Traumatisme

AIR SAFETY AND ACCIDENT ENQUIRIES

Though still young as other means of transport go, air transport has in one short century experienced an extraordinary development. At the same time, it is an activity that has become extremely safe, even if some rare disasters, like that of Swissair Flight 111, regularly remind us that safety is a fragile thing that can very quickly deteriorate if we do not continually fight to maintain or improve it. (1)

It is with this in mind that all aspects of civil aviation, from aircraft design to air traffic control, have been standardised, regulated and controlled to ensure the safety of air transport. Very early on, the need was recognised for a permanent feed-back in order to seek the maximum efficiency of this organisation. This required the systematic exploitation of all unusual events, both accidents and incidents, and an organised collection of all the lessons learned into a collective know-how bank. Because they enable such a use of experience, enquiries have become an essential and central tool for air safety.

1 - Organisation of the enquiries

Over the years, a complex organisation has been set up on a global scale under the auspices of the International Civil Aviation Organisation (ICAO), to facilitate the exploitation of accidents and incidents. It is a fact that in relation to aeronautical safety international aspects cannot be ignored, there can be no borders. An incident today in Mexico can forewarn of and help to prevent an accident in India tomorrow, and vice-versa.

In this context, national States are obliged to open an enquiry in case of an accident or serious incident affecting civil aviation. The general organisation of enquiries is codified, as are the rules for notifying events and for presenting and circulating the results. It is specified that an enquiry concerning an accident or incident has as its chief aim the prevention of future accidents or incidents and that moreover this activity in no way aims to determine faults or responsibilities. Lastly, it organises the close association between the State representatives of the aircraft's operator and manufacturer: on the one hand, it provides the State conducting the enquiry with useful competencies and information, and on the other hand, from a safety point of view, it facilitates the fastest and most complete feed-back possible.

In Europe, a November 1994 directive defines the legal framework of accident enquiries that is directly derived from the above rules. It states that the enquiry must be conducted by a permanent specialised organisation (or under its control) that is functionally independent from the aeronautical authorities in charge of airworthiness, air traffic control, etc. as well as any other interested party, and that the report must be made public.

In France, the body in charge of enquiries is the Bureau Enquêtes-Accidents (BEA - Accident Enquiry Bureau), a component of the Inspection Générale de l'Aviation Civile et de la Météorologie (General Inspectorate for Civil Aviation and Meteorology). A lightweight specialist organisation (currently some sixty-five people), but disposing of powerful means, BEA is totally independent in the context of its mission (2); well integrated in the world of civil aviation, it carries out accident and serious incident enquiries and, selectively, certain other incidents (3) and establishes the relevant reports and recommendations. Made up of aviation professionals specially trained for enquiry work, BEA relies on a vast network of civil aviation personnel who have been prepared for this task - trained investigators - and when needed, and under its authority and responsibility, calls on competencies existing within the DGAC (Civil Aviation Authority), for example in the field of air traffic control, the Ministry of Defence, industry, carriers or other professionals. In accordance with international regulations, BEA also participates in enquiries abroad when they involve a "French" designed aircraft (this designation includes Airbus, ATR, Super Puma, etc.), a French operator or French nationals (4).

One should add that a draft law, based on international provisions, confirms the role of the safety enquiry, also called technical enquiry, defines its relations with the judicial authorities and updates applicable provisions, more particularly in the area of communication and dissemination of information. Adopted by the Lower House (Assemblée Nationale) last June, its reading by the Senate is planned for early next year.

2 - The enquiry

In broad outline, we may distinguish three inseparable stages in the handling of unusual events:

- identification and notification of the event
- the enquiry itself (that is the collection and analysis of the facts)

- information concerning the results and lessons obtained.

2.1 - The first stage, obviously, is the quick notification of any accident or incident affecting or likely to affect the safety of an aircraft. Paradoxically, it is generally this first stage which presents the most difficulties. First of all, someone needs to recognise that the event does indeed carry lessons for safety, and this is not always obvious, far from it, specially when we are dealing with incidents. Next, the transmission of information often runs into indifference, individual or collective selfishness, the fear of sanctions or of "ridicule". It is therefore often slow or incomplete, particularly at the international level.

2.2 - After the accident, or in certain cases incident, an official enquiry is opened. For the enquiry proper we may distinguish three slightly overlapping phases:

- the findings: before any interpretation, it is necessary to collect all the relevant facts for a correct understanding of the event. The quicker the evidence is gathered by qualified professionals, the more useful and complete will be the findings. The investigators will ask that the wreck stays in place until they have been able to examine it, as well as the site; they will gather parts or equipment for these to be expertly examined under their control; they will ask that all files concerning the aircraft and its flight crew and air traffic information be communicated to them, they gather testimonies, etc. In a case involving public transport, they also collect flight data recorders. It should be underlined that these flight recorders are extremely valuable tools, considering the complexity of a modern airliner and the need to understand and exploit accidents.
However, and contrary to a widely received opinion, such tools are not the only basis for an enquiry nor even an essential element in the absence of which the enquirers would be totally disarmed (5).

