Editorial New paradigms in ergonomics

New paradigms

A paradigm is a world view underlying the theories and methodology of a particular scientific discipline; in our case, ergonomics. The ability to engage in new thinking and new paradigms is a critical aspect for the lifeblood of any discipline. Thankfully, ergonomics has a long history of innovations in theory, methodology, science and application. Once regarded as an offshoot of experimental psychology, physiology, biomechanics, engineering and computer science, ergonomics now draws from a much broader range of insights to become a discipline in its own right. This is reflected in the development of models and methods that are unique to the discipline. Accident causation models, for example, first emerged in the early 1900s (Heinrich 1931) but have since evolved to consider entire systems and emergent properties therein (e.g. Leveson 2004; Perrow [1984] 1999; Rasmussen 1997; Reason 1990). Methodologies have moved from focusing on human tasks (Taylor 1911) and work (Gilbreth 1912) to entire systems and the constraints shaping behaviour in the world (e.g. Vicente 1999). Human performance itself has shifted from a world view in which it was mathematically characterised in terms of, for example, response times, sensitivity or decision bias (e.g. Green and Swets 1966) to more complex forms of distributed and situated cognition (e.g. Hutchins 1995).

The systems in which ergonomists work continue to evolve, creating emergent problems and often strengthening intractable issues. In order to remain relevant and impactful, the discipline needs its paradigms to evolve too. The aim of this special issue is to provide researchers and practitioners with an opportunity to present and discuss contemporary, forecasted and required paradigm shifts in ergonomics.

Contributions to this New Paradigms in Ergonomics special issue have been grouped into three themes: new paradigms in theories and methods; new paradigms in practice; and new paradigms in domains and values. These themes take the reader on a journey through underlying theories, news ways to apply those theories, emerging domains in which ergonomics is expected to play a greater role and to what ultimate ends. In the new paradigms, in theories and methods theme, are papers on ‘Quantum ergonomics’ (Walker et al. this issue), ‘Nonlinear dynamical systems for theory and research in ergonomics’ (Guastello), ‘Fitting methods to paradigms’ (Salmon et al. this issue), ‘Quantitative modelling in cognitive ergonomics’ (Moray, Groeger, and Stanton, this issue) and ‘Beyond human error taxonomies in assessment of risk in sociotechnical systems’ (Stanton and Harvey, this issue). In the new paradigms, in practice theme, are papers on ‘Detection of error-related negativity in complex visual stimuli’ (Sawyer et al. this issue), ‘Towards continuous and real-time attention monitoring at work’ (Mijovic et al. this issue), ‘Musculoskeletal disorders as a fatigue failure process’ (Gallagher and Schall, this issue). In the new paradigms, in domains and values theme, are papers on ‘The field becomes the laboratory?’ (Sharples and Houghton), ‘Imposing limits on autonomous systems’ (Hancock, this issue) and ‘Nature’ (Richardson et al. this issue). Summaries of each of the contributions together with conclusions for the special issue are in the following sections of this editorial.

New paradigms in ergonomics theory and methods

Walker, Salmon, Bedinger and Stanton (Quantum ergonomics: Shifting the paradigm of the systems agenda) draw inspiration from the world of quantum physics to confront some difficult truths about ergonomics paradigms. Walker et al. present three case studies to demonstrate ‘quantum effects’ in ergonomics science. In the first case study, they show how the nanoscale, individual behaviour of drivers cannot reveal the macro-scale behaviour of the entire road traffic network. The case study reveals the unexpected reverse finding; that more route guidance (in effect, better situational awareness) on the part of individual drivers does not necessarily result in lower CO2 emissions for the whole road network. Sometimes it is worse. ‘Catastrophically counter-intuitive’ findings like these show the difficulty in predicting system-wide emergent properties (often the thing we really want to know) simply from studying the behaviour of individuals and magnifying the resultant findings. In the second case study of a submarine control room, Walker et al. present insights into the behaviour of a command team system that have been hitherto unforthcoming from studying individual operators. The system dynamics have enabled the development of a submarine command room simulator that is currently being used to generate new understanding of how these teams work together in challenging environments (Stanton 2014). Walker et al. argue that we need systems approaches if we are to remain an effective discipline in the new increasingly complex and networked world. They argue that we need to look at new approaches, such as phase spaces (see Stanton, Walker, and Sorensen 2012; Guastello, this issue), as a means of offering some rigour in naturalistic studies (see Sharples and Houghton this issue). In the final case study, Walker et al. present findings from a large-scale operation field trial (see Stanton et al. 2009). Using network analysis methods, they were able to show how the organisational dynamics changed over time. Walker et al. comment that ergonomic methods need to be fit-for-purpose (see Salmon et al. this issue) and offer the maximum of insight for the minimum of effort. This is in contrast to a ‘one size fits all’ approach to method (and theory) selection. The discipline of ergonomics is good at responding to problems in the world – which are in a constant state of flux – so the methods need to adapt and develop to meet this evolving demand. The overall message is one of hope: methods currently exist to enable these challenging systems problems to be tackled in practical real-world circumstances, provided the necessary paradigm shift can be made to occur.

