Content uploaded by Tom Kontogiannis
Author content
All content in this area was uploaded by Tom Kontogiannis
Content may be subject to copyright.
Cognitive Strategies in Emergency and Abnormal Situations Training:
Implications for Resilience in Air Traffic Control
Stathis Malakis
1
and Tom Kontogiannis
2
1
Hellenic Civil Aviation Authority, Rhodes/ Diagoras International Airport, Hellas
emalakis@otenet.gr
2
Cognitive Ergonomics & Industrial Safety Laboratory, Department of Production
Engineering & Management, Technical University of Crete, Chania Hellas,
konto@dpem.tuc.gr
Abstract. The management of emergencies and abnormal situations in the
Air Traffic Control (ATC) system entails substantial challenges for the air
traffic controllers. The inherent complexity and dynamics of the ATC
system can give rise to numerous cases of failure and controllers have to
develop failure-sensitive strategies to counteract and forestall any paths to
failure. Failure-sensitive strategies in the form of individual and joint
cognitive strategies may be considered as an important resource of
resilience in the ATC system. By hypothesizing that these failure sensitive
strategies are observable during the training in handling abnormal
situations, a field study was performed to elicit and document in a reliable
manner the controllers’ individual and joint cognitive strategies. This
paper summarizes initial findings with the potential of providing insights
in cultivating sources of resilience in the ATC system.
1 INTRODUCTION
Resilience represents the ability of a system to adapt or absorb disturbances, disruptions
and changes and especially those that fall outside the textbook operation envelope
(Woods et al, 2007). Emergencies and abnormal occurrences represent critical situations
close to the margins of safe operation that challenges the controller operational practices
and supervisory systems.The joint human and technical system is stretched to
accommodate new demands and this offers opportunities for studying aspects of system
resilience. In this sense, emergencies and abnormal situations are fertile grounds for
stories of resilience, which can stipulate human factors research.
The ATC system is a highly complex safety critical system with countless anticipated
and unanticipated paths to failure. As the controllers at the sharp end of system become
sensitive to the potential paths to failure, they develop failure-sensitive strategies to
counteract failure paths. An emergency presents controllers with many challenging
issues. Is the situation unusual and how far to pursue monitoring of the situation? As
soon as a disturbance is detected, a problem-to-be-solved is formulated and the need to
re-plan for the situation becomes prominent. To respond to an emergency, controllers
should demonstrate problem-detection skills and re-planning strategies. As an
occurrence evolves over time, new threats may appear whilst current threats may change
their demands.The need for gathering new information to fill in the gaps, correct
explanations, clarify assumptions and evaluate candidate hypotheses is amplified. This
calls for strategies in recognizing the situation, anticipating how the situation will evolve
in future, and how to manage uncertainty.
On the other hand, the joint performance of controllers and supervisory systems is also
challenged in an emergency. ATC requires synchronization of many inter-dependent
activities within a short time window and this calls for demonstration of joint cognitive
strategies. Coordination is the main prerequisite for synchronization but it comes at the
cost of information exchange. New tasks are added and ordinary prioritization is altered.
Therefore, increased workload must be balanced by intra-team reallocation of tasks. In
addition, safety critical situations are not tolerant to errors, which implies that controllers
should create their own opportunities for error detection and correction. These individual
and joint cognitive strategies can be seen as important sources of resilience in the ATC
system that would merit from a systematic classification.
Traditionally, the focus of controllers’ training has been on fulfilling regulatory
requirements. Effective handling of emergency and abnormal situations was considered
as a by-product of technical skills training. However, a growing number of recent
incidents and accidents in ATC and aviation have indicated that effective handling of
emergencies requires more than technical skills (Kirwan et al, 2005). A critical need
arises, therefore, to identify and document controllers’ failure-sensitive strategies as
sources of resilience (remaining sensitive to the possibility of failure). To this end, a
field study was undertaken to probe into the cognitive strategies of controllers in
emergency training.
