Operator Fatigue & Cognitive Load:
Monitoring Without Wearables

Operator fatigue is not a welfare concern — it is a mission risk. In environments where a fatigued analyst misses an anomaly, a drowsy UAS pilot loses situational awareness, or an exhausted SOC operator fails to escalate a critical alert, the consequences are operational. Traditional fatigue management relies on duty hour limits, self-report, and supervisory observation — all of which fail in exactly the high-pressure, extended-duration scenarios where fatigue is most dangerous. EchoDepth provides a continuous, objective, camera-based readiness signal that fills this gap.

Deployment environments

What EchoDepth measures: the facial signature of fatigue

Fatigue produces consistent, measurable changes in facial muscle activation that precede subjective awareness and performance degradation by several minutes. EchoDepth tracks four primary AU clusters associated with fatigue onset:

• AU46 (blink rate and duration): Blink rate decreases and blink duration extends as fatigue progresses — a well-documented oculomotor signature that correlates with impaired vigilance. EchoDepth tracks this as a continuous time-series variable rather than a threshold trigger.
• AU43 (lid droop): The downward displacement of the upper eyelid — PERCLOS (percentage of eyelid closure) in the sleep research literature — is a primary fatigue indicator validated across transportation and aviation safety research.
• AU7 + AU5 (compensatory lid activation): As fatigue progresses, operators unconsciously attempt to compensate by widening their eyes — AU5 (upper lid raiser) and AU7 (lid tightener) activating against the fatigue-driven closure. This compensatory pattern is itself a fatigue indicator, not a sign of alertness.
• VAD Arousal trajectory: Sustained low Arousal on the VAD dimensional score — declining from individual baseline — is the primary readiness indicator, integrating the full AU picture into a single monitored variable.

Readiness scoring and alerting

EchoDepth produces a continuous per-operator readiness score on a 0–100 scale, derived from VAD Arousal deviation against individual baseline and weighted by fatigue-associated AU pattern activity. The score degrades incrementally rather than triggering at binary thresholds — enabling graduated responses proportionate to the level of impairment indicated.

Configurable alerting tiers can be set per deployment environment:

• Advisory (score 65–75): Supervisor dashboard indicator. No operational change required. Logged for post-shift review.
• Attention (score 50–65): Supervisor notification. Workload reduction or additional monitoring recommended.
• Action required (score below 50): Active alert to shift supervisor. Relief protocol initiated per standing operating procedure.

All scores and alert events are logged to structured JSON with operator ID (pseudonymised by default), timestamp, readiness score, and contributing AU pattern data. This log feeds crew duty management systems via API, SIEM platforms, or locally stored audit records.

Technical deployment specifications

Integration with crew duty management

EchoDepth readiness scores integrate with crew duty management and fatigue risk management systems (FRMS) via REST API. In UAS and aviation contexts, the EchoDepth readiness score provides a real-time physiological data layer that complements duty-hours-based FRMS models — which predict fatigue from schedule data but cannot measure actual operator state.

In practice, an operator whose duty-hours model suggests low fatigue risk but whose EchoDepth readiness score indicates declining Arousal — perhaps due to illness, personal stress, or prior poor sleep — can be identified for relief or workload reduction before FRMS would trigger. The combination of model-based prediction and real-time biometric observation is more robust than either alone.

Cognitive load monitoring: beyond fatigue

EchoDepth's VAD output provides cognitive load information beyond fatigue detection. The Arousal and Dominance dimensions together characterise operator cognitive state: a person managing high cognitive load shows elevated Arousal and variable Dominance, distinct from the low-Arousal, low-Dominance pattern of fatigue. In SOC environments, monitoring the cognitive load distribution across an analyst team enables real-time workload rebalancing — routing new alerts away from analysts showing high cognitive load toward those with available capacity.

This capability integrates with EchoDepth's operator readiness platform as a whole. The technical explanation of how FACS measures fatigue covers the AU-level detail of what the system detects and why.