Guastello (Nonlinear dynamical systems for theory and research in ergonomics) also argues that the complexity of sociotechnical systems requires a paradigm shift in ergonomics theory (see Walker et al. 2010 this issue) and practice (see Salmon et al. this issue). He proposes the non-linear dynamical systems approach as an alternative paradigm that can cope with this complexity, as it addresses:

•	structures and patterns of variability;
•	underlying dynamics and system changes;
•	both internal and external dynamics;
•	effects of state-dependent changes; and
•	both top-down and bottom-up emergent properties in systems.

Guastello explains how systems do not just have one stereotypical response, rather there are a multiplicity of behaviours with different patterns and outcomes. The effects can be both large and small, depending upon the state of the system. Guastello brings many different systems concepts and methods together under the non-linear dynamical systems framework with the view that these will help ergonomics explain complex systems behaviour. One example is the use of phase spaces to visualise the dynamical processes in systems (Walker et al. this issue; Stanton, Walker, and Sorensen 2012). He provides numerous examples of the practical application of non-linear dynamical systems methods, including (but not limited to): accidents, biomechanics, performance variability, resilience and team coordination and workload. The non-linear dynamical systems approach has much to offer systems ergonomics, both as a theoretical construct and practical methods.

Salmon, Walker, Read and Stanton (Fitting methods to paradigms: are ergonomics methods fit for systems thinking?) question whether current ergonomics methods really are fit for the new systems paradigm which, from the above, is clearly in the ascendancy. In many ways, ergonomics methods have always been about system interactions, from individuals to teams to organisations (Stanton et al. 2013). Yet Salmon et al. argue that the increasing complexity of the modern world (Walker et al. 2010), and challenges that ergonomics faces, may have left the methods wanting. They show that despite the dramatic reductions in accidents over the decades the statistics are plateauing. It is suggested that we may have reached the limits of deterministic methods, so new approaches are required. Salmon et al. explore how well our current methodological toolkit can cope with modern day problems by focusing on five key areas within the ergonomics paradigm of systems thinking: normal performance as a cause of accidents, accident prediction, system migration, systems concepts and ergonomics in design. The ergonomics methods available for pursuing each line of inquiry are explored, along with their ability to respond to key requirements. Salmon et al. come to the conclusion that, although our current suite of ergonomics methods is highly useful, there is work to be done. For example, they conclude that, although providing rich outputs, some of our existing accident analysis methods do not describe accident causation in a manner that is congruent with contemporary models (e.g. Rasmussen 1997). With regard to accident prediction, it is concluded that we currently do not have appropriate methodologies for predicting systemic accidents (although see Stanton and Harvey, this issue, for new developments). Likewise, Salmon et al. argue that assessing the migration of performance towards and away from safety boundaries, a key systems thinking concept, is outside of the capabilities of our current methodological toolkit. They also suggest that further ergonomics problems and constructs may be suited to systems level analysis, and that few ergonomics methods are actually being used directly in system design processes. Salmon et al. close the article in an upbeat manner, highlighting that many seemingly appropriate methods already exist, both in ergonomics and other disciplines, and that research is underway to develop some of the methods required (see Salmon 2016a,b). In closing, they map out the prerequisites for methods development in systems ergonomics. If our discipline is to maintain currency and rise to the contemporary and future design challenges, we need to develop methods that have the entire sociotechnical system as the unit of analysis.

