Firestorm vs. Purple Rhombus
Who wins at UAS production?
Firestorm and Purple Rhombus. Distributed UAS production versus centralized UAS production. Two AFWERKX awards. Two production philosophies. The same broad timeline.
Purple Rhombus won SBIR phase III this past January for “industrial scale affordable mass” with existing American sheet-metal factories. Firestorm raised a $47mm series A in July 2025. More recently they raised their $82mm series B this week and demonstrated xCell-- its containerized factory concept-- to Kirsten Gillibrand at their Rome, NY facility this month. Seeing Firestorm and Purple Rhombus on these parallel paths made me wonder: who will win? Defense tech likely thinks along the same lines (I haven’t seen a portfolio with both bets at this point)-- centralized versus distributed UAS production as an either/ or thesis pick.
It’s not. It’s a portfolio to construct in service of policy and doctrine.
The importance of a “both/ and” framework in UAS production models
Let’s start with the core premise of my argument: mass and resilience are not the same problem. Centralized production solves mass. Distributed production solves resilience. Consequently, they are not substitutes. They are complements.
This requires us to deeply rethink how our modern supply chains support US strategic interests. Let’s think about the problem set at hand: operations in the Indo-Pacific will require sustainment over 6,000 miles from the West Coast to Guam or Japan. We can no longer afford to think of solving supply chain challenges with one single approach. RAND analysis rightfully points out the fault in this way of thinking. The authors describe just-in-time logistics as “efficient but brittle,” and “existing sustainment plans are likely to break down” from direct disruption.
We fall into a trap when we think of UAS production in cost per unit rather than cost per effect delivered in the right place at the right time. The PLA will doctrinally contest the entire kill chain-- “factory to fort to port to foxhole.” A centrally produced airframe that never makes it to the fight has a functionally infinite cost. The AUSA’s contested logistics report calls for turning our supply chain into a “supply web.” We need both centralized and distributed UAS production if we want innovation to be in service of strategy.
A framework for a UAS production thesis
So, if looking at it as zero-sum is wrong, then what’s right? I think there are two factors to consider:
Peacetime (stockpile) versus wartime (replenishment)
CONUS-produced versus theater-produced
These factors create four different scenarios for distributed and centralized UAS production:
The real investing question is: “what companies have operator buy-in, procurement fluency, and the ability to scale this to procurement?” Centralized production wins for the mass required in early operations: day one mass strike packages and large attritable swarms from pre-positioned stocks. In other words, this is the “millions of drones” challenge the Secretary of War directly names.
Conversely, distributed production is what keeps the force resilient in contested environments: replacing attrited platforms, reconfiguring UAS platforms for mission-specific requirements, and allied co-production. Allied co-production is particularly important. The AUSA recommends Congress identify allies in the Indo-Pacific with whom we can produce critical technologies. Firestorm’s open-architecture production stacks answer for that in a way fixed factories don’t.
Why Firestorm and Purple Rhombus are the right answers
Firestorm and Purple Rhombus showed they understand defense GTM early. They built relationships with end users well before hitting the fundraising trail. Firestorm describes xCell as a product that unsophisticated but well-trained end users can use to “launch complete build cycles in contested environments.” Similarly, Purple Rhombus designed their PR-2 to support payload swaps in the field and ATAK integration. The ATAK (combat smartphone) point is easy to miss if you haven’t talked to end users who have actually used ATAKs. Moreover, both built Modular Open Systems Approach (MOSA)-compliant architectures. Purple Rhombus natively integrates ATAK, MAVLink, and Starlin. Neither Firestorm nor Purple Rhombus strand defense buyers in vendor lock-in.
Both companies have also performed in terms of procurement, as both have crossed the valley from BA3 to BA4. Purple Rhombus has a SBIR phase III grant, and Firestorm has a $100mm indefinite delivery, indefinite quantity (IDIQ) contract. Lots of people are in defense tech; very few people in it can say they’ve made it that far. Both also speak the right commercial language. They don’t talk about how they’re cheaper. They talk about how they deliver a capability that makes end-users more lethal. That’s the language that’s easy to get wrong; Firestorm and Purple Rhombus don’t get that wrong. Capital is downstream of pull from end users. The winners-- like Firestorm and Purple Rhombus-- build operator relationships and procurement fluency first and let the private capital follow.
Strategic alignment as a capital allocation roadmap
Serious UAS production allocation looks like a barbell. Centralized, large-scale wins for peacetime mass; distributed wins in sustained wartime environments. To say one or the other will win is to fundamentally misunderstand contested logistics doctrine the Pentagon wants. The bets that win in either model have three properties:
End user pull
Real understand of procurement
Open architecture
Both Firestorm and Purple Rhombus boast all three. That’s why they’ve had the trajectories they’ve had. The DoW is planning for our UAS production to support both central and distributed production. The DoW recognizes supply chain resilience must be multidimensional with no one hard and fast answer. Defense tech capital allocators should recognize this, too. Robust strategy rarely relies on one solution. It’s the defense tech community’s responsibility to innovate according to this principle.