2 METHOD
The research method was based on observations and ratings of human performance in
simulator-training scenarios for Ab-Initio students and operational controllers. This
method of identifying cognitive strategies and rating their quality was preferred to the
analysis of incident and accident reports that focus on technical aspects and operational
errors. Observational data were combined with qualitative data from briefing and
debriefing simulator sessions, focused interviews with controllers and instructors, and
finally a documentation analysis involving key operational documents and training
curricula. These research techniques belong to the “experiments in the field” family of
methods and are based on scaled world simulations that capture critical aspects of the
targeted situations (Woods & Hollnagel, 2006). Four Ab-Initio student and 22
operational area controllers, participated in the study. Student controllers were receiving
their unusual occurrences training and operational controllers were attending their
annual refresher training as part of their competency scheme.
In the first stage of the study, an inventory of cognitive strategies was compiled on the
basis of a literature review from the Naturalistic Decision Making (NDM) and Cognitive
Systems Engineering (CSE) paradigms. Four prominent sources of references were used
for the individual cognitive strategies.The Recognition Primed Decision (RPD) decision-
making model (Klein, 1998), the Recognition/Meta-Recognition (R/M) decision-making
model (Cohen et al, 1996) and the Contingent Operator Stress Model (COSMO)
decision-making model (Kontogiannis, 1999). The fourth was a model of anomaly
response as a multi-threaded process (Woods & Hollnagel, 2006). These models were
selected based on the plethora, the importance of the cognitive strategies they integrate
and the consistency of the research paradigm with the field study requirements.
For the identification of patterns of joint cognitive performance, a compilation was
made of four well-established frameworks from the same research paradigms. The first
one was the Anaesthetists’ Non-Technical Skills, (ANTS) which is a validated and
widely accepted framework (Fletcher et al, 2004). The second framework is more
generic and can be applied in any type of team and organization. The Big Five, (Salas et
al, 2005) is a teamwork model that has been developed by critical review of empirical
studies and theoretical models of teamwork, team effectiveness and team performance
over the last decades. The third framework is the NOTECHS (Non Technical Skills)
from the aviation domain. This framework was the outcome of a research project that
investigated possible ways to evaluate non-technical skills of multi pilot aircrew (Flin et
al, 2003).The fourth model is a taxonomy of classifying shared cognition breakdowns
(Wilson et al, 2007).
To provide a basis for rating human performance, several metrics of performance were
examined that looked deeper into elements of performance that should be demonstrated
as indicators of failure-sensitive cognitive strategies. For instance, planning strategies
could be rated by looking into the following metrics: standard planning (i.e., application
of standard operational action-scripts) and contingency planning (i.e., application of
non-formally prescribed precautionary action-scripts). The elements of each strategy
were balanced (i.e. two elements for each individual strategy and two or three elements
for each joint cognitive strategy) in order to avoid uneven coverage of the collected data.
3 RESULTS
For the ratings of controllers’ performance, we used a 7-point behaviourally-anchored
scale as it was thought to give a good rating sensitivity to subject matter expert
observers. The collected data were submitted to a Principal Component Analysis (PCA)
to establish the construct validity by revealing factor solutions that corresponded to the
hypothesized models of individual and cognitive performance. PCA was used in a
confirmatory factor analysis role and specific hypothesis were tested about the structure
and relationships between the factors that underlie the collected data. With regard to the
individual and cognitive strategies, a five-factor solution and a four-factor solution
respectively emerged that confirmed the hypothesized models of individual and joint
human performance. The five individual cognitive strategies and four joint cognitive
strategies that were identified in the PCA analysis of the 20 metrics of performance are
summarized in Table 1. The 20 metrics of cognitive strategies were illustrated with good
and poor exemplars (i.e., behavioural markers). This refinement of cognitive strategies
was based on interviews with controllers and instructors so that they were able to apply
this method on their own and achieve greater consensus in their judgement. An inter-
rater validity study is currently in progress to test the screening cognitive strategy tool
and promote greater use within the ATC environment.
Table 1. Cognitive and Joint Cognitive Strategies & Metrics of Performance
Individual Cognitive
Strategies
Metrics of Performance
Recognition Noticing Distinguishing Cues
State Projection
Managing Uncertainty Critiquing Situational Models
Critiquing Goals
Planning Standard Planning
Contingency Planning
Anticipation Threat Acknowledgement
Exploiting less Busy Periods to Perform Planning
Managing Workload Prioritizing Tasks
Interruption Management
Joint Cognitive
Strategies
Metrics of performance
Coordination Team Coordination
Shared Situation Understanding
Intent Communication
Information Exchange Unsolicited Dissemination of Proactive information
Provision of Updates on Situation Status & Management
Ensuring an Undisrupted & Ungarbled Information Flow
Error Management Detection of Errors by Other Team Members
Provision of Feedback to Enable Error Correction
Workload Distribution
Management
Detection of Workload Distribution Problems
Situation Driven Reallocation of Tasks
The five individual and four joint cognitive strategies that corresponded to the
hypothesized models as emerged using PCA are analyzed below.