Moray, Groeger and Stanton (Quantitative modelling in cognitive ergonomics: predicting signals past at danger) state that the discipline of ergonomics is sufficiently mature to enable quantitative modelling of performance. A case study of the activities of the Thames Trains driver in the Ladbroke Grove rail accident is presented. All three authors were expert witnesses in this case, so they combined their knowledge for this paper. Moray et al. present accounts of the accident background, context, infrastructure and timeline together with ‘black box’ data from the train. The accident raises questions about why the driver behaved in a particular way on that fateful day, as well as the performance of signals and warnings inside and outside the train cab. Eye movement data were used as the basis for the development of a cognitive model of driver visual attention. The model accounts for attention both inside and outside the train cab. It revealed the difficulty in acquiring the signal in a relatively short amount of time, due to the number of signals, their placement and the speed of the train as well as initial masking of the signals. The ambiguity of the automatic warning system horn further compounded the problem coupled with a strong expectation of the signal being in a non-red aspect and glare from the sun. The quantitative model shows how prediction of visual attention can be used to determine risks associated with signal sighting, which can in turn be used to support guidelines and in-cab display design.

Stanton and Harvey (Beyond human error taxonomies in assessment of risk in sociotechnical systems: a new paradigm with the EAST ‘broken-links’ approach) introduce a new paradigm for risk assessment based on the Event Analysis of Systemic Teamwork (EAST) method. They argue that the approach offers a fundamentally different way of thinking about risk in systems, and is dramatically different from existing human error taxonomies. Rather than treat accidents as the result of ‘human error’ the EAST broken-links approach treats them as information communication failures. Stanton and Harvey provide examples of information communication failures in the Herald of Free Enterprise capsize at Zeebrugge, the British Midland crash at Kegworth and the Thames Train collision at Ladbroke Grove. All of these accidents have the common feature that key information was not communicated to an appropriate agent by the system in an effective manner at the right time. Stanton and Harvey present a case study of an EAST analysis of a RAF Hawk and RN Frigate system (see Salmon et al. this issue, for a call on understanding normative systems). The EAST model comprised normative task, information and social networks together with the composite model. Stanton and Harvey demonstrate that by systematically breaking the task and social links in the networks, risks are revealed by information not being communicated between tasks and social agents. One hundred and thirty-seven risks were revealed by breaking only 12 task and 19 social links. Stanton and Harvey show how the emergent information communication failures transpose risk around the network of agents, actors and artefacts. Furthermore, they show how reducing risk in one part of the system increases risk in another part. For future work, Stanton and Harvey plan to break and/or add multiple links simultaneously to show how compounded information communication failures and short circuits could affect system performance.

New ergonomics paradigms in practice

Sawyer, Karwowski, Xanthopoulos and Hancock (Detection of error-related negativity in complex visual stimuli: a new neuroergonomic arrow in the practitioner’s quiver) have established a link between the error-related negative evoked response potential (ERN-ERP) and the complex task of identifying a motorcycle in a busy visual scene. The innovation in this work is to take visual search tasks beyond simple letter arrays and icon images in binary forced-choice tasks to a more naturalistic task. Motorcycle detection is notoriously difficult and undoubtedly the cause of accidents. Sawyer et al. transferred this task into the laboratory alongside a more traditional letter flanker task (where the target letter was flanked on either side by non-target letters). They demonstrated that both tasks produce similar error rates and workload ratings. Sawyer et al. found that the ERN-ERP waveforms were very similar for the letter flanking and motorcycle identification tasks. In both tasks, the wave was flat for the correct response and had a pronounced negative deflection for the incorrect response. Sawyer et al. propose that this finding shows the ERN-ERP has potential for practical application and they plan to use the procedure for dynamic environments in the future. Potential applications include use in training and naturalistic human–computer interfaces.

Mijovic, Kovic, De Vos, Macuzic, Todorovic, Jeremic and Gligorjevic (Towards continuous and real-time attention monitoring at work: reaction time versus brain response) state that wearable electroencephalogram (EEG) has made objective, online, continuous, monitoring of attentional state a practical possibility. They have applied the use of EEG to examine the P300 (an ERP wave component associated with the process of decision-making) whilst people performed simulated simple and repetitive assembly tasks. The purpose of the research was to test the relationship between attention level (as measured by the P300 amplitude) and the time it takes to complete a task cycle on the assembly line. The simulated assembly task was designed to replicate monotonous work, comprising: picking up two items, connecting them together, placing them in a machine, pressing a pedal to crimp the parts together, removing them from the machine and placing them in a bin. The entire task cycle took less than 10 s and a shift would involve over 2500 cycles. The EEG data were collected simultaneously. The study revealed that the P300 was correlated with attention on the task. High P300 values were found with shorter cycle times and vice versa. Mijovic et al. are keen to point out that there are inter-individual differences which led to some inconsistencies in the data, but the findings hold true for the collective group level. This study has also shown that continuous monitoring of attention is possible and that neuroergonomic approaches have the potential to be used in the workplace (see also Sawyer et al. this issue).

