Validation showcase artifact

Validation Showcase

Flagship degraded-visibility multichannel replay, replay review artifacts, and two-session diff comparison.

Review Session Narrative

high disagreement

Score: 0.770 · Priority: high

Inspect EO/IIR availability intervals and channel dropout timing.
Review warning intervals for offset drift before disagreement.
Review warning bursts even if sustained disagreement was absent.
Trace channel state transitions against event windows.
Verify recovery interval before closing the review.

Heavy fog / small vessel replay review highlights degraded-visibility validation behavior. EO or IIR availability dropped during one or more replay intervals. 1 disagreement burst requires reviewer attention. Channel state transitions indicate changing sensor availability during replay. Agreement recovered before the session ended.

[warning · 00:00.00–00:02.00] EO/IIR offset warning interval observed.
[availability · 00:07.00–00:09.00] EO availability degraded during this interval (EO invalid/dropout interval).
[warning · 00:02.00–00:06.50] Agreement warning burst detected.
[recovery · 00:09.00–00:12.00] Channel disagreement stabilized after recovery.
[channel_state · 00:07.00] EO channel transitioned available -> unavailable at 00:07.00.
[channel_state · 00:09.00] EO channel transitioned unavailable -> available at 00:09.00.

Metrics appendix: disagreement fraction: 0.0 · unavailable fraction: 0.16666666666666666 · event window count: 4 · disagreement burst count: 1 · channel transition count: 2

Scenario Comparison Narrative

high disagreement review recommended

Posture: improved · Delta: -0.320 · Priority: medium

Severity label changed from high disagreement to review recommended.

Confirm improved review posture against left-session baseline.
Review warning intervals for offset drift before disagreement.
Review warning bursts even if sustained disagreement was absent.
Verify recovery interval before closing the review.

EO degraded / IIR valid reduced unavailable intervals relative to Heavy fog / small vessel while keeping disagreement behavior comparable.

[availability_delta] unavailable fraction reduced by -0.167.
[disagreement_delta] disagreement fraction unchanged.
[disagreement_delta] Disagreement burst count unchanged between sessions.
[channel_state_delta] Channel transition count changed by -2 (2 -> 0).
[event_only_left · 00:07.00–00:09.00] Only in Heavy fog / small vessel: EO degraded or unavailable; IIR remains available.
[event_only_left · 00:02.00–00:07.00] Only in Heavy fog / small vessel: EO/IIR offset warning interval.
[event_only_right · 00:03.00–00:08.00] Only in EO degraded / IIR valid: EO/IIR offset warning interval.
[event_only_right · 00:08.00–00:10.00] Only in EO degraded / IIR valid: EO/IIR offset warning interval.

Metrics appendix: disagreement fraction delta: 0.0 · unavailable fraction delta: -0.16666666666666666 · max azimuth delta deg delta: 0.0 · changed event count: 0 · only left event count: 2 · only right event count: 2

Generated: 2026-01-01T00:00:00Z

Scenario count: 4

Review session: Heavy fog / small vessel

heavy_fog_small_vessel_multichannel

Diff: Heavy fog / small vessel vs EO degraded / IIR valid

heavy_fog_small_vessel_multichannel vs eo_degraded_iir_valid_multichannel

Open review cards Open diff review cards Open bookmarks.json Open bookmarks report Open diff bookmarks

Flagship Artifacts

Replay Event Timeline

Channel State Timeline

Disagreement Bursts

Metric Comparison

Valid Fraction Comparison

Event Window Comparison

Synthetic, benign, validation-oriented, and non-operational. No image-content processing, ML detection, tracking filters, guidance, targeting, launch/deployment behavior, terminal behavior, warhead/fuze logic, or weapon-performance optimization.