Reports

# Vantage Weekly Pilot Report (replay demo) - Generated: 2026-06-15 15:22:47Z - Node: `node-controller-1` - Period: 5 replay scenario(s) - Total events: 6 (1 flagged low-confidence) ## Detector summary | Detector | Events | |---|---| | after_hours_presence | 1 | | degraded_signal | 1 | | dwell_anomaly | 1 | | node_tamper | 2 | | restricted_zone | 1 | Severity: high=4, medium=2 ## Events | ID | Type | Zone | Start (UTC) | Duration | Confidence | Flag | |---|---|---|---|---|---|---| | evt-b8bad8f807eb | dwell_anomaly | lobby | 2023-11-14 22:13:20Z | 70.0s | 0.90 | | | evt-775dc7a9a102 | restricted_zone_crossing | restricted-a | 2023-11-14 22:13:22Z | 1.0s | 0.90 | | | evt-11587cdcb571 | signal_degraded | - | 2023-11-14 22:13:25Z | active | 0.20 | low-confidence | | evt-2842d0376453 | node_tamper | - | 2023-11-14 22:13:30Z | active | 1.00 | | | evt-1960f99e69ff | node_offline | - | 2023-11-14 22:13:40Z | active | 1.00 | | | evt-4be126673cea | after_hours_presence | - | 2023-11-14 22:15:00Z | 2.0s | 0.90 | | ## Node health - `node-1`: **signal_degraded** at 2023-11-14 22:13:25Z (conf 0.20) - `node-1`: **node_tamper** at 2023-11-14 22:13:30Z (conf 1.00) - `node-2`: **node_offline** at 2023-11-14 22:13:40Z (conf 1.00) ## Privacy posture - Raw-signal egress allowed: **False** (default false) - Export policy: **deny** - Raw-capture retention: **0h** (encrypted at rest: True) - Consent-gated matching: **disabled** - Local auth required: **True** ## Known limitations - Results in this report are from synthetic/replay data unless explicitly marked as a live capture; live BFI capture is hardware-blocked on the build machine. - Vantage produces anonymous volumetric tracks (x/y/z + confidence) only. It does not recognize, identify, or match individuals. - Detectors are P0 facilities/security only: restricted zone, after-hours presence, dwell, node tamper/offline, and degraded signal. No medical or safety-of-life detection is provided. - Positional resolution from compressed-V-matrix feedback is coarser than raw-channel methods; treat outputs as zone/occupancy-grade rather than fine-grained localization. ## Next steps - Review flagged low-confidence events with the operator. - Tune zone boundaries / thresholds for any adjacent-zone false positives. - Schedule a controlled walk-test before live monitoring.