Operator Situational Awareness
Designing displays around Endsley's three levels — perception, comprehension, and projection — so operators of both machines always know what is, what it means, and what comes next.
Three levels of awareness
Situational awareness (SA) has a well-established structure: Level 1 is perceiving the raw state, Level 2 is comprehending what it means, Level 3 is projecting what happens next. A control room that only shows numbers serves Level 1 and leaves comprehension and projection to the operator's memory — exactly the load that fails during an upset. Kronos L6 explicitly designs a surface for each level.
| SA level | Question | L6 answer |
|---|---|---|
| 1 Perception | What is the state? | Live twin overlay + KPIs from L2/L3 |
| 2 Comprehension | What does it mean? | Anomaly context, envelope margins, confidence |
| 3 Projection | What happens next? | Predictive shadow + precursor lead times |
Projection is the differentiator
The twin runs a 50-100 ms predictive shadow ahead of the real plant, and the anomaly ensembles quote lead times before an event. L6 turns both into a projection surface: the operator sees not just that plasma current is 9.66 MA now, but the twin's forecast trajectory and how close it is drifting to the safe operating envelope. For the burner, the same surface projects end-plug density and ambipolar potential forward so an operator can see a plug excursion forming before it trips. See predictive shadow visualization.
Keeping the envelope visible
Comprehension collapses when constraints are invisible. Every key variable is drawn against its certified envelope with the current margin explicit, so the operator reads distance-to-limit, not just value. Margins shrinking is the earliest honest signal that attention is needed, well before an alarm threshold trips.
- Perception without comprehension is a number wall — avoided
- Comprehension needs margins and confidence shown, not implied
- Projection uses the twin shadow and precursor lead times, labeled as forecasts
- SA is measured in training with SAGAT-style probes and fed back into layout
SA is fragile under cognitive load, so display density is tuned against the load budget — see human factors and cognitive load — and forecasts always show their uncertainty via uncertainty visualization.