- expert examinations and investigations: this involves more particularly the exploiting of in-flight recordings, the examination of breakages or certain instruments, reconstruction, autopsies, simulations, etc. It also involves synchronising the various pieces of information available (e.g. radar recordings, in-flight recordings, possible structural distortions, etc.), the examination of applicable regulations, of company structures… It is a question of putting together all the facts and understanding them, as exactly and correctly as possible. For this, the investigators will obviously expect frank and unreserved cooperation from the flight crew, airline, flying club or manufacturer.

- analysis and conclusions: after the reconstruction of the scenario of events, or sometimes multiple scenarios(6), this phase involves identifying and understanding the various imperfections or faults of the system (in the widest sense) which may have contributed to the event, promoted it or did not prevent it. It should make it possible to draw lessons from the event for improving safety. It should be noted that with the raising of the level aviation safety and the ever-increasing complexity of on-board systems, it is becoming difficult to be satisfied with just a single "immediate" cause, a cause that somewhat as a caricature I would qualify as anecdotal for safety. An accident is more often the result of an array of different causes whose highly unlikely sequence has nevertheless taken place. Correspondingly, to make any progress and prevent such events in the future, it is necessary to understand how this sequence was able to develop and to be able to go to the root causes. And from this point of view, even when the enquiry has been unable to determine with certainty the causes of the accident, each scenario or each hypothesis carries a message.

I would also like to insist on a point, the social importance of which will be continually increasing. This is the support given to the victims and to their close relatives and friends.

Already traumatised by the accident, these people also often feel the trauma is multiplied in the follow-up to the accident by what they perceive as indifference to their suffering or the unwillingness of a company to assume its responsibilities. This is why it is essential to make sure that good direct information is given and that the families are properly heard; this goes from the immediate communication of the passenger list to the setting up of a quick compensation procedure.

3 - Evaluation of the system

We have seen the wealth of concepts that civil aviation has developed to exploit any failings or weaknesses and turn them into an impetus for progress. Does this mean that everything is going as perfectly as it should? Obviously not.

Without going into details, I will limit myself to giving some general trends we encounter.

Technical imperfections are, generally, correctly identified and corrective measures taken, taking into account possible economic imperatives. This is not however the case when it comes to human failings in a highly complex technical environment. Much thought is being given to this issue in order to be able to better identify these failings, if only for the benefit of human science researchers.

Another weakness of the system is the circulation of information. It can happen that a problem identified in one country or by one organisation is ignored by others. This requires improving the way the transmission of information is organised, while not underestimating the problems this may pose, but also making more effective the evaluation and exploitation of such information by potential users. And, above all, we need to limit the effects of the collective amnesia with which we are well familiar.

However, the chief weakness of the system remains its often misunderstood objective.

As previously indicated, the technical enquiry has only one objective: the prevention of accidents. This is the loop of the feed-back system in aeronautical safety. This activity therefore needs to be undertaken first and foremost for the benefit of those in charge of safety, to whom the information and recommendations need to addressed. Unfortunately, this approach, which is far from the reactions of an emotion-driven society, is not always well understood, and this makes it vulnerable. The common error is to expect the indictment of individuals or institutions considered to be responsible for the malfunctions identified by the enquiry, rather than the correction of these malfunctions. In this context, the reconstitution of the causes leading to the accident - causes which then in the reading of the report appear to become obvious - becomes an indictment for prosecution.

This situation is further complicated by the confusion that surrounds disasters. In the inevitable climate of crisis which accompanies such events and at a time when so many people are reacting emotionally, communication and public debate are easily blurred by misunderstandings or special interests. Communication therefore needs to be organised, if possible in advance, and the help of the media is essential not only to inform but also to explain.

The effectiveness of our action is at stake here. Safety is first and foremost a question of behaviour, both individual and collective; it can therefore not be imposed from outside. If we really want to benefit from failings and make them a driving force for progress, it is not enough to study them and understand them; we also need to convince all safety stakeholders, which in this case means the entire aeronautical community. All those who are in a position to implement the lessons learned must take ownership of them completely, accepting the enquiry's conclusions without reservations, however unpleasant they may be. Never will it be said enough to what degree the alas all too frequent temptation of using the conclusions of a report for partisan or sensational purposes constitutes a serious risk for safety.

