Project Maven JADC2 Lacks Cross-Authority Coordination Substrate

Vendor & Product Reality

Project Maven was established in 2017 as the DoD's flagship effort to apply machine learning to full-motion-video (FMV) exploitation, originally under the Algorithmic Warfare Cross-Functional Team (AWCFT) and since 2022 consolidated under CDAO. Its operational mission set includes object detection, classification, and tracking against FMV from airborne and overhead sensors, with model outputs flowing into intelligence-analysis workflows and increasingly into time-sensitive targeting cycles. Maven outputs are a recognized contributor to the JADC2 partner network, where they compose with the service-specific command-and-control programs — Air Force ABMS, Navy Project Overmatch, and Army Project Convergence — that constitute the operational JADC2 fabric.

The technical execution at program scale is mature. Maven has moved through multiple model generations, integrated with operational users including geographic combatant commands, and supported real-world targeting and intelligence missions. JADC2 itself has progressed from strategy document to fielded experimentation, with recurring joint exercises demonstrating cross-service sensor-to-shooter timelines that would have been infeasible a decade ago. The combined deployment scale across DoD components and select coalition partners is significant.

Within a single program's authority — Maven's intelligence-analysis customer base, or a single service's C2 program — coordination works. Inside that boundary, models are trained, outputs are produced, consumers subscribe, and decisions are made under a coherent accreditation and authority regime.

Architectural Gap

Joint and coalition target-decision workflows are inherently cross-authority. A target nomination produced by Maven analytics may pass through an Air Force component decision cell, be cross-checked against a Navy maritime picture, be deconflicted against allied operations, and ultimately be admitted by a combined joint task force commander operating under explicit rules of engagement and national caveats. Each of those authorities owns part of the decision; none owns it alone.

The architectural problem is that the decision chain across these authorities is not cryptographically bound. Maven's output enters JADC2 partner-network exchanges as data, where it inherits the trust posture of whichever fabric carries it. Service C2 programs each maintain their own data fabrics — ABMS dataONE, Overmatch's equivalent, Convergence's mesh — with their own credentialing, accreditation, and policy regimes, federated to one another and to coalition systems through gateways and integration projects rather than through a shared architectural object. When the targeting decision is reviewed — by a commander before commit, by a legal advisor under the law of armed conflict, by a post-strike assessment cell, or by a coalition oversight body — the chain from sensor through Maven model through joint approval to commit is reconstructed procedurally, not enforced architecturally.

The gap is not in any one program. ABMS is doing its job; Overmatch and Convergence are doing theirs; Maven is doing its. The gap is the absence of a layer above all of them: a cross-authority coordination substrate in which contributions from each program enter as credentialed events, compose through a declared multi-party admissibility relation, and produce a target-decision lineage that is verifiable end-to-end across services and coalitions without forcing any of them onto a single program's data fabric. JADC2 doctrine requires this layer; the JADC2 implementation today does not architecturally provide it.

What the N-Party-Coordination Primitive Provides

The n-party-coordination primitive treats each program's contribution to a joint or coalition decision as a credentialed event published under that program's own authority, and composes those events through a declared multi-party admissibility relation. Maven model outputs enter as signed analytic events under CDAO authority. Service C2 contributions — track correlations, deconflictions, fires-cell concurrences — enter as signed events under their own service authorities. Coalition contributions enter under the credentialing regime each partner nation already operates. None of these authorities is asked to migrate onto a shared fabric; each remains sovereign over its own credentialing.

The admissibility relation expresses what composition of these contributions admits a target decision. Joint requirements — for example, concurrent Air Force and Navy concurrence on a maritime target, or coalition oversight on cross-border action — are expressed once in the relation, rather than re-encoded per integration project. The decision, when admitted, carries lineage that traverses every contributing authority, and that lineage is cryptographically verifiable without trusting any single program's fabric.

Three properties follow. Cross-program audit becomes structural rather than discovery-dependent. Cross-service composition becomes a first-class operation rather than a gateway exercise. Coalition operation becomes admissible without forcing partners onto US-program fabrics — the property that JADC2 doctrine has long required and that no current program of record architecturally produces.

Composition Pathway

Composition with the existing Maven and JADC2 substrate is additive. Each contributing program continues to operate its own data fabric, model pipeline, and accreditation regime. The n-party-coordination primitive is introduced as an overlay in which each program's relevant outputs are wrapped as credentialed events under that program's existing authority, and in which the multi-party admissibility relation runs at the point where joint or coalition decisions are admitted — typically a combined joint operations center, a fires cell, or a CDAO-mediated decision surface.

For Maven specifically, the composition wraps model outputs and analytic conclusions as signed events bound to the model version, the input provenance, and the analyst or automated workflow that issued them. For ABMS, Overmatch, and Convergence, the composition wraps the C2-side contributions — the concurrences, deconflictions, and tasking decisions — under each service's own credentialing. For coalition partners, the composition accepts each partner's national credentialing without requiring its replacement.

The decision surface evaluates the multi-party admissibility relation against the composition of contributions present at decision time. Decisions that meet the relation are admitted with full cross-authority lineage; decisions that do not are held, and the held state itself is a credentialed event that subsequent authorities can act on. JADC2 partner-network exchanges continue to carry data; what changes is that target-decision authority no longer rides on data trust, but on cryptographic composition of the authorities that contributed.

Commercial & Licensing

For DoD and the prime contractors building into Maven and JADC2, the n-party-coordination primitive is the architectural answer to a problem the doctrine has stated for years and the integration market has not yet solved. The primitive sits above existing programs of record and increases their joint and coalition value rather than competing with them. CDAO gains a structural decision-lineage substrate for AI-mediated targeting; service C2 programs gain a composition layer that does not require them to subordinate their fabrics; coalition partners gain a participation model that does not require them to adopt US-program credentialing wholesale.

The patent positions the multi-party coordination primitive precisely at the cross-authority layer JADC2 doctrine requires and current programs do not architecturally produce. Licensing pathways include direct CDAO adoption as a JADC2 overlay component, prime-contractor integration into ABMS-, Overmatch-, and Convergence-class deliveries, and coalition-facing variants negotiated under existing partner agreements. In each case the primitive converts an integration-project problem into an architectural property — and converts the cross-authority coordination gap from a recurring program risk into a licensable capability.