You are on page 1of 5

International Journal of Management and Social Sciences Research (IJMSSR) ISSN: 2319-4421 1

Volume 3, No. 3, March 2014

Human Factors and Workload in Air traffic Control


Operations - A Review of Literature
Dr. Jitender Loura, Assistant Director, DGCA, Govt. of India

ABSTRACT was to control the signal timing throughout the city and
monitor the tunnels leading to Manhattan Island for
Purpose: The present paper focuses on the review of incidents. Currently, under the 1996Federal Highway
concepts, theories and empirical underpinnings of human Administration's Model Deployment Initiative, the New
factor in Air traffic control through literature studies York legacy traffic management centres are evolving to
primarily in the light of mental work load. This paper provide travellers with real-time travel information by
reviews what is currently known about workload, integrating of information provided by 14 different area
particularly from the field of ATC. One of the few areas of agencies. Legacy traffic management centres are often
almost universal agreement in the literature on ATC faced with the challenge of taking on new roles and
complexity is that complexity plays a major role in driving responsibilities, causing the need to increase in size and
controller workload. staffing. Whereas before these centres operated
Design/methodology/approach: Workload in ATC is independently, they now need to develop lilies of
generally mental, as opposed to physical, in nature. The communication and partnership agreements with other
paper will therefore briefly review what is known about agencies. Finally, one of the most critical problems faced
mental workload. by legacy centres is dealing with evolutionary technology
Findings: The findings of the paper implicitly point out upgrades. The technology used to monitor older sections
that workload factor include mental work load, ATC Task of the roadway may be completely different than the
Load Factors and ATC Operator Factors. Workload in technology used to monitor the newer roadway sections.
ATC is generally mental, as opposed to physical, in
nature. Task load i.e., the demand imposed by the ATC ATC involves a wide array of equipment, operational
task itself is generally distinguished from workload i.e., procedures, and many thousands of personnel in a variety
the controller’s subjective experience of that demand The of occupational categories. It is difficult to conceptualize
link between ATC task load and workload is a causative the system as a whole and to discuss human factors
(albeit indirect) one, that is influenced by a number of problems relating to the entire man-machine system at one
internal factors. time without other difficulties intruding. The principal
Originality/value: Based on literature review of 25 among these difficulties is the fact that the system is an
empirical articles, the paper makes Managerial ever changing mixture of men and machines. At the
implications that the workload factor include mental work present time, many en route and terminal facilities are
load, ATC Task Load Factors and ATC Operator Factors. being equipped with new NAS Stage A and ARTS I11
All researchers are sanguine that the relationship between hardware and software. However, much of the air traffic
task load and perceived workload can ever be adequately for the next few years will continue to operate within the
expressed. context of older ATC systems. Consequently, for a
substantial period into the future, there will be a mix
comprised of manual, semi automated, and fully
Key Words: automated operations for performing many of the most
Mental Work Load, ATC Task Load Factors, ATC important functions. Thus, highly specific
Operator Factors recommendations concerning studies to be conducted of
particular panels and displays, operational procedures,
INTRODUCTION communications techniques, and the like, would be
inappropriate in a report of this type. Furthermore, FAA
The most prevalent theme echoed in the literature on personnel (particularly those at NAFEC) have an excellent
traffic management centres was that each centre is unique. on-going program for conducting such studies and the
Generally speaking, two types of centres exist: (1) very developing recommendations and concerning specific
new centres and (2) legacy centres. The legacy centres man-machine interface problems. It seems likely that those
generally started smaller in size and staffing, lower in who are close to the operational situation on a day-to-day
technology, and more specific in their mission since they basis can most effectively identify, study, and recommend
were created for more immediate needs. One good solutions to these highly specific human factors problems.
example is the centre in New York City. Its initial purpose

i-Explore International Research Journal Consortium www.irjcjournals.org


International Journal of Management and Social Sciences Research (IJMSSR) ISSN: 2319-4421 2
Volume 3, No. 3, March 2014

