According to the US General Accounting Office (GAO), some licensed professional pilots should never have been given their jobs, and no amount of training will rectify this situation. While many in the industry would agree with that, there are real differences of opinion on just how those pilots' deficiencies should be spotted before they cause accidents - or even whether they can be.
The GAO's observation about the existence of incompetent pilots in the industry appeared in a report published in June 1996. The purpose of that report was not specifically to reveal shortcomings in the nation's systems for quality-control of aviators, but to examine the status of efforts to integrate human-factors (HF) research into US Federal Aviation Administration activities. There is, however, an implied criticism of system weakness.
Since then, the official inquiry into the Birgenair Boeing 757 crash on 6 February, 1996, just offshore of the Dominican Republic has reported its serious concern about the pilots' poor standards of knowledge of the aircraft systems. While acknowledging that the airline's provision of flightcrew training had not broken any laws, the report said that internationally accepted minimum standards for pilot training had fallen behind developments in the air-transport industry (Flight International, 13-19 November, 1996, P14).
Accident horrors
Under the heading "Selected aviation incidents illustrating human factors issues", the GAO lists brief examples of real events, although the accounts are de-identified. One incident is described with chilling simplicity as follows: "An aeroplane crashed because the pilot did not take proper corrective action even though the first officer appropriately advised him-The pilot did not achieve a satisfactory level of performance, despite remedial training." The GAO summarises the human factors issues involved in only one word: "Selection". Finally, describing the fundamental facts behind this accident, the GAO suggests: "The pilot may not have possessed the skills needed to become competent, despite training."
It says a great deal about US aviation's systemic weaknesses that the GAO's de-identified accident could have described at least two recent US regional-airline fatal crashes, if the GAO text is compared with National Transportation Safety Board (NTSB) accident reports. In both accidents, the aircraft stalled and crashed on night approach, and both the captains had histories of needing regular remedial training to pass proficiency checks. In one case (the one which most precisely fits the GAO description), the captain's previous employment contract had been terminated because of inadequate performance, but he had rehired immediately been by another airline. That is, presumably, why "selection" is the key word in the GAO assessment. In both accident reports, poor crew organisation, failure to adhere to procedures, and failure to react correctly when the aircraft stalled were cited by the NTSB as causal factors. In both, the FAA was criticised for inadequate oversight of training requirements and procedures at the airlines.
The task of assessing potential pilot employees for suitability, and of carrying out their recurrent training and competency checks, is largely devolved to the airlines by national aviation authorities everywhere. The only part of the piloting-standards system for which the authorities universally accept direct responsibility is the legal one of issuing pilot-licences for the first time - and even in that case the examining/checking task is frequently contracted out.
The type of event described by the GAO would theoretically be impossible if flightcrew licensing (FCL) systems, airline selection procedures, and periodic assessment during recurrent line-training worked successfully. The implication is that, while the filters work most of the time, there are system failures which not only let poor-quality pilots gain access to flying jobs in commercial flying, but also to stay there.
If this can happen in the USA, with its structured oversight system and accident statistics which always put it among the best in the global safety league, there is clear cause for concern. It is for this sort of reason that the International Civil Aviation Organisation, at its conference in Montreal on 10-12 November of civil-aviation authority directors general, will continue its push for a mandate to take up an executive safety-auditing role. The meeting's theme is: "A global strategy for safety oversight".
FCL and line training are intended to manage human-performance standards. The performance of the pilot mentioned in the GAO report, however, was not good enough, and it is suggested that no amount of training could have brought it up to the necessary standard for commercial flying. Despite such system failures, most pilots who reach line flying are good enough. Good line-training sets out to keep them that way.
The International Federation of Airline Pilots (IFALPA) human-performance (HUPER) committee recently met in Tobago to discuss these issues. Its remit was to examine all the aspects of how an airline pilot's performance is produced through training, recognised with the issue of a licence, assessed by a potential employer at selection, and finally maintained and assessed during recurrent training by the employer.
Recurrent training, however, has a considerable performance-checking component: airlines have to have standards to determine if their pilots are retaining at least the competency for which they, presumably, were selected upon hiring. Judgement on standards, however, is subjective unless it is qualified and quantified, which is easy for purely technical skills and knowledge, but more difficult in the areas which crew-resource management (CRM) purports to cover. Subjectivity allows for abuse - or the fear of it - which is not just bad for quality, but for flightcrew morale, which affects performance.
Hiring criteria
Standards and safety alone, theoretically, govern the decision about whether to hire a pilot, or to keep a line-pilot flying when, for example, his performance remains marginal after remedial training. The decision should be independent of considerations such as aircrew shortages during expansion, or flightcrew surpluses when business falls. Life is not like that, however, partly because few airlines maintain strategic ab initio training programmes, and partly because flying for a living is still a popular enough occupation to ensure that there are always enough people with commercial-pilot licences happy to get an airline job, even if they cannot offer quality.
IFALPA has always said that pilot organisations would not wish to put any limitations on the testing an airline sees fit to apply for initial selection purposes. With the rise in the profile of CRM as a part of initial and recurrent training, however, IFALPA is concerned that CRM could be used as a means of assessment for individual line pilots. IFALPA argues against this, explaining: "When you are talking about selection you are talking about an individual. Assessment during line training is assessment of a crew, not of individuals in isolation." As a part of a total crew-training programme, CRM has a well-defined aim, but it was not designed to be an assessment tool for individuals, insists IFALPA.
