28
- Type
- deployment
- Deal Value
- ~$1B (Arsenal-1 facility); $250M (Roadrunner/Pulsar contract)
- Status
- announced
- Source
- Original report
US Drone Manufacturing Infrastructure: The Industrial Mobilization Behind Three Interlocking Programs
What Happened
The US military is executing a three-layer domestic drone manufacturing buildout simultaneously, each layer targeting a different scale and mission profile. [1] Anduril's Arsenal-1 facility in Pickaway County, Ohio — approximately 1.7 million square feet across two buildings, representing roughly $1 billion in committed capital — is targeting Fury CCA production in Q2 2026 and broader drone/autonomous vehicle output by July 2026. The Army's SkyFoundry program is separately targeting 10,000+ small drones per month from distributed domestic manufacturers. A third pillar — an additive manufacturing strategy using 3D printing at forward locations — aims to field at least one drone to every Army squad by end of 2026.
These are not redundant programs. They are deliberately tiered: Arsenal-1 handles Group 5 and high-complexity systems; SkyFoundry handles mass attritable Group 1-3 production; additive manufacturing handles last-mile distribution and battlefield replenishment where supply chains cannot reach.
The question is whether the US can compress a 24-month industrial buildout into 18 months without the component supply chain to match.
Why It Matters
The production math is clarifying. Ukraine and Russia have each been producing small drones at rates estimated between 50,000 and 150,000 units per month at peak periods, with Russian FPV production reportedly exceeding 100,000/month by late 2024. These figures derive from open-source conflict monitoring and defense intelligence assessments tracking drone consumption rates in the Ukraine conflict. SkyFoundry's 10,000/month target — 120,000 annually — represents a meaningful but still asymmetric baseline against peer-conflict consumption rates. HIGH CONFIDENCE that 10K/month is a floor figure designed to prove domestic industrial capacity, not a ceiling sized to peer-conflict demand.
The additive manufacturing pivot addresses a specific logistics failure mode: traditional supply chains cannot replenish attritable drones fast enough at the squad level in a contested environment. If a squad burns through three drones in a 72-hour engagement, waiting for a resupply convoy is operationally unacceptable. Printing replacement airframes forward — even if electronics still require supply chain support — compresses that cycle.
The NDAA Section 842 problem remains partially unresolved. Section 842 restricts DoD procurement of drones containing Chinese-origin components, specifically targeting DJI and Autel supply chains. SkyFoundry's domestic manufacturing mandate directly addresses this, but the component-level dependency — motors, ESCs, sensors, and battery cells with Chinese supply chain exposure — is not eliminated by assembling airframes in Ohio or Texas. MODERATE CONFIDENCE that SkyFoundry contracts will require component-level attestation that most current US small drone manufacturers cannot fully satisfy by end-2026.
| Program | Scale Target | Unit Class | Timeline | Status |
|---|---|---|---|---|
| Arsenal-1 (Anduril) | Thousands/year | Group 5 CCA, Roadrunner | Q2-Q3 2026 production start | PROTOTYPE → SCALING |
| SkyFoundry (Army) | 10,000+/month | Group 1-3 small UAS | Active procurement | LIMITED → SCALING |
| Additive Manufacturing | Squad-level distribution | Group 1 micro/nano | End-2026 fielding goal | PROTOTYPE |
| Roadrunner (Anduril) | 500 units (initial contract) | Counter-UAS interceptor | Mid-2026 Arsenal-1 ramp | FIELDED → SCALING |
Who Is Affected
Skydio is the most directly exposed US manufacturer to SkyFoundry contract competition. With its R-series platform already on the Blue UAS Framework and active Army relationships, Skydio is positioned for SkyFoundry volume but faces production scaling questions — the company has not publicly demonstrated 10K/month capacity. Shield AI competes on autonomy software for small UAS but lacks the airframe manufacturing depth SkyFoundry requires. Joby Aviation and Wisk are irrelevant at this tier. Firestorm (formerly known as Kratos subsidiary programs) and AeroVironment — with its Switchblade and Puma lines — are the most credible volume competitors for SkyFoundry contracts. AeroVironment's Switchblade 300 is already FIELDED and has demonstrated production at scale, making it the benchmark SkyFoundry vendors will be measured against.
Anduril's Arsenal-1 does not compete directly with SkyFoundry vendors. Its competitive exposure is at the Group 5 CCA tier against Boeing's MQ-28, General Atomics, Northrop Grumman, and Lockheed Martin — all of whom have deeper program-of-record relationships but slower production infrastructure timelines.
Traditional defense primes face structural disadvantage in the additive manufacturing pillar. Companies like Firestorm Labs and Hadrian (precision manufacturing) are better positioned for the distributed production model than Raytheon or L3Harris.
What to Watch
- Q2 2026: Anduril Arsenal-1 Fury production line start — the single most verifiable milestone in this buildout. Any slip beyond June 2026 signals execution risk across the entire Arsenal-1 thesis.
- Q3 2026: SkyFoundry contract awards — watch for whether awards go to established Blue UAS vendors (AeroVironment, Skydio) or open new entrants. Contract structure (IDIQ vs. firm fixed-price) will signal Army confidence in vendor capacity.
- December 2026: Army squad-level drone fielding target — LOW CONFIDENCE this is achieved universally. Watch for a redefined metric (e.g., "brigade-level availability" rather than every squad) as the deadline approaches.
- Ongoing: Component-level NDAA compliance attestations from SkyFoundry awardees — this is where the program's domestic manufacturing credibility will be tested or exposed.
- H1 2026: USAF CCA downselect decisions affecting Fury's program-of-record status — a negative outcome would redirect Arsenal-1 capacity toward Roadrunner and other platforms but would not halt the facility.
Database Context
Anduril carries a DOMINANT intelligence rating with a WIDE moat assessment, driven by Lattice platform stickiness and Arsenal-1's capital-intensive manufacturing barrier. The $250M Roadrunner/Pulsar contract provides near-term revenue anchor while Arsenal-1 ramps. The broader SkyFoundry and additive manufacturing programs represent demand signals that benefit the entire domestic small UAS ecosystem — but Anduril's direct exposure is at the higher-complexity tier. The industrial mobilization pattern here mirrors historical defense surge models: tiered production by complexity, distributed manufacturing for resilience, and forward-production capability for attrition tolerance. The question is whether the US can compress a 24-month industrial buildout into 18 months without the component supply chain to match.
Methodology Note: Production rate estimates for Ukraine and Russia derive from open-source conflict monitoring, defense intelligence assessments, and published analyses of drone consumption in the Ukraine conflict. Confidence levels reflect editorial assessment of available evidence and source reliability.
Sources
- US drone manufacturing infrastructure buildout: Arsenal-1, SkyFoundry 10K/month, additive manufacturing pivot (signal, 3fa2a51a-5807-478e-bc54-54aa92ac1528)