All defense POCs
POC 05DDeterminismBStability guaranteeMulti-node buildWave 3

Coalition-Interoperable Autonomous Fire Control

Three navies, three certification regimes, one engagement decision — bit-identical, rate-bounded, attestable from the same artifact.

1
Unique engagement hashes across 3 navies
1.6245×
Max per-tick excursion observed (bound 2.4623×)
1
Certification artifact for 3 national regimes
200
Decision ticks per engagement

The scenario

Set the picture

A US Navy DDG, a UK Type 26 frigate, and an Australian Hunter-class frigate share a coordinated air-defense engagement against a multi-vector threat. Each ship's combat system runs on different hardware, different sensor calibration, different national software builds, under different national certification regimes.

The engagement decision must be reached the same way on every ship within the doctrinal time budget, and the per-tick command excursion to the effectors must satisfy each nation's certification regime. The same scenario generalizes across coalition operations: AUKUS Pillar 2 autonomous systems, Five Eyes maritime patrol coordination, NATO integrated air-and-missile defense, NORAD bi-national air defense.

What it costs today

Coalition fire-control coordination is a deeply manual problem. Cross-fleet engagement coordination depends on voice procedures, pre-agreed engagement zones, and rules-of-engagement doctrine that compensates for the fact that the ships' combat systems do not actually agree on the threat picture or the engagement geometry.

Each nation certifies its own autonomy and fire-control stack independently. The same effector behavior is reviewed three times for three regulatory regimes that each demand a different evidence package. Certification cost scales with the number of partner nations.

When something goes wrong — a near miss, a friendly-fire incident, a missed intercept — the after-action investigation has three different records of what each ship 'thought' was happening. Reconciliation of those records is months of work and rarely conclusive. Translation overhead in every cross-fleet data fabric (Link-16, Link-22, NATO STANAG) compounds the bandwidth and latency cost.

What changes with SolvNum

Two capabilities, two parts of the coalition-autonomy problem, one substrate.

Dcross-platform determinism

When the track-fusion math, intercept geometry, and engagement timing run on SolvNum, every ship — regardless of national hardware, national software stack, or national operating environment — produces the bit-identical engagement geometry. The fleet picture is a single shared object, not three approximately-aligned national pictures.

Bper-step excursion limit

The per-tick command excursion to the effectors satisfies each nation's certification regime from the same artifact. Each navy's safety review board can accept the same SolvNum-bounded controller because the bound is a mathematical property of the data type, not a property of national-specific code. The coalition operates as a single distributed combat system, certifiable once and re-attestable on every patrol.

Measurable outcome

What we'll claim — and how it survives review

Each line below maps to a captured number in the demo section. Every number is reproducible from the SolvNum validation suite.

  • Three-nation engagement decision provably derived from one math, attestable by a hash all three navies independently produce.
  • Per-tick effector-command excursion bound certifiable under three national regimes from the same evidence package — single artifact replaces triplicated reviews.
  • Cross-fleet investigation time after a near-miss or anomaly reduced from months to a single hash-compare and a state-trace replay.
  • Translation-layer numerical overhead in coalition data fabrics reduced to zero — every receiver decodes to the same bit-pattern the sender produced.
  • Coalition-JADC2 numerical interoperability becomes a shipping capability, not an aspirational milestone.

The demo

What was tested. How. What the script printed.

A three-machine coordinated intercept demonstration. Each machine represents one nation's combat system: x86_64 Linux server (US reference build), ARM SBC (UK), CUDA GPU (AUS). A multi-vector air threat enters the engagement zone. All three machines independently fuse the same sensor inputs, compute the same engagement geometry, issue the same engagement command, and log the same SHA-256 of the engagement decision.

The demo concludes with three printed hashes that match, three certification-bound attestations that satisfy three national regimes, and one engagement decision the coalition reached as a single distributed system — the first multi-national autonomy interoperability artifact ever generated from a single math substrate.

Live simulation

Animated in-browser simulation of what the demo proves. The numbers underneath are the captured demo output.

Coalition engagement decision — same hash, three navies

COMPUTING...

US

United States Navy · x86_64

801cc675be837

0.0000×

limit 2.4623×

UK

Royal Navy · ARM

e2da4e08a70f5

0.0000×

limit 2.4623×

AUS

Royal Australian Navy · CUDA

a90282544ad6a

0.0000×

limit 2.4623×

Captured demo output

The numbers the script actually printed.

Coalition national stacks — same engagement, three nations (12-hex hash prefix)
NationPlatformEngagement hashMax excursionBound
USx86_64 server8bd9f85becb11.6245×2.4623×
UKARM partner8bd9f85becb11.6245×2.4623×
AUSCUDA GPU8bd9f85becb11.6245×2.4623×

Coalition Engagement Attestation — Multi-Vector Air Defense

Engagement hash (all nations)
8bd9f85becb169eb713a8d384ef00668b0debf1aef6e6fc340f6a63410adb258
Targets
6
Decision ticks
200
Per-tick excursion bound
≤ 2.4623× per-step excursion
Bound compliance
YES (all 3 nations)
Cross-national identity
YES
SolvNum table version
core.K=24, TABLE_BITS=11
Certification regimes covered
US, UK, AUS (single artifact)

Evidence pointers

Where the claims live in the repo

These are the files a reviewer should run, read, or grep to re-derive every number on this page.

  • SolvNum cross-platform determinism verification (x86, ARM, WASM, CUDA)
  • SolvNum validation suite — excursion-limit verification
  • SolvNum documentation — coalition narrative
  • SolvNum benchmark suite — identity + stability-bounded control demos — composite verdict

Want to see this in your environment?

Brief us on a program where this POC matters.

ITAR-aware. Air-gapped delivery available. Every claim above traces back to a script in the public repo.

Brief us