The US Battlefield Management System Market: CJADC2, Program Consolidation and the Autonomy Gap
The largest BMS market in the world is mid-transition — from platform-centric command programs to a joint data fabric — and the autonomy layer is the contested ground.
The United States is the largest battlefield management system market on earth and the most difficult to characterise, because it is not one market. It is a set of service-specific command programs being forcibly reconciled into a joint data architecture while simultaneously absorbing autonomous systems at a pace the programs were not designed for.
The organising initiative
CJADC2 — Combined Joint All-Domain Command and Control — is less a system than a policy commitment: any sensor, any shooter, any domain, any coalition partner, connected by a common data layer. Its implementing efforts include the Army's Next Generation Command and Control (NGC2) and its TITAN ground station, the Air Force's ABMS and cloudONE/deviceONE components, the Navy's Project Overmatch, and DIU's autonomy and orchestration lines including the Replicator initiative.
The supplier landscape
Legacy primes. Lockheed Martin, RTX, Northrop Grumman, General Dynamics Mission Systems, L3Harris and Leidos hold the mission-system baselines, the accreditation pathways and the sustainment revenue. Their structural advantage is that they know how to deliver inside the acquisition system; their structural weakness is release cadence.
Software entrants. Palantir dominates the data-integration layer and has converted analytic strength into programs of record. Anduril has built the most complete autonomy-native command product in the market, tied to its own hardware line. Scale AI, Applied Intuition, Shield AI and Govini occupy adjacent layers.
The middle tier. A large population of venture-funded autonomy and C2 companies now reaches pilots through OTAs and the software acquisition pathway in months rather than years. This tier did not meaningfully exist a decade ago and is where architectural experimentation is happening.
The three real problems
1. Data model fragmentation. CJADC2 requires a common data fabric. Each service arrived with an incompatible one. Most current spending is reconciliation work — valuable, but not new capability.
2. The autonomy command gap. DoD can now field hundreds of autonomous systems. Doctrine and command software for employing formations of them at scale lags badly. Exercises repeatedly show operator saturation well before the technical asset limit is reached: the constraint is human cognitive load, not airframe count.
3. Vertical integration risk. The most capable autonomy command products are coupled to their vendors' hardware. That is efficient for the vendor and a strategic dependency for the government, which has spent thirty years learning what sole-source lock costs.
Where the market is heading
Procurement language is converging on three demands: modular open systems approaches with government-owned interfaces, demonstrated performance in denied and degraded environments, and operator-to-asset ratios that improve rather than degrade as formations grow.
That third metric is the one that will separate winners. A system that requires one operator per four autonomous platforms is a research demonstrator. A system that holds one commander per formation regardless of size is a capability.
The KhanBMS position
KhanBMS addresses the autonomy command gap directly and structurally. The decimal hierarchy of the Mongol army — Arban (10), Zuun (100), Minghan (1,000), Tumen (10,000) — caps the number of objects any commander supervises at ten per tier. Formation size becomes an implementation detail rather than a cognitive burden.
Two further design commitments align with where US requirements are moving. First, hardware agnosticism: a ROS 2 node-isolation model lets any compliant platform from any vendor occupy a tier, keeping the command layer government-favourable rather than vendor-locking. Second, denied-environment doctrine: intent is pre-decomposed and distributed, so a Zuun cut off from its Minghan continues lawful execution and reconciles state on reconnection.
The US market's money will keep flowing through the primes for the foreseeable future. Its decisive capability question is narrower and more urgent: when a brigade commander is handed a thousand machines, what does the screen in front of them look like?