Attention, Vigilance and Distraction

The mind is always attending to something, except in sleep. The danger for a pilot is rarely a lack of attention; it is attention in the wrong place. The crew that circles in the dark working on a warning light while the aeroplane descends, the pilot who stares at the chart while the heading wanders, and the crew chatting during taxi who line up on the wrong runway all had their attention fully occupied, just not with the task that mattered.

Attention is the limited capacity that decides what information is processed and which task gets effort. Vigilance is the ability to keep it on a task over time, so as to catch rare but important signals. Distraction is whatever pulls it away. Human performance training treats all three as limits to be managed, not failings of character, and much of standard operating procedure, from the sterile cockpit to the written checklist, exists to protect attention.

Attention as a limited resource

Attention works with the information processing chain: it selects what passes from the sensory stores to perception and working memory. Because the capacity is limited, ATPL texts identify poor management of attention, attending to the wrong item at a given time, as the major attentional danger for pilots.

How much attention is available depends on arousal, the readiness to respond. Performance follows an inverted U against arousal, the Yerkes-Dodson relationship: an under-aroused pilot scans slowly and misses things, an over-aroused pilot narrows attention and makes errors, and the best performance lies in between. The more complex the task, the lower the arousal at which it starts to degrade, so the moment a flight becomes complicated is the moment to simplify it. See stress.

View full schematic

Selective and divided attention

Human performance texts recognise two types:

• Selective attention samples inputs continually for relevance and lets the important ones through. A pilot's own callsign is especially attention-getting.
• Divided attention time-shares the central decision maker between tasks, as when a pilot on a visual approach alternates between the lookout and the instruments.

Divided attention is not true parallel processing. The pilot switches between tasks, and the tasks that coexist smoothly are those run as well-learned skills, which need little attention.

Single-channel processing

In the single-channel model of attention, the central decision maker can handle only one decision at a time. What feels like multi-tasking is rapid switching, and each switch costs time. When the demand exceeds the channel's capacity, tasks are shed, often without the pilot noticing which ones, and the task dropped is not necessarily the least important. A pilot hand-flying a demanding approach may simply fail to process ATC calls, however clearly they are transmitted.

The defence is to reduce demand rather than to try harder: aviate, navigate, communicate, in that order. Keep the flight path under control, tell ATC to "stand by", delay non-essential tasks, use the autopilot to free capacity, and move work such as briefings and avionics set-up to quiet phases. See workload management.

Saliency and the cocktail party effect

The cocktail party effect is the ability to pick out one's own name from the background of many conversations. It shows that unattended inputs are screened for relevance at an early stage. In the cockpit it lets a pilot hear the aircraft's callsign through radio chatter, or notice a smell of burning while busy with something else.

Saliency works the other way. Attention is drawn most strongly to cues that are loud, bright, recent, centrally placed and easy to interpret, so vital information that is none of these can be overlooked. Auditory stimuli attract attention more readily than visual ones, but under extreme stress auditory information is commonly the first to be discarded, which is why a warning can sound clearly and still go unnoticed. At Kegworth in 1989 the vibration indicator of the failed engine read maximum, but the crew shut down the other engine without assimilating it: it was a small display on the new electronic engine instruments, and the AAIB recommended adding an attention-getting feature to draw the eye to a vibration indicator at maximum.

Fixation and attentional tunnelling

Fixation, or channelised attention, is the concentration of attention on one source of information to the exclusion of all others. A student pilot, uncertain of position, spends four minutes head-down comparing chart and ground and looks up to find the aeroplane 20° off heading and 400 ft low. Under high arousal the effect is automatic: attentional narrowing, or tunnelling, contracts the range of cues sampled, and warning flags and standby instruments go unseen. The instrument pilot knows the same failure as one of the three classic scan errors: fixation (staring at one instrument), omission (leaving one out) and emphasis (relying too heavily on one).

ATPL texts describe blinkered concentration as becoming so involved in a single readout that situational awareness is lost, and point out that in the vast majority of cases an alarming reading can be cross-checked against other instruments.

The textbook case is Eastern Air Lines Flight 401 in December 1972. On approach to Miami the crew of a Lockheed L-1011 found that the nose gear indicator had not shown green, entered a hold at 2,000 ft and became absorbed in the indication. The autopilot's altitude hold disengaged, the aeroplane began a slow descent that nobody monitored, an altitude alert chime went unanswered, and the crew noticed the lost height only seconds before the aeroplane struck the Everglades. The lesson taught ever since is that somebody must always fly the aeroplane, and that a minor problem does not need every pilot on board.

