Editorial: Ergonomics and Human Factors in Aviation

Introduction

Ergonomics and Human Factors (E/HF) in Aviation is essential for the safety and efficiency of commercial airlines, passenger, cargo and military operations, and for the well-being of their passengers. However, it also extends beyond the aircraft to air traffic control and management, maintenance, regulatory bodies and policy makers. E/HF has a long history of innovations in theory, methodology, science and application. For example, approaches have evolved from the examination of the activities of individual pilots, through to crew resource management to considering entire aviation systems and their emergent properties. Similarly, ergonomics methodologies have moved from focussing on individual tasks to entire systems, the constraints shaping behaviour and the culture of organisations. The aim of this special issue is to provide researchers and practitioners with an opportunity to discover the very latest research trends for E/HF in Aviation.

In a previous special issue of Ergonomics on this topic, Harris and Stanton (2010) made the point that aviation is a system of systems. Some of this complexity is characterised in Figure 1, which shows that the aviation sociotechnical system comprises airports, aircraft, airlines, air traffic management (ATM) and air traffic control (ATC). All of these systems interact with each other within the rules and regulations of the aviation authorities around the world. The papers within this special issue offer good coverage of topics across the systems highlighted in Figure 1, including ATC, passenger comfort, automation, pilot mental workload, hypoxia, crewing of aircraft, design of controls and displays, and crew resource management.

Figure 1. Aviation as a system of systems.

The main themes for this special issue on E/HF in Aviation are: system approaches, crew resource management, helicopter operations, design of input/output devices, and last, but not least, passenger comfort. Each will be presented in turn.

Systems approaches

The systems paradigm is gaining ground in Human Factors and Ergonomics (Salmon et al. 2017; Stanton and Harvey 2017; Stanton, Salmon, and Walker 2017; Walker et al. 2017) as represented in this special issue. Indeed, the importance of the systems approach has been recognised in aviation (Harris and Stanton 2010; Schmid and Stanton 2018). Methods have been developed to present a systems perspective, such as EAST and STAMP. Stanton, Plant, Revell, Griffin, Moffat and Stanton (Distributed cognition in aviation operations: a gate-to-gate study with implications for distributed crewing) present the findings of an observational study of cargo operations. The researchers recorded flight operations for three outbound and two inbound flights over a five-day period. From this, they constructed task, social and information networks to represent four key phases of flight: (i) pre-flight cheques and engines start (ii) taxi and take-off, (iii) descent and landing, and (iv) taxi, park and shutdown. The networks present a detailed analysis and work audit of contemporary operations. This analysis led the authors to consider what task, social and information network requirements that might need to be for distributed crewing of the future (see Stanton, Harris, and Starr 2016). As the analysis shows, it is not as simple of just moving one of the pilots to a ground-based station; rather it involves a comprehensive reconceptualization of the work from a variety of perspectives (e.g. dispatch, maintenance, loading, ATC, ATM and the flight deck). A complete reallocation of roles and tasks over the four key phases of flight is required if distributed crewing in cargo operations is to be successful.

It has been argued that the distributed crewing concept is technically feasible for commercial aviation with the technology that is currently available (Harris 2007; Stanton, Harris, and Starr 2016). However, to make this a reality will mean redesigning both flight operations and the aircraft. It is not simply a matter of removing one pilot from the aircraft (see Stanton et al.: Distributed cognition in aviation operations: a gate-to-gate study with implications for distributed crewing). In addition, all of the emergency scenarios need to be assessed to see if the distributed crew can manage them effectively. Distributed crewing needs to be as least as safe as the current multi-crewing system (Stanton, Harris, and Starr 2016). Revell, Allison, Sears and Stanton (Modelling distributed crewing in commercial aircraft with STAMP for a rapid decompression hazard) use the System Theoretic Accident Model and Processes (STAMP) approach together with the System-Theoretic Process Analysis (STPA) to model the situation in the Helios hypoxia incident involving a Boeing 737-300. STAMP was used to model the current and potential future distributed system comprising the regulator, airline, crew, aircraft, ATC/ATM and (in the future) a ground station crewed by another pilot. The STPA was used to model and compare the performance of the current multi-crew operation with distributed crew operations for a rapid decompression event. The findings suggest that recovery from rapid decompression involving hypoxia on the pilot(s) on the aircraft is more likely with distributed crewing. As such, the addition of a ground station and the distributed crewing concept has the potential to improve safety in commercial aviation. The authors recommend that other emergency scenarios should be investigated to see if these benefits are generalisable.

