Human Factors In Air Traffic Control

V. D. Hopkin

• Publisher: CRC Press
• Year: 2017
• ISBN:
  9781351440172
• Edition: 1
• Topic: Technology & Engineering
• Subtopic: Industrial Health & Safety
• Index: Technology & Engineering

This text discusses the skills and abilities that air-traffic controllers need. Its approach is international as air-traffic control practices throughout the world have to be mutually compatible and agreed.

1

Historical introduction

1.1 Human factors

The origins of Human Factors as a profession with its own domain of expertise are not all self-evident. Various influences have fostered its emergence as an independent discipline and contributed to its consolidation, but the gifts of hindsight are needed to retrace them. It was not apparent beforehand that these influences would coalesce into the single discipline called Human Factors, any more than it is apparent now what the long-term future of Human Factors as a discipline will be. In retrospect, the following themes deserve some mention as formative influences on Human Factors.

Human Factors can be subsumed under Applied Psychology, particularly if Applied Psychology is interpreted, according to the distinction drawn by Hearnshaw (1987), as the extension of psychological methods to real problems rather than the extrapolation of laboratory findings to real-life settings. Early texts on Industrial Psychology (e.g. Myers, 1929) heralded many topics that have become familiar human factors themes, although their connotations have changed. Industrial Psychology evolved into Occupational Psychology as its influence spread to non-industrial jobs, and a broader range of factors was accepted as relevant. Sometimes Industrial Psychology and Organizational Psychology are equated (Reber, 1985), although the latter now extends beyond industrial and non-industrial jobs to embrace a diversity of social structures (Bradley and Hendrick, 1994). Ergonomics, evolving contemporarily with Human Factors, also recognized the mutual interactions of humans and equipment within the workspace, but usually viewed them from a more interdisciplinary perspective (Murrell, 1965).

A different kind of influence on Human Factors was the treatment of humans and machines as comparable components of a human–machine system. The principles of scientific management enunciated by Taylor (1911) and the study of human movement patterns at work (Gilbreth, 1919) which developed into time and motion study (Barnes, 1958) exemplified this approach, with its quest to improve productivity and motivation and its development of techniques for job analysis and design. Similar measures were applied to describe humans and machines. The machines might be modified to increase the pace, smoothness and efficiency of human movements, but not solely to satisfy human aspirations. However, some individual human characteristics, associated for example with skill, fatigue or errors, were represented in Craik’s (1947) pioneering theory of the human as an engineering system, which was in tune with the beginnings of information theory, cybernetics and computer technology (Turing, 1950). The classification and allocation of functions as more suitable for humans or machines, which was originally formulated for application in air traffic control (Fitts, 1951a), has endured far longer than its originators advocated, for they foresaw that technological advances could invalidate rigid function allocation. They did not recognize such inherent limitations of the approach as its competitiveness (Jordan, 1968), its exclusion of some kinds of human–machine relationship (Hopkin, 1982a), and its over-emphasis on those human functions with a feasible machine equivalent.

A further influence came from the performance of the human tasks that resulted from technological innovations in equipment, especially within large systems. The main impetus originated during the Second World War, particularly with radar displays and vigilance tasks (Mackworth, 1950), when human rather than equipment limitations seemed to determine what was achievable. The extensive early psychological research on equipment design sponsored by the United States Armed Forces was initially labelled Engineering Psychology or Human Engineering (McCormick, 1957), then was called Human Factors Engineering, and eventually became designated as Human Factors, partly because of the eponymous journal. Originally, the main purpose of Human Factors was to aid design: the Tufts College handbook of Human Engineering Data (1949) and Woodson’s (1954) text were for designers to use, and Sinaiko’s (1961) collection of papers emphasized design. Fitts’ chapter about equipment design in Stevens’ definitive Handbook of Experimental Psychology (Fitts, 1951b) testifies alike to the already established role within experimental psychology for Human Factors, and to the confusion over what it should be called. Perhaps it was applied experimental psychology (Chapanis et al., 1949). Human Factors became concerned extensively with the development and evaluation of the human contributions to the functioning of large systems, expressed in systems terms (Meister and Rabideau, 1965). Many studies of large systems as functioning entities were conducted, including studies of air traffic control systems, yet it often proved difficult to trace whether the findings produced by such major efforts and lavish resources had any significant lasting influence at all on the system design or functioning (Parsons, 1972), and studies of that magnitude became less common.

