Loss of situation awareness is often discovered only when intervention is required
Aliases: latent awareness loss · takeover exposure · recovery fragility
What it is
Latent situation-awareness failure is a cognitive gap hidden by normal automation performance until anomaly handling or takeover requires human judgement. Task success does not prove process understanding when automation maintains the outcome.
Why it happens
Closed-loop automation isolates human model error from execution, preventing feedback. At intervention the protective layer weakens and human understanding becomes control input; discovering mode, environment, or goal error then leaves no reconstruction margin. Low incident rates can hide fragile recovery.
Studying it
After long nominal runs, varied anomalies should measure pre-cue awareness, takeover correctness, reconstruction, and near misses. Normal performance and anticipated anomalies overestimate ability. Safe freeze probes reveal gaps before deployment.
Where it stops holding
Frequent probes disrupt work and fatigue, so sampling follows risk. If automation safely handles an anomaly, forced takeover is unnecessary, but recovery paths still need audit. Awareness scores also do not guarantee retained control skill.
Applying it
- Use low-frequency context checks and simulated takeover during normal operation.
- Increase process information and participation early in degradation rather than transfer at the limit.
- Track pre-event model accuracy, stable-control recovery, and tail failures in safety review.
Related
- Same group: X4.07.1 Situation awareness covers current state, change trends, and future projection · X4.07.2 Situation awareness decays during long uneventful supervision · X4.07.3 Both excessive and insufficient information density weaken situation awareness
- Adjacent: X4.05 Takeover and handoff · X4.02 Human intervention
- Search terms:
latent situation awareness failure·takeover performance·recovery capability