Accident reports revealed that landing was the most critical stage in flight operations. Human error is the most significant contribution to aviation safety (Stanton et al. 2009). Many studies have indicated that the pilot is the last defence of safety in flight operations. Currently, there are few pieces of research focussed on the pilot’s performance and operational environment based on real flight data. The Quick Access Recorder (QAR) data can record all kinds of flight and operational parameters of an aircraft. The CAAC regulations require that all commercial aircraft shall install airborne QAR to monitor pilots’ performance and systems’ behaviours. Wang, Zhang, Dong, Sun and Ren (A method of applying flight data to evaluate landing operation Performance) aim to develop a method of evaluating a pilot’s performance during the landing phase using QAR data. The verification process demonstrates that the model is suitable for evaluating landing performance; the extended application shows that this model could be used to forecast potential risks associated with the landing phase across an airline fleet. Furthermore, the model is able to compare the landing performance among different fleets. The Flight Landing Operation Performance Evaluation System (FLOPES) was developed based on the evaluation model. It could be used as a practical tool for airlines to manage landing risk quantitatively and to provide practical support to improve training and safety management systems in aviation.

The Human Factors Analysis and Classification System (HFACS) is currently the most widely used human error-coding framework for accident and incident analysis in the aviation industry. The approach is upon Reason’s organisational model of human error which suggests that there is a ‘one to many’ mapping of condition tokens to unsafe act tokens. It is this ‘one to many’ relationship that is problematic for predicting human error outcomes from their organisational precursors. Traditional statistical approaches cannot be used in such instances however Harris and Li (Using neural networks to predict HFACS unsafe acts from the pre-conditions of unsafe acts) demonstrate that artificial neural networks may be a viable approach for the prediction of the HFACS categories of unsafe acts from their immediate psychological pre-cursors. They argue that this approach is also commensurate with theoretical basis of error proposed by Reason. Using an artificial neural network, in an analysis of 523 aircraft accidents, it was observed that almost three-quarter of all unsafe acts could be predicted from information about their psychological preconditions. It is argued that this analytical approach reflects the complexity of modern operations where there are many interacting factors at work. Neural networks have the potential to predict simultaneously many outcomes from many input factors reflecting the complexity of the system as a whole, rather than characterising errors as the outcome of a simple ‘cause and effect’ relationship.

Advances in aviation safety have improved significantly in more recent years (Harris 2011). One important contributor to this improvement has been the identification of factors involved in accidents and eliminating the causes where possible. Therefore, analysis of near misses is an important part of safety management activities. In addition, learning from successful safety outcomes, or what went right, is an important emerging component of maintaining safety systems. Accordingly, there are increasing calls to study normal operational performance in near misses as a part of safety management activities. Despite this, there is limited guidance on how to accomplish this in practice. Thoroman, Goode, Salmon and Wooley (What went right? An analysis of the protective factors in aviation near misses) demonstrate that Rasmussen’s risk management framework can be used to identify networks of protective factors that prevent accidents. Rasmussen’s theory of accident causation in sociotechnical systems describes the processes involved in accident prevention, this was undertaken through an analysis of 16 serious incident reports using Nvivo, a qualitative software tool. It is concluded that protective factors occur throughout the sociotechnical system, which support accident prevention. Near misses appear to share many of the same properties as accidents. While it is possible to identify protective factors from current investigation reports, this limits the information available on the influences of normal work. Safety practitioners can use the framework described to discover and support a system-wide network analysis of protective factors.