Gallagher and Schall (Musculoskeletal disorders as a fatigue failure process: evidence, implications and research needs) state that musculoskeletal disorders are a major cause of workplace injury and, although much is known about the causes, little theoretical work has been presented. By way of contrast, fatigue failure in materials is well established and well understood. Fatigue failure in materials is a function of differential loading characteristics and the frequency of cycles. This is analogous to the cumulative trauma disorder in human tissue, which has led Gallagher and Schall to the fatigue failure hypothesis. Some evidence from recent studies has lent support to this hypothesis, that musculoskeletal tissue also shows signs of fatigue failure. In particular, a large epidemiological study has shown that forceful repetition leads to carpal tunnel syndrome. Gallagher and Schall offer a unifying framework for musculoskeletal disorder risk factors together with validated methods for assessing the risk of cumulative damage. They argue that the fatigue failure process offers the possibility to assess risk and develop effective interventions. Ultimately, musculoskeletal disorders could be predicted as our understanding of the fatigue failure process develops. This requires a much larger data-set of fatigue failure on musculoskeletal tissues (much like the data held on material fatigue). It also requires a better understanding of the dynamical properties of the musculoskeletal system (see Guastello, this issue). Gallagher and Schall are working towards a fatigue failure theory of musculoskeletal disorders and a consequent reduction in workplace injury.

New ergonomics paradigms in domains and values

Sharples and Houghton (The field becomes the laboratory? The impact of the contextual digital footprint on the discipline of E/HF) argue that the ubiquity of data collection in the digital world has had a dramatic effect on the volume of information available for ergonomics research and development. They go so far as to suggest it could even change the nature of how the discipline goes about its practice in terms of new methods and insights (see Walker et al. this issue; Guastello, this issue; Salmon et al. this issue). The digital exhaust (called ‘contextual user data’ by Sharples and Houghton) that is emitted by people in their work, social and home lives could provide a richer understanding about the behaviour of individuals and collectives, but with that comes a range of ethical, moral and privacy issues. Sharples and Houghton present three scenarios, a hospital as a digital workplace, distributed crowd-sourced activity, and managing journeys. In each of these scenarios, Sharples and Houghton show the potential for the technology to be both beneficial (improving human well-being) and detrimental (making work harder, less enriching, disengaging and potentially harmful). They argue that, used appropriately, ergonomics methods could make use of the contextual user data in a manner to improve the design of work, social and home lives consistent with the values of the discipline. Big data analytics are likely to play an important role in the discipline, and its methods, going forward.

Hancock (Imposing limits on autonomous systems) warns of the impending march from automated rule following technologies that require human supervision to autonomous systems, those intelligent enough to not necessarily require human supervision). He states that it is necessary for the ergonomics community to engage with the debate, and to rigorously research and design these systems if we are to ward off the (many) potential pitfalls of the technology. There are some foreseeable problems, such as the so-called ‘mode error’, where human supervisors confuse the mode the automated (or autonomous) system is in (e.g., Stanton, Dunoyer, and Leatherland 2011). More worryingly are the unforeseeable problems, those that emerge through complex system interactions (Walker et al. this issue; Guastello, this issue). Hancock is quick to point out that each potential benefit afforded by technology is accompanied by drawbacks (see Sharples and Houghton, this issue). He goes much further to suggest that automation could have deeper and more profound societal effects. One of these is to deepen the societal divide between rich and poor. History has shown that technology can have dramatic effects on work, social and leisure lives. Hancock warns that technological Darwinism may not always be to the benefit of society at large. By way of mitigation, he argues for an ethical approach to automation, which the ergonomics community should be leading. Autonomy brings moral philosophy into sharp focus, with Hancock’s depiction of the ultimate delegation of authority to automation that could lead to very inhuman acts. He argues that, as ergonomics is about human betterment especially with technological development, we should not shirk our responsibilities., Thus the paradigm is not merely theoretical or methodological, but strikes to the heart of the discipline’s underlying value base.