THESIS STATEMENT in dividing attention. In what has been termed the


“cocktail party phenomenon,” people are able to attend to
This paper reviews what is currently known about mental only one source of information, unless some salient
workload, particularly from the field of ATC. One of the stimulus— such as their own name—is broadcast. The
few areas of almost universal agreement in the literature notion of channel capacity was adopted to explain
on ATC complexity is that complexity plays a major role limitations of the information processing system.
in driving controller workload. Workload in ATC is
generally mental, as opposed to physical, in nature. The Based on mounting evidence, theories of attention have
paper will therefore briefly review what is known about been significantly refined over the years. The 1970s saw
mental workload. the emergence of “resource models” of attention, which
postulate that all cognitive processes demand resources
that are available only in limited supply. Whereas channel
OBJECTIVE OF THE PAPER theories had assumed a structural bottleneck on
information processing, resource models held that
The vast complexity and the diffused nature of the air limitations were functional, in the form of attention or
traffic control systems makes it necessary to adopt some effort (Sanders, 1979). If task demands exceed available
organizing structure in order to obtain an overview for resources, performance declines. Conversely, if task
considering the system as a whole. The main objective of demands fall short of supply, then the amount of residual
the paper is to review the underpinning of human factors resource provides a measure of spare mental capacity.
in ATC in mental work load perspectives. Perhaps the most widely accepted current model of human
attention is the Multiple Resources model, as proposed by
MENTAL WORKLOAD Wickens (1980). According to this model, tasks differ on
the basis of demands they place in terms of: input
Interest in defining and developing metrics of mental modality (visual versus auditory); data input code (spatial
workload has grown dramatically since the mid-1970s /verbal) stage of processing(encoding/central/response);
(Sanders & McCormick, 1987; Wickens, 1980). Most and response type (manual versus verbal)characteristics.
attempts to define mental workload have grown by way of
analogy out of the concept of physical workload According to this multiple resource model, the degree of
(Meshkati, Hancock &Rahimi, 1990). interference between two tasks can be characterized by the
tasks' compatibility along the dimensions outlined above.
The lack of a clear definition is reflected in the This model has proven useful in predicting what sorts of
disagreement over appropriate metrics of mental tasks (do they rely on spatial or verbal information? Do
workload. It seems generally agreed that mental workload they require verbal or manual response? Is information
is not a unitary, but a multi-dimensional concept (Leplat, auditory or visual?) will interfere with one another.
1978; Moray, 1979; Kramer, 1991), that taps both the Further, workload in this model is the demand that tasks
difficulty of a task and the effort (both physical and place on a limited supply of attention resource.
mental) brought to bear (Gopher & Donchin, 1986). It
therefore represents an interaction between task and ATC TASK LOAD FACTORS
operator, that can vary for different task-operator
combinations (Leplat, 1978). Such factors as time Task load (i.e., the demand imposed by the ATC task
pressure, noise, stress, and distraction can all influence the itself) is generally distinguished from workload (i.e., the
'human costs' of performing a given task (Hancock controller’s subjective experience of that demand). A
&Chignell, 1988; Jorna, 1993). Aptitude, skill, experience, number of studies have attempted to identify traffic-related
operating behaviours, and personality traits have all been workload factors for ATC. Of many prospective task load
cited as determinants of subjective workload in ATC indices, the number-of-aircraft-under-control (i.e., traffic
(Bisseret, 1971; Sperandio, 1978). Clearly, the same given load) has shown the clearest predictive relationship to
task might represent a reasonable amount of workload for workload measures (Hurst & Rose, 1978;Stein, 1985). As
an experienced operator, yet overtax a novice. The with complexity, traffic density does not appear to fully
distinction is generally made between task load (the capture workload. Some of the other (airspace-related)
objective demands of a task) and workload (the subjective
• Sector flow organization (Arad, 1964);
ATC task load factors include:
demand experienced in the performance of a task).Inherent
in the notion of mental workload has been the concept that • Number of traffic problems (Kalsbeek, 1976);
the human operator has a limited capacity to process • Number of flight altitude transitions (Cardosi &
information. Information processing models of the 1950s
• Mean airspeed (Hurst & Rose, 1978);
Murphy, 1995);
grew out of the field of communications engineering.
• Sector area (Arad, 1964);
Experiments into “dichotic listening” by Colin Cherry in
the early 1950s demonstrated the difficulty humans have