Crew resource management

In aviation, non-technical skills (i.e. those skills not related directly to control or the aircraft but are required to interact with other people and make decisions) are as important as the skills required to fly the aircraft. Crew Resource Management has had been successful in the training of non-technical skills but has been criticised for not integrating this with the technical skills training. Typically, the training of non-technical skills is undertaken after the technical skills have been mastered. Arguably, these skills should be trained together as they are dependent on each other. Ideally, the training of both sets of skills would be integrated into the operational environment, as is the case in commercial airlines Line Oriented Flight Training (LOFT). Tsifetkis and Kontogiannis (Evaluating non-technical skills and mission essential competencies of pilots in military aviation environments) seek to correct this position by identifying behavioural markers for mission essential competencies for military F-16 pilots. The mission essential competencies were identified through a review and workshop with pilots as mission planning, flying tactics, battle tactics, weapon handling, information management, uncertainty and de-briefing. Four broad classes of non-technical skills were found to be reliable and valid predictors of performance across the six mission essential competencies (i.e. managerial skills, team cooperation, situation assessment and decision making). This also demonstrated the strong relationship between technical and non-technical skills. The findings reinforce the point that the training of both types of skills should be better integrated rather than one preceding the other.

Crew Resource Management (CRM) has become a ‘way of life’ in the aviation industry; it evolved as an operating concept after a series of accidents where the main cause was attributed to a failure to utilise effectively the human resources available on the flight deck. CRM introduced the disciplines of applied social psychology and management science into the cockpit. Bennett (The training and practice of crew resource management: recommendations from an inductive in vivo study of the flight deck) utilised an approach not often utilised in Human Factors studies to examine CRM in action. Using an ethnographic approach, he observed CRM in a naturalistic setting using non-participant observations, gathering data during 20 airline flight sectors. This gives an insight into the role and processes of CRM not usually reported in Human Factors journals. Using this real-world approach, it was observed the CRM processes were robust even under moderately stressful situations and that it served to facilitate and shape social relations on the flight deck. However, while CRM has promoted safe operations on the flight deck, and as a result, its principles have also been adopted by other professions in aviation (e.g. air traffic control; maintenance and ramp operations), it was observed that poor inter-profession co-ordination and teamwork still occurred. It was suggested that a more system-wide approach to aspect of CRM awareness was required.

To many, the job or airline pilot would seem to be glamorous, an extremely desirable occupation to which many aspire. However, there is also a down side. The demands on professional pilots are steadily increasing in many respects. Significant time away from friends and family; anti-social hours; disturbed sleep patterns and high levels of responsibility to name but a few. The paper by Demerouti, Veldhuis, Coombes and Hunter (Burnout among pilots: psychosocial factors related to happiness and performance at simulator training) found that in a large sample of airline pilots surveyed, 40% experienced high burnout. This affected their performance in the flight simulator, where recurrent cheques were made on their capabilities as a pilot to maintain their licence. Burnout did not impact upon their flying performance per se. The authors report that pilots experiencing higher levels of burnout were less inclined to use stratagems that could help them to improve their work characteristics. Job crafting strategies, such as seeking challenges and optimising demands, were found to be used more frequently by pilots performed better during their simulation check rides. Pilots experiencing burnout were also less happy with their life in general. It was concluded that pilots’ work/life imbalance and future insecurity were of major concern.

The measurement of mental workload is an E/HF metric used when assessing the utility of any system. There are several basic methods by which mental workload may be determined but two of the most common are based upon physiological measures related to ECG (electrocardiogram) and subjective rating techniques (based upon simple univariate of more complex multi-dimensional scales). All such techniques have advantages and disadvantages, based around their sensitivity, diagnosticity and task intrusiveness (which is essential to avoid during fast-paced, safety-critical tasks). Mansikka, Virtanen and Harris (Comparison of NASA-TLX scale, Modified Cooper-Harper scale and mean inter-beat interval as measures of pilot mental workload during simulated flight tasks) used the NASA-TLX, modified Cooper-Harper (MCH) scale and the ECG-derived inter-beat interval (IBI) to evaluate fighter pilot workload in a series of operational scenarios flown in a high-fidelity flight training device. All measures of workload were correlated and there was an explicit link demonstrated between pilot performance and IBI, NASA-TLX and the MCH scale. All measures were able to differentiate the high-performance category participants (based upon Instrument Landing System performance scores) from the medium- and low-performance categories. In addition, the NASA-TLX and MCH were also able to differentiate the medium- and low-performance. IBI measures were not able to differentiate the medium-performance category from the low-performance category suggesting that while they are less task-intrusive, they are also less sensitive.