Anticipation: A cognitive strategy that enables a controller to timely and accurately
detect and respond to a threat. Anticipation engages with response planning during low
tempo periods. It is the process of recognizing and preparing for difficult challenges and
brings forward the notion of threats. Threats can be defined as events or errors that occur
beyond the control of the controller and must be managed in order to maintain the
margins of safety.
Recognition : A cognitive strategy that enables a controller to timely and accurately
detect early signs of an impending emergency and play out mentally the progression of
events. Emergencies and abnormal situations can occur suddenly when the flight crew
formally declares an emergency or may evolve slowly over time. In the first case,
recognition is effectively reduced to the level of an accurate classification of the
emergency type (i.e. a symptom-fault matching). In the latter case when the emergency
is evolving over time, a pattern of cues is available.
Managing Uncertainty: A cognitive strategy that enables a controller to timely and
accurately assemble and assess a model of the situation and establish safety related
goals. Emergencies and abnormal situations are closely associated with information-
based uncertainty due to their dynamics. The controller has to assemble a model of
situation, formulate goals and correct any tentative explanations or assumptions seeking
information that may not be available or accessible. Flight crews are in general reluctant
to provide conclusive information during emergencies and communication with the ATC
is not their first priority. Even if they are willing to communicate their status, this may
not be technically feasible.
Planning : A cognitive strategy that enables a controller to employ standard and/or
contingency planning for the unfolding situation. Controllers have to make a plan and in
certain cases to re-plan their actions in order to cope with the demands of the unfolding
situation. Planning may have the form of standard and/or contingency planning.
Depending on the situation, a minimal set of prescribed action-scripts in documented
forms (e.g. checklists) are normally available in all ATC units. Controllers are trained in
certain types of emergencies and this annual process is a major part of their competency
scheme. Nevertheless, in many cases the need for contingency planning arises. It may be
a textbook case of an abnormal situation but certain characteristics may warrant an
additional form of precautionary planning (contingency planning) in order to counteract
a possible escalation of the situation.
Managing Workload: A cognitive strategy that enables a controller to timely and
accurately sequence the required tasks and respond to interruptions and distractions.
From the onset of an emergency, the workload increases significantly due to a notable
increase in the number of tasks, the available time and the importance of the tasks to be
completed. Workload management functions as a mental task regulator enabling
controllers to cope with the complexity of the situation. Issues related to switching
attention between normal and situation related tasks as well as judging interruptibility
are regulated by workload management.
Coordination : It refers to the extent to which controllers direct and coordinate other
team members, establish situation assessment congruence and clarify intent. The
structure of a team (as defined by the nature of the team’s tasks and their allocation) can
generate lateral (intra-team) and vertical (inter-team) dependencies, which require
coordination to achieve orchestrated action. The importance of coordination
requirements increases with the severity of the unfolding situation. The building blocks
of coordination are the shared mental models and the concept of intent. The more
congruent the shared mental models of a team, the more congruent the situation
assessment and performance of a team. ‘Communication of intent’ serves to fill-in any
gaps and/or tentative assumptions of the shared mental models.
Information Exchange : It refers to the extent that proactive information is disseminated
between controllers and regular updates are made on the situation status without
disruptions and garbling. Coordination requirements generate a pressing need for
communication. Communication is depended mainly on information exchange and
requires both sufficient time and cognitive resources to be accomplished. The team
members exchange information to articulate their planning, their actions and their
responsibilities. Therefore, the role of information exchange is crucial to the ability of
team to achieve coordinated action and perform effectively in critical situations.
