Pilot performance in emergencies: why it can be so easy, even for experts, to fail

On February 4, 2015, about 1054 Taipei Local Time, TransAsia Airways (TNA) flight GE 235, an ATR72-600 aircraft, registered B-22816, experienced a loss of control during initial climb and impacted Keelung River, three nautical miles from its departing runway 10 of Taipei’s Songshan Airport. Forty-three occupants were fatally injured, including three flight crew, one cabin crew, and 39 passengers. The remaining 13 passengers and one cabin crew sustained serious injuries. One passenger received minor injuries. The aircraft was destroyed by impact forces. The aircraft’s left wing tip collided with a taxi on an overpass before the aircraft entered the river. The taxi driver sustained serious injuries and the only taxi passenger sustained minor injuries.


Photo: Frame of dashboard camera video excerpted from the Final Report- Use of the video was authorized by the TVBS

The accident was the result of many contributing factors which culminated in a stall-induced loss of control. During the initial climb after takeoff, an intermittent discontinuity in engine number 2’s auto feather unit (AFU) may have caused the automatic take off power control system (ATPCS) sequence which resulted in the uncommanded autofeather of engine number 2 propellers.

Following the uncommanded autofeather of engine number 2 propellers, the flight crew did not perform the documented abnormal and emergency procedures to identify the failure and implement the required corrective actions. This led the pilot flying (PF) to retard power of the operative engine number 1 and shut down it ultimately.

The loss of thrust during the initial climb and inappropriate flight control inputs by the PF generated a series of stall warnings, including activation of the stick shaker and pusher. After the engine number 1 was shut down, the loss of power from both engines was not detected and corrected by the crew in time to restart engine number 1. The crew did not respond to the stall warnings in a timely and effective manner. The aircraft stalled and continued descent during the attempted engine restart. The remaining altitude and time to impact were not enough to successfully restart the engine and recover the aircraft. (See: TransAsia Airways Flight GE235 accident Final Report)


Being the continuation of  When the error comes from an expert: The Limits of Expertise


“Emergencies and other threatening situations require pilots to execute infrequently practiced procedures correctly and to use their skills and judgment to select an appropriate course of action, often under high workload, time pressure, and ambiguous indications.

The performance of even the most skilled experts can be impaired by situational stress, applying equally to the skilled performance of almost all experts, from surgical teams to firefighters.

The term stress refers to the effects and the term stressful situations refer to the causes of a well-defined picture of two neural/hormonal systems that respond to the threat with characteristic changes that prepare the body for “fight or flight” e.g. increased heart rate and hard breathing. However, the psychological mechanisms associated with stress are less clear.

The effects of stress can be explained using the cognitive appraisal model (Lazarus and Folkman, 1984)when individuals encounter challenging situations they orient both their cognitive and physiological resources to deal with the situation.

Physiological responses, such as increased heart rate and force, faster breathing, and restriction of peripheral blood flow, prepare the body for ‘fight or flight.’ Cognitively, the individual focuses attention on the challenging situation, mentally preparing for whatever tasks may be required. Up to this point, the individual’s resources are mobilized to deal with the challenge, the individual can manage the situation effectively and performance may actually improve.

However, if the situation becomes threatening—physically or socially—and the individual is uncertain of his or her ability to manage the threat, anxiety arises. This anxiety plays a central role in altering the individual’s cognitive processes and overall performance, and is maladaptive because it disrupts the individual’s ability to manage the threatening situation, particularly by degrading attention and working memory, both of which are crucial for managing challenging situations effectively (Eysenck, Derakshan, Santos, and Calvo, 2007).”

Stress not necessarily directly caused accident pilots’ errors, but stressful conditions made these errors more likely to occur.

Attention and Working Memory

“Attention is the focus of one’s mind on one task or thought or stream of sensory input from a myriad of other possibilities. Basically, we can only fully attend to one stream of information at a given moment. If we must deal with multiple tasks, we are forced to switch attention back and forth among them, somewhat like a spotlight.

Working memory is a very small subset of the vast store of an individual’s long-term memory, momentarily activated so that it can be quickly accessed and manipulated. It consists of two components: the information stored and the control processes used to manipulate the information. These control processes are known as executive processes and are also involved in directing attention.

Attention and working memory are known as limited cognitive resources; their capacity for processing information is quite small compared to the vast store of information in long-term memory.

The content of working memory (that is, short-term memory store) is generated by the interaction of perceptual input with activation of a very small portion of long-term memory. Central executive processes and some involuntary processes, such as the startle reflex, control movement of the spotlight of attention over this limited store and update its content, holding task-relevant information readily available and updating that information.

The effects of anxiety on attention and working memory are consistent with the attention control theory (Eysenck, Derakshan, Santos, and Calvo,2007).