Helicopter operations

Helicopter operations are over-represented in aviation accidents and very constrained by reduced visibility conditions for safe operation (Stanton, Harris, and Starr 2016). Unlike their fixed wing counterparts, currently, helicopters do not operate with an Instrument Landing System (ILS: a precision runway approach aid based on two radio beams which together provide pilots with both vertical and horizontal guidance during an approach to landing). Without this aid, the pilots of helicopters are heavily dependent on cues outside the cockpit to assist in landing their aircraft safely. Stanton, Plant, Roberts and Allison (Use of highways in the sky and a virtual pad for landing head up display symbology to enable improved helicopter pilots situation awareness and workload in degraded visual conditions) explore the use of computer-generate synthetic cues to replace real cues in conditions of very poor visibility. Normally, helicopters would be prevented from operating in these conditions but the augmentation of the external visual environment with head-up cues can greatly assist pilots in finding the airport and landing the aircraft. In their study, Stanton et al. used highway in the sky ‘hoops’ to assist pilots in navigating their way to the landing pad. In addition, other instruments were displayed head-up to help the pilots manage torque, air speed, ground speed, pitch, yaw, roll, drift, altitude and vertical speed. The landing zone was annotated with a virtual pad and pyramid markers (to assist with depth perception). Helicopter pilots flew the route with and without the head-up display in both clear and foggy conditions. The new head-up display did not interfere with flights in clear condition and showed considerable improvements in the foggy conditions. Such displays could extend the operational capability of helicopters in the future.

Martin and Nixon (Helicopter pilots’ views of air traffic controller responsibilities: a mismatch) examined another facet of the aviation system where lines of responsibility and cross-sectoral misunderstandings occurred. Air Traffic Controllers (ATCOs) and pilots must work together to promote safe and efficient flight. However, significant differences between ATCOs and helicopter pilots were identified in the perception of controller duties with regard to the responsibility for aircraft separation in class D airspace and along the helicopter routes in the London Control Zone. There were variations in pilot understanding of controller responsibility compared to the formal regulations defining controller responsibility. ATCOs have no responsibility for aircraft separation within Class D airspace. Similarly, helicopter pilots had high expectations of ATCO responsibility for terrain and obstacle clearance; regulation define controllers have no responsibility for these tasks either. The variation between pilot perceptions of controller responsibility was interpreted using the concept of a shared mental model, whereby team members have a common expectation of other personnel in the wider system allowing them to coordinate their actions. The paper by Martin and Nixon, and the study by Bennett, again emphasise that safety is a system problem. While great steps can be made within the various individual components of the aviation industry (as attested by the exceptionally low accident rate) perhaps one of the greatest challenges now lies at addressing the interactions between organisations.

Hypoxia is a major threat to all aspects of the cognitive function of pilots during high altitude flight. Cabin depressurisation can result in impaired judgement; problems in psychomotor control and memory impairment, and at the cruising altitudes typical of modern commercial aircraft, can also rapidly result in a loss of consciousness. Bustamante-Sánchez, Delgado-Terán and Clemente-Suárez (Psychophysiological response of different aircrew in normobaric hypoxia training) describe the different effects on the symptoms of hypoxia experienced by different groups of aircrew. Differences in cognitive tests performance while experiencing hypoxia were observed between helicopter pilots, transport aircraft aircrew and fighter pilots. Their results showed that both tests of Working Memory and Matching to Sample tests were impaired by hypoxia. Physiological responses to hypoxia indicated that transport pilots were more negatively affected in terms of their higher heart rate and decreased the function of breathing muscles compared to helicopter pilots suggesting that this may be attributable to differences in their previous training, giving some indication by which a degree of tolerance may be induced in pilots.