i-Explore International Research Journal Consortium www.irjcjournals.org


International Journal of Management and Social Sciences Research (IJMSSR) ISSN: 2319-4421 3
Volume 3, No. 3, March 2014



Mean aircraft separation (Arad, 1964); of separation or, worse, a collision). How a controller
Aircraft mix (as it relates to differences in aircraft chooses to prioritise tasks, or the compensatory strategies


performance envelopes); used to respond to workload fluctuations (e.g., shedding or
Variations in directions of flight (Wyndemere, deferring tasks, and deciding which tasks to handle first, or
becoming more cautious in bad weather), all influence the

1996);
Proximity of aircraft and potential conflicts to controller’s workload(Koros et al., 2003). The research of
Hilburn&Jorna, 2001 depicted a simplified schematic of

sector boundaries(Wyndemere, 1996); and
Weather (Scott, Dargue&Goka, 1991; Mogford et the relationship between ATC task load and controller
al., 1994). workload. Though many other factors e.g., time pressure,
motivation, efforts are omitted in his research. The
Airspace factors are clearly not the only contributors to literature review also revealed a number of compensatory
ATC task load. Such other considerations as the ATC cognitive strategies that controllers use in accommodating
position (e.g., oceanic versus terminal (Wickens, Mavor& changes in complexity and workload.
McGee, 1997)), and the controller interface (including
both the visual display and the data entry system) are CONCLUSION
critical in determining a controller’s task load. They do not
always do so, however, in predictable or beneficial ways. The enormous complexity and the diffused nature of the
As research from other domains has demonstrated, a air traffic control systems makes it necessary to adopt
system’s interface itself can impose additional task some organizing structure in order to obtain an overview
demands. For instance, automated tools can have the for considering the system as a whole. The paper was
unintended effect of raising task load (Selcon, 1990; review on the workload factors task factor. All researchers
Kirlik, 1993). The potential for such situations appears are sanguine that the relationship between task load and
increasingly likely as more sophisticated “advisory” types perceived workload can ever be adequately expressed.
of decision aids emerge within ATC. By presenting the And it has been found a simplified schematic of the
controller the additional tasks of (1) considering the relationship between ATC task load and controller
system's advice, and (2) comparing the system's solutions workload. The literature review revealed a number of
to those he/she must continue to generate (if he/she is to compensatory cognitive strategies that controllers use in
remain “in the loop”), such decision aiding automation accommodating changes in complexity and workload.
might paradoxically force an additional task upon the
controller (Hilburn, Jorna&Parasuraman, 1995), or lead SCOPE FOR FURTHER RESEARCH
the controller to feel “driven” by the system (Whitfield,
Ball &Ord, 1980).
However, this study takes in to account some of the major
aspects of workload and task load factors but still there is
ATC OPERATOR FACTORS much more to do in this field of study. This study is
limited to workload and task load factors and its effect on
The link between ATC task load and workload is a ATC. Though various studies have shown a strong
causative (albeit indirect) one that is influenced by a relationship between complexity factors and controller
number of internal factors. In the past, attempts to assess workload but this aspect has not been undertaken in the
ATC workload have sometimes equated measures of direct study. Many other factors (e.g., time pressure, motivation,
task performance (such as time to perform discrete ATC effort) are omitted from this study.
tasks) with workload. Such observable workload
(Cardosi& Murphy, 1995), however, provides only a REFERENCES
partial picture of the workload experienced by a controller.
For instance, a controller’s observable performance cannot [1] Arad, B.A. (1964). The controller load and sector
always convey the cognitive task demands—such as design. Journal of Air Traffic Control, May, 12-
planning, decision making, and monitoring—imposed by 31.Attention and performance VIII. Hillsdale, New
ATC. Factors such as skill, training, experience, fatigue Jersey: Erlbaum.
and other “stressors” all mediate the relationship between [2] Bisseret, A. (1970). Operational memory and
task demands and the workload experienced by a structure of work. Bulletin de Psychologie, 24,
controller. Further, strategy plays an especially important 280-294.
part in determining a controller’s workload. Notice that [3] Bisseret, A. (1971). Analysis of mental processes
within very basic system demands, there are few involved in air traffic control. Ergonomics, 14(5),
constraints on how a controller should handle air traffic 565-570.
(Cardosi& Murphy, 1995; Leroux, 1992; Reason, 1988). [4] Bisseret, A. (1981). Application of signal detection
As result, the system can accommodate various control theory to decision making in supervisory control.
strategies without suffering a negative outcome (i.e., a loss Ergonomics, 24(2), 81-94.