Attention is known to be controlled by two different brain systems: a top-down system that directs attention to support the individual’s currently active goals, and a bottom-up system that draws attention to environmental stimuli, especially stimuli that are salient, abrupt, or threatening. Attention control theory posits that anxiety disrupts the balance between the two attentional systems, giving the bottom-up system more weight. Consequently, attention is less under the control of task goals and is more easily pulled away by salient or threatening stimuli. Thus, the individual is more easily distracted from task goals. However, if the threatening stimuli are central to the task’s goals, focusing might actually be improved.

Individuals under stress are less able to manage their attention effectively. They are more likely to be distracted from a crucial task by highly salient stimuli, such as an alarm, or by threatening aspects of a situation. They may process information less fully and may have difficulty switching attention among multiple tasks in a controlled fashion, and consequently, their management of the overall situation may become disjointed and chaotic.

Because anxious thoughts tend to preempt working memory’s limited storage capacity, the individual may have difficulty performing computations that would normally be easy and have difficulty making sense of the overall situation and updating the mental model of the situation (i.e. situation awareness). In studies of accidents, by far the most common -23%- category of errors involved inadequate comprehension, interpretation, or assessment of the ongoing situation.

To understand how stress affects the skilled performance of pilots, especially in emergencies (which by their nature involve novelty, uncertainty, and threat), one must understand the distinction between the automated performance of highly practiced tasks and the effortful performance of less familiar tasks that draws heavily on attention and working memory. If the threat produces anxiety, pilots’ performance is likely to be undermined in specific ways.

Attention and working memory are essential for tasks involving novelty, complexity, or danger. Performing tasks requiring these two limited cognitive resources is typically slow and effortful. Highly practiced skills, such as manual operation of flight controls, are less vulnerable to stress because they are largely automated and are less dependent on attention and working memory. Studies show that inadequate execution of a physical action occurred only in <5% of accidents.

However, emergencies almost always require interweaving highly practiced tasks with less familiar tasks, novel situational aspects, and uncertainty. Thus, in an emergency situation, overall demands on attention and working memory are very high at a time when these limited cognitive resources may be disrupted by anxiety; consequently, tasks such as decision-making, team performance, and communication that depend heavily on attention and working memory are likely to be impaired.”


Photo:  GE235 loss of control and initial impact sequence, excerpted from the Final Report


“Decision-making under stress becomes less systematic and more hurried, and that fewer alternative choices are considered when making decisions. However, in highly practiced situations experts make decisions largely by automatic recognition of the situation and retrieval of the appropriate response from the long-term memory of previous experiences. This is why pilots are required to practice responding to some emergency situations. Thus, experts such as pilots are protected from impairment from stress under very familiar situations, at least to some extent.

For example, airline pilots are often given an engine failure during recurrent simulator training, and so pilots are typically fairly reliable in executing the appropriate response when experiencing an actual engine failure emergency in flight, even though the situation is somewhat stressful.

Unfortunately, most emergency situations are not rehearsed. Even in cases where the emergency procedures are practiced, the decisions that the pilot needs to make to respond appropriately in a particular emergency may be unique, and thus the required decision-making is not rehearsed. For example, the immediate responses to an engine fire in flight are practiced in recurrent training and are likely to be fairly reliable. But, the decisions about the next steps to take depend on where the aircraft is, fuel remaining, weather, and many other variables. Consequently, deliberate thought is required about these aspects, and such necessary deliberation may be impaired by the stress that is induced during the emergency.

The decisions made by pilots involved in accidents are often criticized. Indeed it is easy to identify, after the fact, what the pilots could have done to avert the accidents. But, as have previously argued, that kind of assessment suffers from hindsight bias (Dismukes, Berman, and Loukopoulos, 2007). In current studies of accident errors, has been found relatively few examples of poor decision making or poor choice of action. Therefore, it is suspected that —at least in the case of experienced airline pilots— “poor decision-making” may be used as a catch-all category, and it is suggested investigations would be better served by a deeper analysis of underlying cognitive factors.”

Team Performance and Communication

“Under acute stress team members search for and share less information, tend to neglect social and interpersonal cues, and often confuse their roles and responsibilities. Stress hinders team performance, including decision-making, primarily by disrupting communication and coordination. Coordination, of course, lies at the heart of effective team performance. Stress significantly reduces both the number of communication channels used and the likelihood that teammates will be provided needed information.

Poor communication and coordination can lead to downstream errors by team members. Studies found that 14% of errors in accidents involved inadequate or improper communication, 17% involved poor management of the competing task demands, and another 17% involved inadvertent omission of required actions. It is suspected that most of the errors of all types may have resulted from an underlying cause already mentioned: disruption of pilots’ executive control of attention and working memory.”