Design of input/output devices

Touchscreens offer much greater flexibility than traditional hard mechanical controls and displays, as they can both display information as well as be the interface for control input. This dual use makes it possible to have layered displays and controls which has the potential to reduce the space requirements on the flight deck. Nevertheless, the flight deck is part of a moving platform and can be subject to turbulence in flight. Typically, turbulence is short-term, caused by natural phenomenon or other aircraft, resulting in frequent changes to the velocity of air through which the aircraft is travelling. Touchscreens can be more difficult to operate in turbulent conditions than traditional hard mechanical controls. Coutts, Plant, Smith, Bolton, Parnell, Arnold and Stanton (Future technology on the flight deck: assessing the use of touchscreens in vibration environments) sought to assess the effectiveness of touchscreens being operated in different levels of simulated turbulence (no- turbulence, light chop, light turbulence and moderate turbulence (experimental ethics prevented testing in heavy simulated turbulence due to potential harm that might come to participants)). Participants undertook four different tasks (multi-directional tapping, tracking, sliding and swiping) on touch screens in three different locations (side, centre and overhead). Generally, performance on the touchscreens was found to be equivalent to that reported on other input devices. The centre mounted screen had superior performance to all other positions, but the side screen was reported to be the most comfortable to use.

ATM systems contain decision support tools called safety nets to enhance ATCO’s monitoring tasks. The COOPANS ATM system which is deployed in five European countries, Austria, Croatia, Denmark, Ireland and Sweden, contains three types of safety alert, all signalled with the same acoustic BEEP-BEEP sound which can cause ATCO confusion. Effective decision support systems require human-centred design with effective stimuli to direct ATCO’s attention to critical events. It is necessary to provide ATCOs with specific alerting information to reflect the nature of the critical situation in order to minimise the side-effects of startle and inattentional deafness. Inappropriate alert design can induce startle, reduce or disrupt situational awareness, all of which delay responses to critical alerts in time critical situations. Human centred design can improve ATCO performance, situational awareness and reduce cognitive workload leading to increased cognitive capacity to manage complex tasks. Kearney, Li, Yu and Braithwaite (The impact of alerting designs on air traffic controller’s eye movement patterns and situation awareness) demonstrated that ATCOs’ visual scan patterns and situation awareness had significant differences between acoustic and semantic alerts. ATCOs exhibited different visual behaviours on fixations number, fixation duration and saccade velocity. The context-specific semantic alerts can improve ATCO’s situational awareness and speed up their response time to aircraft conflicts, thereby contributing to the goal of the European ATM Masterplan.

Eye scan pattern is one of the methods for assessing a pilot’s cognitive process in respect of situation awareness and decision-making in the cockpit. The visual information captured by eye trackers provides the opportunity to investigate the relationship between eye movement fluctuations and attentional shifts while performing tasks. Pilots are made familiar with the principles of scanning flight instruments and assessing relevant information in the early stages of initial training. Upset prevention and recovery training (UPRT) is intended to improve the ability of pilots to recognise and avoid situations that can lead to aeroplane upsets and to enhance their ability to recover control of an aeroplane that has exceeded the normal flight envelope. EASA strongly recommends implementing UPRT on a voluntary basis for general aviation pilots to reduce LOC-I occurrences, though ICAO only requires UPRT for commercial pilots. In two studies, Ryffel, Muehlethaler, Huber and Elfering (Eye tracking as a debriefing tool in upset prevention and recovery training (UPRT) for general aviation pilots) investigate the effectiveness of using eye trackers as a debriefing tool in UPRT scenarios. Eye tracking technology appears to be a powerful method to visualise instrument scanning techniques, supporting instructors with objective debriefing material and to fostering self-awareness in cognitive information processing. However, eye tracker hardware and software need to be improved in order to be more robust and efficient and to be an optimal feedback tool for aviation training.