i-Explore International Research Journal Consortium www.irjcjournals.org


International Journal of Management and Social Sciences Research (IJMSSR) ISSN: 2319-4421 4
Volume 3, No. 3, March 2014

[5] Buckley, E., & Stein, E. (1992).Simulation research Technical Publication TP96999. Amsterdam,
variable specifications (Unpublished manuscript). Netherlands:NLR.
[6] Cardosi, K. & Yost, A. (2001). Controller and pilot [19] Hilburn, B., Bakker, M.W.P., Pekela, W.D.
error in airport operations: A review of previous &Parasuraman, R. (1997). The Effect of Free Flight
research and analysis of safety data on Air Traffic Controller Mental Workload,
(DOT/FAA/AR-00-51). Washington, DC: FAA Monitoring and System Performance. Proceedings
Office of Aviation Research. of the Confederation of European Aerospace
[7] Cardosi, K. M & Murphy, E.D. (1995). Human Societies (CEAS) 10th European Aerospace
factors in the design and evaluation of air traffic Conference, 20-21 October 1997, Amsterdam, The
control systems (DOT/FAA/RD-95/3). Cambridge, Netherlands.
Massachusetts: U.S. Department of Transportation, [20] Hilburn, B., Jorna, P.G.A.M. &Parasuraman, R
Volpe Research Centre. (1995). The effect of advanced ATC strategic
[8] Fuld, R.B., Liu, Y., &Wickens, C.D. (1987). The decision aiding automation on mental workload and
impact of automation on error detection: Some monitoring performance: an empirical investigation
results from a visual discrimination task. In in simulated Dutch airspace. In Proceedings of the
Proceedings of the Human Factors Society, 31st Eighth International Symposium on Aviation
Annual Meeting (pp. 156-160). Psychology, April 1995, Columbus, Ohio: Ohio
[9] Gromelske, S., Davidson, L. & Stein, E.S. (1992). State University
Controller memory enhancement: Field facility [21] Hilburn, B., Jorna, P.G.A.M. &Parasuraman, R
concepts and techniques. DOT/FAA/CT-TN92/7. (1995). The effect of advanced ATC strategic
Atlantic City, New Jersey, USA: FAA. decision aiding automation on mental workload and
[10] Guttman, J.A., Stein, E., &Grimelski, S. (1995). monitoring performance: an empirical investigation
The influence of generic airspace on air traffic in simulated Dutch airspace. In Proceedings of the
controller performance.DOT/FAA/CT-TN95/38. Eighth International Symposium on Aviation
Atlantic City, New Jersey: FAA. Psychology, April 1995, Columbus, Ohio: Ohio
[11] Hancock, P. & Desmond, P. (Eds.) Stress, workload State University
& Fatigue. Mahwah, New Jersey, USA: [22] Hopkin, V.D. (1979). Mental workload
[12] Hancock, P.A. &Caird, J.K. (1993).Experimental measurement in air traffic control. In N. Moray
evaluation of a model of mental workload. Human (Ed.) Mental workload: its measurement and theory.