A more human–centred influence on Human Factors concerned the attributes of work that could satisfy human needs and aspirations. Job satisfaction and job enrichment, the quality of working life, and work as a social activity were studied, and theories propounded (Herzberg, 1966; Maslow, 1976). Morale, degree of autonomy, status, responsibilities, self-esteem, and the esteem of colleagues seemed relevant. The significance of team roles, ethos, and professional norms and standards was sometimes conceded grudgingly, when more traditional and orthodox influences had failed to account adequately for what actually occurred. While the motives for acknowledging their significance were occasionally humanitarian, they were more commonly associated with the unacceptable costs of high staff turnover rates, strike-prone workforces and poorly motivated staff. Users’ attitudes, as well as technical efficiency, could be crucial to the successful implementation of changes, although their willing adoption of new equipment did not guarantee that they would actually use it in accordance with the designers’ intentions.

Another influence was the study of individual differences, generally for the purpose of curtailing them as a source of unwanted variance. Selection procedures sought to choose from a wider population a more homogeneous group possessing identified measurable attributes. Training procedures then instilled common knowledge, skills and practices so that all individuals performed tasks similarly. The allocation of people to their initial jobs, and subsequent career development procedures, acknowledged that some residual individual differences remained, and attempted to utilize them by matching individuals with job requirements, thereby reducing further the individual differences between people doing the same job.

Operators often had little control over the task demands imposed on them by the system. The effects of systems on the operators who worked within them constituted another seminal influence on Human Factors. These effects, whether couched in such system concepts as efficiency, safety and system integrity, or in such human concepts as stress, boredom, health and well-being, were mediated by general differences between categories of people, as in their age or their experience, and by differences between individuals, as in their capabilities, adaptability or tolerance.

A final kind of influence on Human Factors dealt with organizational settings and contexts. Broad examples include conditions of employment, management practices, legislative requirements and organizational structures. More specific aspects are work-rest cycles, policies on supervision and assistance, the provision of supporting services, opportunities for mutual help, and physical and environmental features of the workspace.

Gradually Human Factors as a discipline has evolved to embrace all the above influences, to permit them to interact, and to absorb further influences as they arise. Recent additional influences have come from computer sciences and artificial intelligence, from cognitive theories, from changed social expectations about work and its rewards, from cultural ergonomics, and from repercussions on legal responsibility or financial accountability.

1.2 Air traffic control

When commercial aviation began, aircrew lacked ground-based navigation aids. To avoid becoming lost, they relied on direct observation, their own deductions and simple instruments (Williams, 1990; Kendal, 1990). Only in the immediate vicinity of airports was there any assistance, as the experiences of pioneer aviators make clear (Saint-Exupéry, 1939). The Second World War, with its requirements for air operations in poor visibility and at night, led to rudimentary air traffic control based on procedural principles similar to some of those still followed beyond radar coverage. The safe, orderly and expeditious flow of air traffic became established as the objective of air traffic control. Safety was paramount. Aircraft receiving an air traffic control service would remain safely separated from each other at all times, and would not collide with the ground or with other obstacles. Their pilots would never become lost, but would be guided to their destinations and warned while airborne of hazards such as severe weather.

Towards the end of the Second World War, the Convention on International Civil Aviation, subsequently dubbed the Chicago Convention, led to the formulation of agreed standard practices as a prerequisite for international air traffic control, and to the founding of the International Civil Aviation Organization (ICAO) as the regulatory body for the specification and implementation of common air traffic control procedures and practices and for compliance with them. The need for international agreements and collaboration was thus acknowledged from the early days of air traffic control. It predates most technological advances and the vast expansion of air traffic and air users. A judicious balance between national sovereignty and the international regulation of air traffic was struck, taking account of national differences in geography, traffic demands, and the political and financial priorities accorded to air traffic control. At the end of the Second World War, aeronautical, technical and navigational advances had made long flights feasible and reliable, aircraft construction had become a major industry, airfields that could handle many large aircraft had been built, and plenty of experienced pilots were keen to continue flying.