Passenger comfort

Increased financial pressure on airlines has spurred changes in their passenger seating strategies. Airlines are reducing seat width to maximise profits at the same time that passengers are getting larger. The confined interior of the aircraft cabin is designed to maximise passenger capacity. Flights are increasingly crowded with load factors at a record high, as airlines are constantly seeking new approaches to increase the number of passengers in the aircraft while maintaining some minimum level of comfort. This study presents a method of analysing how passengers fit in their seats on aircraft. Miller, Lapp and Parkinson (The effects of seat width, load factor, and passenger demographics on airline passenger accommodation) investigated how passengers fit in their seats on the narrow-body aeroplane. Unlike most previous research which focussed on the spatial requirements of the pilots in the cockpit, this study examines the experience of passengers. The passenger modelling method is a flexible tool for discovering the impact of different factors on accommodation. The number of passengers, ratio of men to women, seat width, seat format, accommodation logic, and seating logic can all be manipulated to represent different scenarios for improving seat design. In general, passengers may have different concepts of accommodation and dis-accommodation. Some may define dis-accommodation as any intrusion into a passenger’s seat space, while others might defer that classification until physical contact occurs. It provides a flexible environment for discovering the impact of different factors on accommodation, therefore it can facilitate the design of more comfortable seating on the aeroplane.

Conclusions

Over the past decade, since the last special issue on aviation (Harris and Stanton 2010), there have been many developments in E/HF research and practice. It is fair to say that the theory and methods have matured over those ten years and it is pleasing to see that E/HF has been applied to all manner of problems. So far, these applications of E/HF have led to interesting findings and encouraging results. The take-home messages from the papers presented in this special issue are as follows:

• Distributed crewing of aircraft for commercial aviation will require a reconceptualization of the roles and tasks of the human and technical systems in variety of functions, including dispatch, maintenance, loading, ATC and ATM as well as the flight deck;

• Distributed crewing for commercial aviation could improve safety in emergency scenarios, such as a rapid decompression hazard;

• FLOPES can provide practical support to enhance training and safety management systems in aviation;

• Artificial Neural Networks provide a potential method for the prediction of unsafe acts from their psychological precursors in HFACS;

• Learning from successful safety outcomes or what went right is an important emerging component to enhance safety systems;

• Non-technical skills should be taught together with the technical skills of flying the aircraft as they are highly interdependent;

• CRM processes on the flight deck were robust even under moderately stressful situations;

• In a sample of airline pilots surveyed, 40% experienced high levels of burnout, which impacted on their performance and general well-being;

• Subjective measures of pilot workload are more sensitive than physiological approaches;

• Head-up displays of highways in the sky and virtual landing pad can be used to extend flight operations of helicopters to conditions of degraded visibility;

• Differences exist in their understanding of responsibilities between Air Traffic Controllers and helicopter pilots with regard to some aspects relating separation in certain classes of airspace;

• Differences in cognitive test performance while experiencing hypoxia was observed between types of pilot suggesting that this may be attributable to differences in their previous training;

• Touchscreens are a viable alternative to traditional hard mechanical controls but positioning on the flight deck affects performance and comfort;

• Provision of semantic audio alerts improves air traffic controllers’ response time and situational awareness in the COOPANS ATM system;

• Eye tracking technology appears to be a powerful method to support instructors to provide objective debriefing material for upset prevention and recovery training; and

• Predictive models of comfort, safety and biomechanics provide a useful tool for understanding possible passenger configurations in aircraft.

In summary then, it is concluded that E/HF has much to offer aviation research and practise. It is hoped that the papers and further the research themes will inspire readers of this special issue. They are encouraged to report on any developments, showing the substantial contribution that E/HF can make to aviation, within this journal.

Additional information

Funding

Professors Neville Stanton and Don Harris’s time on this special issue was co-funded by Innovate UK, the UK’s Innovation Agency, with support from the UK Aerospace Technology Institute (grant reference number 113108).