Airspace Intelligence Core Platform

1. Observations arrive at different times
2. They are correlated into candidates
3. One source goes stale and the claim softens
4. The source recovers and the state is restored
5. The moment is replayed with what was known then

Sources report in different formats, at different times and with different confidence. A console that draws them as dots shows what arrived, not what can be reviewed: which observations belong together, how old each one is, and what is still unknown.

• Observations from many sourcesEach arrives with its own time, format and quality. None of them is trusted because it arrived.
• Context and historyEarlier states, site context and source health are kept, so that a state can be compared with what was known before.
• Explicit unknownsGaps and stale inputs are recorded as unknown, never as zero and never as clear.

1. Correlates the observations before any fusion, so that only what belongs together is combined.
2. Maintains one state object whose fields carry freshness, confidence, evidence and the unknowns.
3. Keeps the history, so that any moment can be replayed and any state shared under governance.

• Unified does not mean certain: the state carries its own confidence band.
• Known then and known now are kept apart, and the difference can be inspected.
• When an input degrades, the claim is softened before anything is redrawn.

The core platform is the layer in the middle of the truth chain. It receives what the sensing projects observe, forms one time-aware state from it and hands that state, with its evidence, to fusion, to replay and to the interfaces that need it.

Read the architecture

Data status: Illustrative

Evidence ladder

1. E0 · Concept reached
2. E1 · Architecture Defined reached
3. E2 not reached
4. E3 not reached
5. E4 not reached
6. E5 not reached

Project assurance panel

Mission and context

Give every environment a single airspace state that people can explain, review and hand over, instead of several unreconciled views.

Primary contribution

A shared state model with defined fields for the identity hypothesis, position band, freshness, sources, confidence, evidence and unknowns.

Decision value

Helps a reviewer see what is supported, how old it is and what is missing before deciding what to do. It does not make the decision.

Evidence status

Evidence level: E1 · Architecture Defined A frozen architecture specification exists. No prototype, test result or deployment is claimed.

Next validation step

Run the state model against synthetic multi-source scenarios in simulation and record the outcome as the first evidence record.

Behaviour when degraded

When a source slows or drops, its freshness is shown, the confidence band widens and the claim softens first. The state stays, marked as less certain, and recovers when the source returns.

Role in integration

Publishes the state and its history to fusion, to replay and, through the data interface, to other systems. It consumes observations and does not command sensors.

Human review

A person reviews every state that leads to an action. The platform prepares the picture and its evidence; it does not act on it.

Public-safe limitation

Everything described here is a design at the architecture stage, and every value shown is illustrative.

Open unknowns

How the state fields behave with real sensor timing, how much history operators will need and which handover formats sites will accept are not yet known.

When a source slows or drops, its freshness is shown, the confidence band widens and the claim softens first. The state stays, marked as less certain, and recovers when the source returns.

• A protected site that needs a handover record

  Why it matters here

  Shift changes lose context when every view is separate.

  What it contributes

  One reviewable state and its history for the incoming shift.

  What it does not replace

  The site's own procedures and the people who apply them.

• A region combining several local pictures

  Why it matters here

  Local views differ in timing and in detail.

  What it contributes

  A common state model that keeps each site's evidence attached to it.

  What it does not replace

  Sensing at the sites, which still has to observe.

Publishes the state and its history to fusion, to replay and, through the data interface, to other systems. It consumes observations and does not command sensors.

Run the state model against synthetic multi-source scenarios in simulation and record the outcome as the first evidence record.

• The platform forms a state from what is observed. Where nothing is observed, the state says unknown; it cannot supply an observation that no source made.
• Confidence depends on the sources connected, their timing and the local environment. A state is only as strong as its inputs.
• At this stage the evidence is a written specification. No measured result supports a claim about performance.
• Unification makes a picture easier to review; it does not make the picture correct, and it does not replace human judgement or certified systems.

• Multi-Sensor Fusion EngineEvidence level: E0

  Fusion is the step that combines compatible observations inside the state.

• Airspace Data & Intelligence APIEvidence level: E0

  The data interface carries the state out with its evidence still attached.

• Passive RF Sensing NodeEvidence level: E1

  A passive RF node is one of the sources that feeds the state.

• Passive Coherent Location (PCL) RadarEvidence level: E1

  Passive coherent location adds spatial observations whose confidence depends on geometry.

• Critical Infrastructure Airspace ProtectionEvidence level: E0

  Site protection is the first environment that needs a reviewable continuity record.

• Near-Ground Airspace Intelligence NetworkEvidence level: E0

  Near-ground airspace is where a coherent state is hardest to build and most needed.