The final paper of this special issue is provided by Richardson, Maspero, Golightly, Sheffield, Staples and Lumber (Nature: a new paradigm for well-being and ergonomics) who propose ‘nature’ as a new paradigm for ergonomics that is required to promote physical and mental well-being as a core principle of the discipline. They argue that the progressive urbanisation and removal of our experience with the natural world (a global trend) has a negative effect on health and well-being. From a review of the corpus of evidence, Richardson et al. are able to demonstrate the benefits that arise from the ‘nature experience’. This evidence shows that people who live in urban environments generally have poorer health than those in rural settings. Green spaces in urban environments have offset some of those negative effects for city dwellers, leading to improvements in mood, mental health, well-being, anxiety, vitality, job and life satisfaction. Similar benefits have been found for physical health as well, with reports on improvements in heart rate, blood pressure, muscle fatigue, sickness and stress. Even virtual scenes of nature seem to have short-term restorative effects, presumably because of the associations they evoke. Richardson et al. argue that the degree of connectedness people feel with nature has an effect on their psychological well-being in the workplace. They offer ‘nature’ as a cost-effective, simple, ergonomics intervention strategy with positive physical and mental health benefits. It has the potential to be a transformational paradigm shift that is consistent with the mission of the ergonomics discipline as a whole.

Conclusions

This collection of papers from leading researchers and practitioners in our discipline represents a tantalising glimpse into a fascinating ergonomics future – or possible futures. Given the shifting nature of the systems in which ergonomists work, there is clearly a strong need to revisit the ergonomics unit of analysis and a groundswell of support for a broader, stronger, and more rigorous approach to systems thinking. There is a corresponding need, and opportunity, to enhance our ergonomics practices, with several important advances now afforded by new technology and the discipline’s maturing state of knowledge in critical areas. Finally, there is an equally strong need to combine these paradigm shifts in theory, methods and practice with continued scrutiny of where ergonomics should be applied, how it should be applied, and to what ends. In other words, to keep asking ourselves ‘what do we stand for?’. With this in mind, the collection of papers provides a series of important take-home messages. These include:

•	The need to face up to the ‘quantum effect’ in ergonomics science and analyse systems at the appropriate level and with appropriate methods.
•	That non-linear dynamical systems approaches offer a way of dealing with some of the true complexities of sociotechnical systems.
•	That autonomous systems pose a significant threat to our way of life and we need to ensure they are implemented in an ethical manner.
•	That work is required to develop ergonomics methods that are capable of modelling key systems thinking concepts such as normal performance, future accidents and system migration.
•	That quantitative modelling on cognitive ergonomics is now able to predict performance and can be used to assess risk and design systems.
•	That new modelling approaches can model normal performance and assess risk in complex sociotechnical systems.
•	That the sheer quantity of contextual user data available from our digital work, social and home lives will provide a richer understanding of human, organisational and system behaviour.
•	That neuroergonomics methods can be used for practical visual search tasks and measures of attention in applied settings.
•	That the fatigue failure process could explain musculoskeletal disorders and reduce workplace injury.
•	That connection with nature can offer transformational physical and mental health benefits for urban dwellers and office workers, and is an emerging ergonomics criterion.

It is hoped that readers of this special issue will be inspired by these new paradigms in ergonomics and seek to push the boundaries even further. The lifeblood of the science depends on continual evolvement and developments to take on the challenges we face in complex sociotechnical systems design and evaluation. Perhaps the most significant take-home message is that all these papers have demonstrated how theory maps onto practice. As such, perhaps the only remaining paradigm shift is for these ideas and concepts to be taken up more widely and the discipline advanced, until the next paradigm shift occurs.

Neville A. Stanton
Transportation Research Group, Civil, Maritime, Environmental Engineering and Science Unit, Faculty of Engineering and the Environment, Bouldrewood Innovation Campus, Burgess Road, University of Southampton, Southampton, UK
[email protected]
Paul M. Salmon
Centre for Human Factors and Sociotechnical Systems, Faculty of Arts, Business and Law, University of the Sunshine Coast, Maroochydore, Queensland, Australia
Guy H. Walker
School of Energy, Geosciences, Infrastructure and Society, Heriot-Watt University, Edinburgh, UK