Ways to Reduce Error Vulnerability

“The design of airline operating procedures, training, and cockpit interfaces have evolved and improved steadily over decades of operational experience. However, there could be a hidden vulnerability in the design of three crucial aspects of safety—operating procedures, training, and interfaces—when non-normal situations are encountered. There seems to be an implicit assumption by designers that experienced pilots in emergency situations will be able to perform “normally”: that is to say pilots are assumed to process information, communicate, analyze situations, and make decisions as well as if they were sitting safely on the ground. That assumption is wrong.

Therefore the vulnerability to make errors in stressful situations could be reduced by developing tools to help flight crews:

  1. Recognize, interpret, assess and comprehend the full implications of a challenging situation that may change dynamically.
  2. Keep track of where they are in a procedure or checklist.
  3. Shift attention among competing tasks without becoming locked into just one task.
  4. Identify and analyze decision options.
  5. Step back mentally from the moment-to-moment demands of the flight situation to establish a high-level (meta-cognitive) mental model that guides action.
  6. Continuously update that mental model as the situation unfolds.
  7. Maintain the cognitive flexibility to abandon a previously selected procedure or course of action that has become inappropriate for the situation.

To a large degree, these seven objectives could be supported by revising existing flight deck operating procedures, checklists, and training to reflect diminished attention control and working memory function in threatening situations. This would best be accomplished by collaboration between human factors experts and the operational community. In addition, a longer-range approach would be to support these objectives in the design of future flight deck displays and automation interfaces.

Pilots’ resilience to stressful situations could also be improved by stress exposure training. In its simplest form this training would explain the physiological and cognitive changes that occur in stressful situations, which might help pilots be less disconcerted when they experience the physiological effects and be on guard for the cognitive effects. More advanced training could be incorporated into existing Line Oriented Flight Training (LOFT), allowing pilots to examine their own performance in stressful scenarios.”

Implications for NextGen

“The NextGen environment will present flight crews with operating procedures and demands that could increase stress and the consequences of stress, especially in non-normal situations. Complexity and traffic density will increase in this environment, and thus margins for error and time to respond may decrease. Therefore, it is crucial to identify human factors challenges that may arise during implementation and to develop appropriate countermeasures.

The increased navigational precision and reduced aircraft spacing required for NextGen may sometimes reduce the time flight crews have to interpret emergency situations and to select appropriate courses of action. The complexity of choosing an appropriate course of action may also increase for crews encountering emergencies because options may be constrained while conducting NextGen operations, such as closely spaced parallel operations.

New technologies will generate new failure modes that may increase stress and cognitive demands on flight crews. Research would allow these failure modes to be characterized, well-anticipated, and thoroughly covered in training that is designed to mitigate stress effects on flight crew performance in the NextGen context. Existing alerting features on flight decks may not be adequate for NextGen procedures and failures.

As the airspace system evolves and grows more complex and crowded, the need for ways to help flight crews deal with the heavy cognitive demands of non-normal situations becomes even more important. Transition to complex new technologies poses human factors challenges, and those in NextGen are particularly critical to its successful implementation. Difficulties will be worked out as they appear, but the transition period, including learning new procedures to proficiency, is likely to be especially cognitively demanding on flight crews; thus realistic simulation research to characterize the human factors challenges and develop mitigations should be conducted before NextGen systems are fielded. After NextGen technologies are in operation, it will be important to carefully monitor operations for indicators of latent human factors problems, particularly related to the effects of stress in normal and non-normal operations.”


  1. Excerpted from Dismukes, R.K., Goldsmith, T.E., & Kochan, J.A. (2015).  Effects of Acute Stress on Aircrew Performance: Literature Review and Analysis of Operational Aspects.  NASA Technical Memorandum TM-2015-218930.  Moffett Field, CA: NASA Ames Research Center.
  2. Stress and Flightcrew Performance: Types of Errors Occurring in Airline Accidents, R. Key Dismukes, Janeen A. Kochan, and Timothy E. Goldsmith, July 2014
  3. Aviation Safety Council Taipei-Taiwan Aviation Occurrence Report, 4 February 2015 TransAsia Airways Flight GE235, ATR72-212A, Loss of Control and Crashed into Keelung River Three Nautical Miles East of Songshan Airport. Report Number: ASC-AOR-16-06-001Date: June 2016. English report released on July 1st, 2016.


  1. When the error comes from an expert: The Limits of Expertise
  2. Multitasking in Complex Operations, a real danger
  3. Shutting down the wrong engine
  4. TransAsia Airways Flight GE235 accident Final Report


minime2By Laura Duque-Arrubla, a medical doctor with postgraduate studies in Aviation Medicine, Human Factors and Aviation Safety. In the aviation field since 1988, Human Factors instructor since 1994. Follow me on facebook Living Safely with Human Error and twitter@dralaurita. Human Factors information almost every day 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s