Cognitive lockup

Cognitive lockup is the tendency to stay with the task already in hand rather than switch to a new and more urgent one. It is most likely when the current task is engaging and partly successful: a pilot nursing a rough-running engine with mixture and carburettor heat may not address the line of cumulonimbus developing on track. The countermeasure is a deliberate, periodic step back, asking what else is happening and what will matter in ten minutes, and, in a crew, making sure that the pilot flying is not the one solving the problem.

Vigilance and hypovigilance

Vigilance, sustained alertness to infrequent signals, is one of the foundations of situational awareness. It declines when there is too little to do. On a long, quiet, automated sector the scan slows, attention wanders and vigilance for rare events decays, the vigilance decrement: radio calls are missed and a slowly developing fuel imbalance goes unnoticed. Time on task works by driving arousal down, so underload is a genuine hazard rather than a rest. Humans are poor at sustained monitoring of a system that is usually right, and automation complacency, a drift into passive monitoring, is the result. The countermeasure is to manufacture work: a disciplined scan cycle at fixed intervals, regular fuel and systems checks, log entries, position and weather updates, and periods of hand-flying.

Hypovigilance, in EASA human performance texts, is a state in which sleep patterns appear on the EEG while the person seems to be working. Monotony, low workload, simple repetitive tasks, constant monotonous noise, low light, warmth, isolation, sleep debt, fatigue and a recent meal all promote it. Crew contact, moving about and staggering meals help for a while; only sleep removes the cause. See fatigue.

Interruptions and distractions

An interruption leaves a gap, not a pause. In well-practised, highly automatic tasks such as reading a checklist, an unexpected interruption is frequently followed by an omission, because it breaks the sequence and the pilot is a poor judge of where it stopped. ATPL texts list the main sources of checklist error as responding automatically, seeing what one expects to see rather than what is there, external interruptions such as a radio call, and simple slips of the place-marker. After a checklist interruption, resume only from an item positively confirmed as completed, or restart the checklist, as the operator's procedures require.

Interruptions also end reviews. At Kegworth the commander's attempt to review the engine symptoms was interrupted and never resumed. Distraction management follows the same principles as workload management: one pilot keeps flying while the other deals with the distraction, the problem is delegated rather than shared by both, the automation is used to buy capacity, and head-down tasks are done in short glances with the flight path checked in between.

The sterile cockpit

The sterile cockpit rule removes avoidable distraction from the phases where it is most dangerous. Under 14 CFR 121.542, applied to commuter and on-demand operators by 135.100, no crewmember may perform any duties during a critical phase of flight except those required for the safe operation of the aircraft, and the pilot in command may not permit any activity that could distract a crewmember. Critical phases are all ground operations involving taxi, take-off and landing, and all other flight operations below 10,000 ft except cruise flight. That excludes non-essential conversation, paperwork, eating and non-operational calls, and since a 2014 amendment, 121.542(d) also bars Part 121 flight crew from personal use of wireless devices and laptops at the duty station during flight time.

EASA's air operations rules (Regulation (EU) No 965/2012, Annex I) define the critical phases of flight for aeroplanes as the take-off run, the take-off flight path, the final approach, the missed approach and the landing, including the landing roll, plus any other phase determined by the pilot-in-command or commander. Under CAT.GEN.MPA.105 the commander must not permit any crew member to perform any activity during those phases except duties required for the safe operation of the aircraft.

Two NTSB investigations show why. Before the 2006 wrong-runway take-off of Comair Flight 5191 at Lexington, the crew held a conversation unrelated to the flight during taxi; the NTSB named that non-pertinent conversation, which resulted in a loss of positional awareness, as a contributing factor. In the 2009 Colgan Air Flight 3407 accident, the crew continued an unrelated conversation after descending through 10,000 ft, and failure to adhere to sterile cockpit procedures was a contributing factor. A single pilot can apply the same discipline, briefing passengers before departure to stay quiet during taxi, take-off, approach and landing.

Library articles are written for study and exam preparation. They do not replace your aircraft's approved documentation, your operator's procedures or the regulations themselves.