The greatest initial demand for commercial flights was between main centres of population, and the most direct routes were naturally preferred. Many of these routes were eventually marked by ground-based beacons emitting signals that could be sensed by aircraft flying along the route. Separate and approximately parallel routes might be allocated to aircraft flying in opposite directions. Alternatively, minimum permissible lateral separations were applied between opposite direction traffic within the same route, airway or air corridor. Aircraft in level flight were separated also by height, with different flight levels for different types of aircraft and for traffic flying in different directions. Minimum longitudinal separations between consecutive aircraft at the same flight level on the same route could be expressed as distances or times, with some provision for faster aircraft to overtake slower ones, subject to an adequate lateral or height separation between them. Safe separation between aircraft involves the three spatial dimensions and time. The actual magnitudes of the separations required for safety are not universal but depend on the nature and quality of the navigational information.

When the advent of radar improved the quality of that information, the minimum lateral and longitudinal separations between aircraft could safely be reduced, so that the controller could handle more air traffic within a given airspace while complying with the new standards. Beyond radar coverage, for example in oceanic regions, very large separations between routes and between consecutive aircraft at the same height on the same route still had to be maintained. However, the information about consecutive aircraft on final approach to the same runway at an airport might be of such high quality that the minimum separation between them was no longer determined by it but by the regulations that required a following aircraft to be sufficiently separated from the aircraft ahead of it not to be put at hazard by wake vortices and turbulence.

As the number of aircraft and the demand for air traffic control services increased, so did the workload of the controller. The concept of the sector was applied to the region of airspace, defined by geographical and height boundaries, within which one controller or one team of controllers was responsible for providing the air traffic control service. For a time, further increases in demand could be accommodated by reducing the size of sectors. However, the handover of responsibility for the control of each aircraft as it left one sector and entered the next imposed communications workload on both the controllers and on the pilot. At some point, further partitioning of sectors becomes counterproductive as a response to increased traffic demands because of the extra coordination and liaison tasks introduced by sectorization. Other solutions have to be devised that enable the controller to maintain safe separations between aircraft and provide an efficient air traffic control service without becoming overburdened.

Several developments have seemed helpful. One is the improvement in navigational information, expected to continue with the future provision of data derived from satellites. Another is the automation of tasks, particularly those associated with the routine gathering, assimilation, collation, updating and discarding of information. A further development is the provision of computer assistance for decision making, for problem solving and for prediction, often by presenting ready-made solutions for the controller to accept, reject or modify, instead of requiring the controller to fulfil these functions unaided. A current trend is the evolution from tactical to strategic air traffic control, associated with planning in advance to prevent problems from arising rather than resolving problems that have already arisen, and with organizing air traffic into flows rather than dealing with aircraft singly.

The extent of direct human involvement in air traffic control is a further issue. Technological and computer advances, particularly in software, continually extend the feasible options. Fewer technological limitations now impose inappropriate roles on humans simply because no machine can fulfil them. Many air traffic control functions could in principle be wholly automatic or wholly manual, or could employ any intermediate stage of automated assistance. A broader range of questions about suitable future roles for controllers and pilots can be posed because the technological means to implement alternative policies on preferred human and machine roles are becoming available. Technology has also revived the issue of how far air traffic control should remain exclusively ground-based, when sensors on board aircraft can detect other nearby aircraft and direct how to avoid them. Such a tactical aid is not easily reconciled with the evolution of air traffic control towards the longer-term planning of traffic flows, which presumes that single aircraft will not manoeuvre unexpectedly at short notice.

Developments in navigation, in communications, in computer technology and software, in system planning and strategic control techniques, in artificial intelligence, in human–machine interface design, and in many other technical fields combine to offer a wealth of options for future air traffic control. Few of these developments are exclusively for air traffic control. For most of them, air traffic control is one of many possible applications. Some technological innovations may have no air traffic control applications. Others may seem feasible yet bring no major benefits. A few may lead to spectacular improvements. Almost all will require considerable adaptation before they can benefit air traffic control fully.

