Engine IA

Firmware/API output governance architecture.

The engine is organized around adapter normalization, sweat quality, device-event disturbance modeling, marker policy, fail-closed output routing, Evidence Packet generation, and replay verification.

1. OEM Biofluid Adapter

Normalizes vendor payloads into marker, unit, candidate presence, calibration, fill, quality, sampling time, and raw access policy.

2. Sweat Quality Frame

Evaluates sweat rate, fill state, contamination, calibration, drift, reagent age, skin temperature, and ambient state.

3. Device Event Frame

Models BLE, haptic, ADC, battery, charging, CPU/MCU load, firmware, SDK/API request, and frame movement events.

4. Disturbance Vector

Combines quality and event state into marker-aware disturbance components used before release.

5. Marker Policy Package

Controls allowed output modes, quiet windows, artifact windows, trend windows, raw policy, and evidence required fields.

6. Output Router

Routes to full value, limited value, trend-only, warning, request remeasurement, raw access blocked, or non-value response.

Readiness JSON

loading
VerifyAdapterEvidence