Error Management: It is the extent to which controllers can develop task monitoring
and/or augmented monitoring strategies that enable them to detect errors and provide
feedback for error correction. In handling critical situations, errors can be committed
that vary from minor ones to major ones that complicate the situation and reduce the
safety margins. Errors can be detected and corrected not only at the individual level but
also more effectively through the team structure. The error detection process is based on
monitoring strategies, which ran parallel to the normal tasks at the cost of cognitive
resources (mainly attention and memory). In mature teams, the members employ
efficient monitoring strategies that have been crafted during years of operational day-to-
day experience and accumulated expertise in handling the available systems. These
monitoring strategies enable them not only to “catch” promptly an error but also to
correct it and/or provide feedback for error correction without hindering their individual
and the other team-members’ flow of tasks.
Workload Distribution Management: It is the extent to which controllers have
developed workload balancing strategies that enable them to detect and counteract
workload problems of team members. Workload is not a constant parameter and it
naturally follows the changing requirements of the escalation pattern of a critical
situation. The steeper the escalation pattern, the steeper the increase of the workload for
both the controllers. The task sequence may be altered while new tasks (those induced
by the critical situation requirements) are added in the task backlog. The controllers have
to manage, not only the normal traffic in their sector, but also the critical situation and
the interactions between them. The criticality of the situation increases and diversifies
the normal distribution of the workload and generates imbalances between the tasks of
the controllers. Therefore, a critical need arises for the implementation of strategies that
balance and keep the workload below saturation point for all members of the operating
team.
4 CONCLUSION
Hollnagel and Woods (2006) argued that we can measure the potential for resilience but
not resilience itself. In line with this reasoning, we conclude that these failure-sensitive
cognitive strategies provide important practical examples of the potential for resilience
in two levels. Firstly, by providing insights on how adaptations by local actors in the
form of cognitive strategies are employed to support resilience in cases of safety critical
events. Secondly by using these cognitive strategies as the foundation blocks in the
development of an advanced safety training program with the aim of cultivating sources
of resilience in the ATC system.
REFERENCES
Cohen, M.S., Freeman, J.T. & Wolf, S.P. (1996). Meta-Recognition in time stressed
decision making: Recognizing critiquing and correcting. Human Factors, 38, 206-219.
Fletcher, G., Flin, R., McGeorge, P., Glavin, R., Maran, N., & Patey R. (2004). Rating
non-technical skills: Developing a behavioural marker system for use in anaesthesia.
Cognition Technology and Work, 6, 165-171.
Flin, R., Martin, L., Goeters, K., Hoermann, J., Amalberti, R., Valot, C., & Nijhuis, H.
(2003). Development of the NOTECHS (Non-Technical Skills) system for assessing
pilots’ CRM skills. Human Factors and Aerospace Safety, 3, 95-117.
Hollnagel, E. & Woods, D. (2006). Epilogue: Resilience Engineering Precepts. In E.
Hollnagel, D.D. Woods, & N. Leveson (Eds.). Resilience Engineering: Concepts and
Precepts. Ashgate Publishing. Aldershot. UK.
Kirwan, B., Rodgers, M., & Schafer, D. (Eds.). (2005). Human Factors Impacts in Air
Traffic Management. Ashgate Publishing. Aldershot UK.
Klein, G.A. (1998). Sources of Power: How people make decisions. Cambridge, MA:
MIT Press.
Kontogiannis, T. (1999). Training effective human performance in the managing of
stressful emergencies. Cognition Technology and Work. 1, pp7-24.
Salas, E., Sims, D.E., & Burke. C.S. (2005). Is there a “big five” in teamwork? Small
Group Research. Sage Publications, 36, 555-599.
Wilson, K.A., Salas, E. Priest, H.A., & Andrews, D. (2007). Errors in the Heat of
Battle: Taking a Closer Look at Shared Cognition Breakdowns through Teamwork.
Human Factors. 49, 243-256.
Woods, D.D. & Hollnagel E. (2006). Joint Cognitive Systems: Patterns in Cognitive
Systems Engineering. CRC Press. Taylor & Francis Group. Boca Raton FL.
Woods, D.D., Patterson, E.S., & Cook, R.I. (2007). Behind Human Error: Taming
Complexity to Improve Patient Safety. In P. Carayon. (Ed.). Handbook of Human
Factors and Ergonomics in Health Care and Patient Safety. Lawrence Erlbaum
Associates, Mahwah, New Jersey.