Be that as it may, the safety debate picks up strength with each new disaster. Often simplistic and rarely escaping from the old choice between fatality and guilt, fuelled by misunderstandings or interested designs, it threatens the stability of a complex and fragile system. No country, no system is able to avoid this. Under such conditions, do we not run the risk of forcing the responsible administration to in turn seek the favours of the media or to give pledges to its critics, and all the more so if its administrative status predisposes it for such dealings? It is safety that must be the issue of political debate, not the enquiry itself (7).

4 - And what about justice and judicial issues?

This brief presentation would not be complete if it did not also make mention of the objectives of justice. The fact is that we cannot feign to ignore that an accident brings its trail of suffering and damages and for these to receive reparation it is essential that possible faults or responsibilities be identified. In general, this is the role of the courts and to be able to do this they need, just like the rest of us, to understand what happened and why.

This means that after a disaster we are faced with several types of enquiries, having different objectives and approaches which are often contradictory, not to say incompatible. This does not make for an easy situation. I might merely mention the risks of dissimulation or of fierce debates which result from the unavoidable cohabitation of such differing approaches. No type of organisation can avoid this.

Nevertheless, the fact that in France we have two separate enquiries right from the outset offers an advantage that is rarely highlighted: aeronautical investigators do not need to justify the independence of their approach nor do they have to establish a case that is legally receivable; the judge is there for that and it is he who will guarantee the protection of the interests of the parties. While we wait for the new law, instructions have been given to coordinate and organise the action of technical and judicial investigators; experience shows that between responsible persons of good will it is always possible to find a solution for the difficulties that may arise.

To conclude, I could say that safety is something that is built with the bricks and mortar of time and cooperative dialogue, not heated confrontations and media clashes. People in charge, investigators, judges, flight crew, journalists, lawyers, we all have a role to play and a place to fill in the system. But conversely, it is also possible for each of us to pervert it and cause it to lose its effectiveness, out of selfishness and lack of understanding. When it comes to safety, clear-headedness and a sense of the common interest are essential.

Paul-Louis ARSLANIAN

Chief of the "Bureau Enquêtes Accidents"

Inspection Générale de l'Aviation Civile et de la Météorologie - France

Translation by Andrew WILES

1 As a rough guide, here are some of the orders of magnitude: 25 fatal accidents for 15 million annual departures in regular airline service, excluding the CIS; 1000 fatal casualties in public transport for 1.5 billion passengers a year.

2 The draft law on enquiries states that it neither receives nor solicits instructions from any authority.

3 In parallel, all incidents are systematically analysed by the administration, airlines and manufacturers.

4 Thus, over and above its intervention in a little more than seven hundred events a year on French territory, BEA is currently involved in some fifty enquiries abroad.

5 In this regard, please allow me to make a brief comment on the importance given by the media to the shutdown of the recorders on the Swissair MD 11 a few minutes before its final crash. This shutdown, in and of itself, is one of the elements of the enquiry. It is something to be studied and could, like any other fact, contribute to an understanding of the unfolding of the event. However it is at the very least premature to conclude that this loss of information in the last moments of an accidental process, and at a moment when the aircraft was most likely already doomed, is a hard blow for the enquiry. If among the lost data, some elements were important for understanding what happened and thus contribute to improving safety, they are most likely be found just before the shutdown. As for saying that "it would be easy to ensure that the recorders continue right on to the end", let us wait and see how and why they stopped.

It will then also be necessary to establish priorities. Priorities of use of the last fractions of energy of an aircraft, for example, or more modestly the priorities in terms of the needs of an enquiry. My colleagues from NTSB have several times regretted that two aircraft disasters in the US, involving Boeing 737s, remain as yet unexplained, several years later (and here we are only really dealing with assumptions, which are insufficient for obtaining safety improvements), in part because the recorders of the doomed aircraft were only recording a very limited amount of data, contrary to what is observed among the larger European airlines.

6 This point is worth a comment. During the preceding phase, studies and research have made it possible to determine and reconstruct the events which make up the sequence of an accident in the widest meaning of the term.

There may nevertheless remain some moments or aspects which are not determined. During the analysis, the investigators will therefore seek to compensate for these gaps by reasoning (the touchstone being consistency with those points that have been clearly established).

Bearing in mind the inevitable errors that can mar certain data, bearing in mind the subjectivity which is inseparable from any human activity, bearing in mind too the weakness that sometimes affects available evidence, the established scenario may well not be unique and in fact several "probable" scenarios may need to be presented.

7 To assess the effectiveness of safety systems, in particular those relating to feed-back from experience, they need to be studied according to three criteria: their ability to identify the events that carry lessons to be drawn, their ability to analyse them and to exploit the results of such analyses. By insisting on only one of these points of view, which is often the case when one seeks to respond to an isolated concern, one runs the risk of being less effective.