Factors,3, 413-429. New York: Plenum Press.
[13] Hancock, P.A., Chignell, M.H. & Kerr, G. [23] Hurst, M.W. & Rose, R.M. (1978). Objective job
(1988).Defining task complexity and task difficulty. difficulty, behavioural response, and sector
Presented at the XXIV International Congress of characteristics in air route traffic control centres.
Psychology, Sydney, Australia. Ergonomics, 21, 697-708.
[14] Hart, S. &Wickens, C.D. (1990).Workload [24] Jex, H.R. (1988). Measuring mental workload:
assessment and prediction. In H.R. Booher (Ed.) Problems progress and promises. In P.A. Hancock
MANPRINT: An emerging technology, advanced and N. Meshkati (Eds.), Human Mental Workload.
concepts for integrating people, machine, and Amsterdam: North Holland.
organization. New York: Van Nostrand Reinhold. [25] Jorna, P.G.A.M. (1993). The human component of
[15] Hart, S.G. &Stavelend, L.E. (1988). Development system validation. In JA Wise, VD Hopkin&P
of the NASA-TLX (Task Load Index): results of Stager (Eds.) Verification and validation of
empirical and theoretical research. In P.A. Hancock complex systems: human factors issues. Berlin:
& N. Meshkati (Eds.) Human mental workload. Springer-Verlag.
Holland: Elsevier.Harwood, K., Roske-H [26] Kalsbeek, J.W.H. (1976). Some aspects of stress
[16] Hilburn, B. & Flynn, M. (2001). Air traffic measurement in air traffic control officers at
controller and management attitudes toward Schiphol Airport. Symposium on stresses of air
automation: an empirical investigation. Presented at traffic control officers (University of Manchester,
the 4th USA/EUROPE ATM R&D Seminar. Santa Dept. of Postgraduate Medical Studies), 39-42.
Fe, NewMexico, USA. [27] Kirlik, A. (1993). Modelling strategic behaviour in
[17] Hilburn, B. &Jorna, P. (2001).Workload and Air human-automation interaction: Why an aid can and
Traffic Control. In PA Hancock and PA Desmond should go unused. Human Factors, 35, 221-242.
(Eds.) Stress, Workload and Fatigue: Theory, [28] Knecht, W., Smith, K. and Hancock, P. (1996). A
Research and Practice. Hillsdale, New Jersey, USA: dynamic conflict probe and index of collision risk.
Erlbaum. Proceedings of the Human Factors and Ergonomics
[18] Hilburn, B., (1996). The impact of ATC decision Society 40th Annual Meeting.
aiding automation on controller workload and [29] Koros, A., Della Rocco, P.S., Panjwani, G.,
human machine system performance.NLR Ingurgio, V., & D’Arcy, J.F. (2003). Complexity in

i-Explore International Research Journal Consortium www.irjcjournals.org


International Journal of Management and Social Sciences Research (IJMSSR) ISSN: 2319-4421 5
Volume 3, No. 3, March 2014

