GPS-Denied & PNT Integrity Evidence Assurance
Review whether positioning and navigation evidence remains trustworthy when GPS or other PNT evidence is degraded, missing or contradicted by other sources.
Who this is for
Navigation engineers, sensor-fusion leads and PNT assurance reviewers preparing degraded-navigation evidence before field escalation.
The problem you already have
Several navigation sources can look independent while sharing the same hidden dependency
Part 80 introduces source-dependency and independence checks for PNT evidence. For this product, that supports a bounded offline review of whether GNSS, inertial, visual, terrain, timing or other positioning sources actually provide independent corroboration, rather than simply repeating one upstream clock, map, calibration or earlier fused bias.
A confident fused state can still be invalid because frame, time or freshness evidence no longer matches
Part 80 keeps reference frame, time base, uncertainty, latency, freshness and dependency semantics explicit before heterogeneous PNT evidence is treated as one state. This helps surface cases where a source is locally plausible but unusable for the current review because a required transfer is missing, stale or contradictory.
Degraded navigation should narrow authority instead of silently preserving the previous state
The patent lineage distinguishes bounded holdover, dead reckoning and degraded-PNT modes from full-support operation. The public product uses that idea only for offline evidence review: it can flag that a scenario should be downgraded or re-reviewed, but it does not control a vehicle, aircraft or hardware system.
Real-World Operational Scenario: The Navigation State That Stayed Confident After Its Independent Sources Disappeared
Scenario: GPS drops out, inertial and visual sources continue, and the fused navigation state still reports high confidence — but both surviving sources depend on the same stale map/calibration chain.
The navigation team needs to know whether the replay still supports the agreed degraded-PNT review question or whether the state should be downgraded, held for review or treated as unsupported.
WITHOUT this gate: source count and confidence can make the state look resilient even when true independence, freshness or frame/time evidence has collapsed.
WITH this gate: source independence, frame/time transfer, freshness, contradiction, holdover age and fallback evidence are reviewed together. The result remains a software-pilot evidence disposition; no live navigation or route authority is created.
What BivectorAI Provides
PNT Source Independence Review
Reviews whether the supplied navigation sources provide genuinely distinct support or share hidden clock, map, calibration, preprocessing or upstream-state dependencies.
Frame / Time / Freshness Consistency Review
Uses Part 80 technical lineage to surface stale timestamps, unresolved reference-frame transfers, latency issues and other evidence conditions that can invalidate an otherwise plausible fused state.
Degraded-PNT / Holdover / Fallback Review
Separates bounded degraded states from full-support navigation evidence and identifies where additional validation or human review is required before field escalation.
Offline-Only Navigation Authority Boundary
Returns customer-safe evidence findings while live flight/vehicle control, route generation, evasion guidance, targeting, actuator authority and field certification remain outside scope.
Prove the Gate Before Broader Integration
Back-test the degraded-PNT evidence gate against frozen historical or synthetic replay cases with known engineering disposition.
Freeze the accepted source fields, dependency relationships, frame/time-transfer requirements, freshness rules, holdover/fallback conditions and downgrade/review criteria before revealing the historical disposition. Broader integration is justified only if the same bounded contract reproducibly surfaces material PNT integrity failures before field escalation. This is a proposed validation methodology, not field certification or proof of autonomous navigation safety.
What you provide
Customer-provided or synthetic offline logs, replay records and sensor summaries containing only the positioning and navigation evidence required for the agreed review.
What BivectorAI actually hands you
Navigation evidence-readiness assessment
A bounded disposition tied to the exact supplied offline/replay PNT evidence package and agreed review question.
Sensor-consistency / contradiction findings
A concise record of source disagreement, dependency, frame/time and freshness conditions that support or weaken the navigation-state evidence.
Degraded-PNT / fallback review findings
An explicit list of holdover, dead-reckoning, fallback or downgraded-state conditions requiring further engineering validation or review.
Hash-bound customer-safe evidence record
A retained input-linked artifact tying the findings and limitations to the exact supplied offline evidence for technical review.
What we will never ask you for
The current scope is offline and replay evidence review, so live vehicle control, actuator authority, targeting inputs and hardware credentials are not required.
- Live flight or vehicle control.
- Autopilot, actuator or motor-command authority.
- Target-routing, evasion or weapon-control inputs.
- Live hardware credentials.
How it actually runs
Start with a bounded evaluation, not a large commitment
Fixed scope / quoted per engagement. We define the evidence boundary, inputs, outputs and acceptance criteria before work begins.
If the offline degraded-PNT review does not establish that the evidence layer is useful for the agreed navigation question, there is no reason to move into a broader commercial commitment.
Environment mapping, bounded data contract and agreed evidence-delivery scope.
Annual commercial access within the agreed product and deployment boundary.
Fixed program scope where a bounded evidence engagement is preferred.
Current status: customer software-pilot evidence layer for offline/replay evaluation. It is not a live flight-control or live hardware-control system and is not field-certified. It does not generate target routes, evasion guidance, intercept paths, weapon control or autonomous engagement authority.