Why a ‘Competence-based’ Approach to Crew Resource Management Training?
The thesis presented in this paper is straightforward: one of the principal causes of incidents and accidents in civil jet transport operations is the lack of effective management of available resources by the flight-deck crew.
John K. Lauber (1979)
In this chapter, I want to address three questions: what do we mean by ‘crew resource management’ (CRM)?; what might the ultimate goal of any training intervention be?; how do we set about designing training to develop CRM skills? Despite being involved in the design and delivery of CRM since 1989, and was also CRM Instructor Examiner for the UK CAA for part of that time, I have never been convinced that it was a discipline sufficiently distinct from more fundamental approaches to aircraft pilot performance development. Undoubtedly, it is a fascinating field of research and clearly overlaps the broader ‘safety’ domain but from the narrow perspective of a safety-related training intervention, it remains a muddy concept. As this book unfolds, I hope to make my case more fully. For now, however, I want to set out some broad principles. I start with a brief overview of the CRM concept from its origins, through various iterations to the current position, which is best summarised as ‘confused’. Given that the domain, in my opinion, lacks coherence, it follows that we need an analytical framework to help clarify the requirement. In the second part of this chapter, I will propose that the goal of training is to develop expertise and, therefore, I will try to outline what ‘expertise’ looks like in pilots. As with CRM, generally, I suggest that the current position on training in aviation is equally confused. Therefore, the chapter concludes with an outline of some approaches to training design and a discussion of the challenges to the assessment of performance that flow from my analysis.
I set out in this book to explore CRM through the ‘competence’ paradigm, by which I mean the development of transferable skills that support expert performance. Unless training achieves some measurable output, it cannot be considered a valid use of the increasingly scarce resources of time and money. Furthermore, by taking the competence route, I hope to show that it becomes very difficult to sustain the technical/nontechnical duality of pilot performance. Therefore, as I develop my arguments in this chapter, the discussion will move away from dealing with CRM, specifically, to a consideration of competence in a more generic sense. Finally, the management of training is, itself, a ‘competence’. While the content of this chapter might be of passing interest to line pilots and students, it should be central to the role of anyone involved in the management of the training process.
A Short History of CRM
The broad concept of CRM training has been around for over 35 years and, in addition to now being applied in just about every area of aviation, has been adopted by other domains such as maritime and medicine. So, why the need now for a specific competency-based approach to the topic given that it is firmly established as a training requirement? There are two main creation myths attached to ‘CRM’. If you are European, CRM started with the Tenerife disaster in 1977 (Project Tenerife), and if you are North American, it was the United Airlines Flight 173 (NTSB, 1979) crash at Portland Oregon in 1978 that was the trigger. The consensus at that time was that aircraft were now so reliable that it was the human that had become the problem. We will see later that this was yet another myth. The actual catalyst for action was undoubtedly the joint NASA/industry workshop - ‘Resource Management on the Flight Deck’ (NASA, 1980) - in June 1979, at which John Lauber offered the view quoted at the start of this chapter. The workshop resulted in the first course, run by United Airlines, delivering what we would now recognise as CRM training. That course was, in turn, the template for almost all other airlines’ initial implementations. Interestingly, as Table 1.1 illustrates, the airline industry has historically tended to seek technological solutions to safety issues. CRM remains one of the few attempts to increase safety through fundamental behaviour change. The problem is that it was never entirely clear what ‘CRM’ actually was nor if the initiative really worked. Soon after the FAA published its initial proposals, it announced that it was not intending to progress to a full rule-making (Diehl, 2011). The major US airlines objected. As a training intervention, CRM was seen to be of real value.
Despite the concept’s longevity, the aviation industry has yet to offer a coherent definition of ‘CRM’. Lauber’s ‘making full use of all available resources’ persists but, unfortunately, offers little in the way of guidance to training developers. Attempts
Catalysts for Change
to elaborate on the concept have generated lists of topics such as communications, teamwork, decision-making and safety culture, which form the basis of curricula around the world. From the outset, ‘CRM’ appears to have been a generic umbrella label and, over the years, it has been interpreted in a number of different ways. I do not propose to offer my own definition. Instead, I hope to make it clear that ‘CRM’ is a perspective on, not a subset of, the skills needed to fly aircraft safely. Helmreich et al. (1999a) make the point that ‘some have argued that CRM training should ultimately disappear, as it becomes fully integrated into technical training’. While the authors admit to once supporting that position, they were then of the opinion that CRM should remain separate on the grounds that it forms ‘a critical and continuing component of a safety culture’. My position is that they were broadly correct in their original analysis. That said, once training has been designed to address real competence, there remains the question of what underpinning knowledge in the broad area of CRM is needed to support performance.