IFALPA uses the generic term "non-technical skills" for the human-performance category outside the areas of stick-and-rudder skills and technical knowledge. Some non-technical skills come under the heading of CRM skills.
One of the HUPER committee members was tasked with drawing up a discussion paper on the assessment of non-technical skills (NTS) for presentation at the Tobago seminar. IFALPA emphasises, however, that it contains proposals which may yet be rejected, because the meeting did not come up with a definitive policy on NTS assessment; the Association hopes to have produced a draft policy by the end of the year.
The discussion paper clearly shows the complexity of the International Air Transport Association's concern to keep pilot standards high in all skill areas, yet prevent the exploitation of pseudo-psychological assessments under the umbrella of CRM. The document begins: "Where human factors are mentioned as causes for accidents, often the non-technical skills of the pilots are part of it. When the non-technical side of our job appears to be so important, it is not enough to receive training only; it has to be ascertained that minimum requirements regarding non-technical skills are met by pilots."
Before CRM, IFALPA points out, examiners or check pilots used to assess flightcrew for "airmanship and captaincy". In reality, however, these are imprecisely defined terms, just as CRM is, suggests the document. Captaincy or good CRM is easy to recognise when it is there, just as its absence is obvious, but defining when non-technical skills have reached an adequate standard for safe line-operation is difficult.
IFALPA's NTS discussion paper suggests: "When it comes to non-technical issues which they want to address, examiners and instructors cannot properly mark their opinions on examination check forms. Terms used in non-technical issues vary widely and are often vague."
Jarring views
The European Joint Aviation Authorities, in its Joint Aviation Requirements for flightcrew licensing (JAR FCL), has framed rules for checking non-technical skills during certain examinations: for example, those for an air-transport pilot's licence (ATPL) and proficiency checks on multi-pilot aircraft.
The JAR FCL will be implemented on 1 July 1999, but Des Payton, head of flightcrew licensing at the UK Civil Aviation Authority, points out a problem. Licences are legal documents, and the JARs as they stand do not have the force of law. The European Commission is in the process of framing JARs for European Union (EU) legislation, but this will not have been completed by 1999.
Meanwhile, individual authorities like the CAA are trying to frame the JAR FCL for national legislation before the implementation date. Converting JARs to legal terminology is complex, involving interpretation of meaning and intention. This is exacerbated by the opportunities for distortion in the translation to different languages. All this worries professional onlookers such as IFALPA, which says that it has already observed that "whole tranches" of JAR FCL are missing from the draft EU version. IFALPA officials believe that the omissions are genuine errors, but add that they will now have to study the EU JARs line-by-line.
The JAR FCL, referring to required pilot performance in the flying test for multi-crew type-ratings or award of an ATPL, states clearly the minima for handling accuracy and the need "to apply aeronautical knowledge".
On the non-technical side, the JAR FCL defines requirements in a fairly traditional way, stating: "The following matters shall be specifically checked when testing/checking applicants for the ATPL(A) or a type-rating for multi-pilot aeroplanes [as] pilot-in-command: management of crew co-operation; maintaining a general survey of the aeroplane operation by appropriate supervision; and setting priorities and making decisions in accordance with safety aspects and relevant rules and regulations appropriate to the operational situation, including emergency situations."
In addition, says JAR FCL, the pilot should:
exercise good judgement and airmanship;
understand and apply crew-co-ordination and incapacitation procedures, if applicable;
communicate effectively with other crew members, if applicable.
In operational JARs (JAR Ops), IFALPA's discussion paper observes, NTS issues are "mainly focused on education in CRM skills." CRM cannot be a catch-all, however. For example, a first officer might show good CRM skills in that role, yet not be command material.
Arguments against NTS assessment at airline level are presented in IFALPA's discussion paper. These include:
not enough is known about NTS to carry out formal assessments of them;
cultural differences preclude NTS assessment standardisation;
the assessment of NTS is subjective;
NTS assessment is easily abused.
Pilot experience
IFALPA's Capt Ted Murphy, a senior captain with Aer Lingus, says that the Association is happy about the way in which the new JAR FCL insists on a formalised structure for the crucial parts of ab initio training for self-sponsored pilots going for commercial-pilot licences, killing off the traditional "self-improver" route which largely relied on accumulating hours by whatever means possible, and "cramming" for the examinations. Murphy asks: "Why should the airline industry be unique in allowing its professionals to train themselves? Would that be permitted for doctors or architects?"
IFALPA is not alone in being uneasy that a large traditional supplier of formally trained pilots to the airlines - the military - is drying up as a source. The airline industry is massive and growing fast as the military supply shrinks.
Murphy points out, meanwhile, that even those airlines which, traditionally, used to have a philosophy of training a proportion of their own pilots from the ab-initio stage "are distancing themselves from it". He gives the example of Lufthansa, which has a training school with its own name, but which has now been structured as a third-party training organisation.
"Pay-for-training", for example to gain a type rating on joining a new airline, is a big issue among pilot unions in the USA, although its use is fading slightly as the industry, now desperate for pilots, starts to foot the bill more often. The US pilots have, however, found it difficult to counter the argument that, on gaining a type rating, a pilot has uprated his earning power, so why should he not invest in it?
In the end, however, the pilots' objection mainly comes from a gut feeling that pay-for-training is just not right. Perhaps Murphy has put his finger on the nature of the problem: if he is right that airlines are "distancing themselves" from flightcrew training, pilots would not be the only ones with that same gut feeling.
Source: Flight International