Air Traffic Control towers: A field study. Part 1: (DOT/FAA/CT-TN96-16). Atlantic City, NJ:
Complexity Factors. DOT/FAA/CT-TN03/14. DOT/FAA Technical Centre.
Atlantic City, NewJersey: FAA. [40] Sperandio, J.C. (1971). Variations of operators'
[30] Leplat, J. (1978). The factors affecting workload. strategies and regulating effects on workload.
Ergonomics, 21, 143-149. Ergonomics, 14, 571-577.
[31] Leroux, M. (1993).The PATS cooperative tools. In [41] Stein, E.S. (1985) Air traffic controller workload:
The proceedings of the scientific seminar on An examination of workload probe (Report
[32] Moray, N., Lootsteen, P., &Pajak, J. (1986). FAA/CTTN90/60). Atlantic City, New Jersey:
Acquisition of process control skills. IEEE FAA.
Transactions on Systems, Man, and Cybernetics, [42] Stein, E.S. (1987). Where will all our air traffic
16, 497-504. controllers be in the year 2001? In Proceedings of
[33] Murphy, E.D. &Cardosi, K.M. (1995).Issues in the 4th Symposium on Aviation Psychology. pp.
ATC automation. In Cardosi, K. M & Murphy, 195-201. Columbus, Ohio: Ohio State University.
E.D.(Eds.). Human factors in the design and [43] Stein, E.S. (1988). Air traffic controller scanning
evaluation of air traffic control systems and eye movements—A literature review.
(DOT/FAA/RD-95/3). Cambridge, Massachusetts: (DOT/FAA/CT-TN 84/24). Atlantic City Airport,
US Department of Transportation, Volpe Research New Jersey: Federal Aviation Administration
Centre. Technical Centre
[34] Parasuraman, R. (1987). Human computer [44] Stein, E.S. (1991). Air traffic controller memory: a
monitoring. Human Factors, 29(6), 695-706. field study (DOT/FAA/CT-TN90/60). Atlantic City,
[35] Pawlak, W.S., Brinton, C.S., Crouch, K. & New Jersey: FAA Technical Centre.
Lancaster, K.M. (1996). A framework for the [45] Stein, E.S. (1992). Air traffic control visual
evaluation of air traffic control complexity. scanning (FAA Technical Report DOT/FAA/CT-
Presented at the AIAA national TN92/16). Washington, DC: US Department of
Conference.PHARE.Braunshweig, Germany: DLR Transportation.
Institute for Flight Guidance. [46] Warm, J.S., Dember, W.N. & Hancock, P.A.
[36] Reason, J. (1988). Cognitive under-specification: its (1996). Vigilance and workload in automated
varieties and consequences. In B. Baars (Ed.)The systems. In R. Parasuraman& M. Mouloua (Eds.)
psychology of error: a window on the mind. New Automation and Human Performance: Theory and
York: Plenum Press. Applications. Mahwah, NJ, USA: Lawrence
[37] Selcon, S. J., & Taylor, R. M. (1989).Evaluation of Erlbaum.
the situational awareness rating technique [47] Whitfield, D, Ball, RG, and Ord, G (1980). Some
(SART)as a tool for aircrew systems design. human factors aspects of computer-aiding concepts
AGARD Conference Proceedings No. 478, Neuilly- for air traffic controllers. Human Factors, 22(5),
sur-Seine,France. 569-580.
[38] Selcon, SJ (1990). Decision support in the cockpit. [48] Wickens, C.D. (1980). The structure of processing
Probably a good thing? Proceedings of the Human resources. In R. Nickerson and R. Pew (Eds.)
Factors Society 34th Annual Meeting, pp. 46-50, [49] Wickens, C.D., Mavor, A.S. & McGee, J.P. (Eds.)
Santa Monica, California: Human Factors Society. (1997). Flight to the future: human factors in air
[39] Sollenberger, R. L., Stein, E. S., &Gromelski, S. traffic control. Washington, DC: National Academy
(1997). The development and evaluation of a of Sciences, National Research Council
behaviourally based rating form for the assessment [50] Wyndemere (1996). An Evaluation of Air Traffic
of air traffic controller performance Control Complexity, Final report contract NAS2-
14284. Boulder, Colorado USA: Wyndemere.

i-Explore International Research Journal Consortium www.irjcjournals.org