The honeymoon did not last long. Within lOyears, the major US carriers were questioning the cost of training with no discernible reduction in adverse events. Professor Bob Helmreich, at the University of Texas, Austin, had participated in the 1979 conference and his research group was asked to come up witli a new approach. Helmreich’s team did two things. First, they formulated the ‘generational’ model of CRM development (Helmreich et al., 1999b). This was another creation myth in that the ‘generations’ they identified only make sense from a North American perspective. Indeed, an early CRM conference asked if the concept even applied outside of the USA. However, the ‘generations’ model culminated in ‘error management’ as the ultimate goal of CRM, which neatly led into the teams’ second initiative. The Texas Research Group developed the Line Operations Safety Audit (LOSA) process. Between the first, prototype, LOSA run by Delta Airlines and the second iteration, adopted by Continental Airlines, the ‘threat and error’ (ТЕМ) conceptual framework had been developed (Klinect, 2005). In the mid-1980s, ТЕМ was adopted by ICAO, a move helped by the fact that Captain Dan Maurino, the ICAO Safety Manager, had been a thesis supervisor for the LOSA/TEM project.
The relationship between ТЕМ and CRM was confused from the outset. Helmreich, himself, recognised that ТЕМ was a data collection process and that CRM was a training intervention aimed at resolving issues identified in the audit. He argued that data were only of use when fitted into a theoretical framework: ТЕМ was that framework. The shift in the centre of gravity from CRM to ТЕМ can be traced, in part to a paper entitled ‘Defensive Flying for Pilots: An Introduction to Threat and Error Management’ (Merritt & Klinect, 2006). In it, the authors describe ТЕМ as promoting a ‘proactive philosophy’ and providing ‘techniques for maximising safety margins’. In their view, ТЕМ had moved from being a static framework of analysis to a dynamic model of pilot behaviour. They state that ‘ТЕМ tools work best when pilots adopt ТЕМ techniques’ and ‘many of the best practices advocated by CRM can be considered ТЕМ countermeasures’. What was not clear, however, was what constituted а ТЕМ ‘tool’ or ‘technique’ nor how a ‘practice’ differs from a ‘countermeasure’ in any substantive way. The European Aviation Safety Agency (EASA) pilot training requirements now require ТЕМ to be included in all phases of training such that pilots will demonstrate the ‘attitudes and behaviours appropriate to safe conduct of flight, including recognising and managing potential threats and errors’. Apart from cosmetic changes in language, the core of the EASA training requirement remains the same. The latest iteration of the LOSA process has substituted the competence-based training (CBT) initiative’s ‘competencies’ for the original FAA countermeasures, suggesting that ‘ТЕМ’ is not a fundamental truth and can be changed to suit circumstances.
The confusion between CRM and ТЕМ persists. EASA, in its updated CRM regulations (EASA, 2015) rather trips over itself in trying to reconcile the two initiatives. Of interest, while CRM training persists, no airline (to my knowledge, at least) has managed to develop a ‘pure’ ТЕМ training course that does anything more than simply define the key concepts of a threat, an error and an outcome. Later in this book, I do draw' on data collected in LOSA projects I have managed. The LOSA process is a valuable source of information about normal operations, as was claimed by Bob Helmreich. The benefits that accrue from substituting one acronym for another, however, remain unproven. Despite continuing concerns about the cost-effectiveness of training and the progressive reframing of the issue in a new jargon, the fundamental approach to CRM training has never really changed. The broad themes of an early course are still recognisable today.
Thirty-five years have, however, seen changes in other areas. Our understanding of ‘error’ as a symptom, not a cause, of failure has changed. New paradigms, such as ‘Safety-II’ and ‘resilience engineering’, recognise that human behaviour is adaptive and that we should look more at how pilots respond to change, rather than simply try to prevent the repetition of past failures. The enormous cost of initial and recurrent training, especially in a challenging financial climate, is causing airlines to employ new technologies to deliver training. Finally, advances in aircraft automation have changed the nature of the job of an airline pilot. Today, airline pilots are system managers, more akin to the operator in a power production plant control room. The prospect of autonomous, remotely piloted commercial aircraft is on the horizon and, in all probability, this book is aimed at the last generation of pilots who wall travel in the same craft as their passengers. Nonetheless, that generation will still have to contend with the same challenges as their predecessors while, in many cases, carrying larger passenger loads over greater distances. It is probably timely to take stock of the skills they w ill need.
The fundamental question we need to ask is what are we trying to achieve by exposing pilots to the cluster of ideas that fall under the umbrella of CRM? Furthermore, what do we want pilots to do differently once they have had that exposure? Ultimately, we could say that we w'ant to increase ‘safety’. In an industry that is already incredibly safe, as measured in terms of hull loss events and passenger fatalities, this seems a daunting task. Of course, the actual occurrence of and the potential for a future occurrence are not the same things and so risk mitigation might be the goal. In either case, behaviour change must be a candidate target for any intervention. CRM courses have tended to be ‘educational’ in nature in that they have been rather broad brush and informative rather than requiring trainees to show some incremental improvement in performance. It is not unreasonable to expect to see an output from training that could be measured in behavioural terms but this has never really been the goal. Although performance is assessed in a simulator, there is usually no direct connection between the classroom inputs and the observed outputs. At this point, it is probably fair to say that the historical isolation of CRM as a domain is part of the problem. By requiring ground training to satisfy an externally imposed curriculum, the link between operational need and training design is severed. To restore this link, we need to look more holistically at the problem. To this end, I now want to examine the concept of expertise.