Air traffic control is so complex that any innovation can be expected to introduce both advantages and disadvantages, so that its adoption must depend on a favourable balance between them. It is therefore vital to identify all of them before any changes are made. Safety takes precedence. Any lapses make headlines, a reflection alike of their potentially catastrophic consequences and of their current rarity and resultant newsworthiness. Air traffic control must evolve to cope with future traffic, but its safety record must be maintained and preferably enhanced.

1.3 Human factors in air traffic control

The emergence of Human Factors as a discipline cannot be traced to any single event, but the application of Human Factors to air traffic control can be. Although a few small studies related to air traffic control preceded it (Hopkin, 1970), the report by Fitts and his colleagues (1951a), which formulated a long-term programme of human factors research on air traffic control, became a classic reference, and was widely cited for many years even by those with no interest in air traffic control. Fitts deliberately proposed research capable of yielding findings that could be applied broadly and would not be invalidated by relatively trivial system changes. As a result, his report has been applied beyond air traffic control to many large human–machine systems, and his principles and recommendations, including the assignment of roles to men and to machines that gained the sobriquet “Fitts’ list”, have been treated as universals, though not by Fitts himself who was much more circumspect. With some minor changes of terminology but not of substance, almost all the research objectives in Fitts’ report are still valid, a testament to its prescience and perspicacity in discerning key issues and to the universality of many human factors problems within large human–machine systems.

Fitts advocated a systems approach, but it was not as narrow as the approach actually adopted in air traffic control (Parsons, 1972) and elsewhere (Meister and Rabideau, 1965), although the need to consider broader influences was recognized and voiced (Sinaiko and Buckley, 1957). In all these early human factors studies on air traffic control, the main emphases were on the performance of tasks and the selection of controllers. The limitations and underlying assumptions of their mechanistic approach were sometimes recognized, but even criticisms of the approach as “Procrustean” (Taylor and Garvey, 1959) achieved few modifications of it. Nevertheless much valuable and durable work on such topics as workspaces and task designs was accomplished. A series of formal inquiries into the origins of widespread disquietude and grievances among air traffic controllers, particularly in the United States in the 1970s, exposed a gulf between the real human factors issues deemed by controllers to be important and the issues actually addressed during systems design or in contemporary research. Human factors studies of air traffic control paid insufficient heed to preferences or individual differences, tended to presume that every question must have a single optimum answer (Hopkin, 1970), and lacked subtlety in their treatment and classification of controllers’ skills (Older and Cameron, 1972).

Over 20 years ago, the human factors literature on air traffic control had already become diffuse and rather inaccessible (Hopkin, 1970), and no comprehensive bibliography of it existed. Despite modern storage and retrieval systems, this remains true and for the same reasons. Much relevant work is never published or widely disseminated or else appears in obscure sources. While the results of simplistic laboratory studies of dubious relevance to real life were being applied, the findings from major air traffic control systems experiments that employed prodigal facilities and resources languished, even after Parsons’ (1972) description and critique of them.

Human factors studies of air traffic control form two main categories. Some studies belong to continuous work programmes extending over many years, which utilize dedicated air traffic control simulation facilities and in-house resources or employ contractors under the auspices of national or international agencies. Other studies apply the known relevant expertise of a contractor or academic department to a particular air traffic control problem for a short time, without becoming involved in wider air traffic control issues. This dichotomy has characterized the human factors studies on air traffic control in most nations, including the United States, which potentially has the largest resources for such studies. The influence of psychology on these air traffic control studies has been on their measures and methodologies and their reliance on experimental data, rather than on the application of psychological theori...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• List of figures
• Preface
• 1. Historical introduction
• 2. The air traffic control system
• 3. The human as a system component
• 4. Human cognitive capabilities and limitations
• 5. Matching human and system
• 6. Human factors contributions during air traffic control system evolution
• 7. Air traffic control jobs and tasks
• 8. The selection of controllers
• 9. The training of controllers
• 10. The work environment
• 11. Air traffic control displays
• 12. Input devices
• 13. Communications
• 14. Forms of computer assistance
• 15. Common human factors implications of computer assistance
• 16. Effects of the system on the individual controller
• 17. Conditions of employment
• 18. Measurement
• 19. Research and development
• 20. Human factors implications of other functions related to air traffic control
• 21. The future
• List of sources
• References
• Index