The report of the investigation into the Concorde crash in France has laid bare the conflict between judicial requirement and flight safety.
Publication of the final report on the 25 July 2000 Aerospatiale/British Aerospace Concorde crash has given the UK Air Accident Investigation Branch (AAIB) an opportunity to record, formally, its dissatisfaction with the French investigation process.
It would be easy to assume that this was a culture clash. In a sense it was, but the main problem at issue was procedure: specifically the manner in which the French legal system impinges, from the start of an investigation, on a process which the UK believes should start as a gathering of technical facts and operational information by specialists. Once that technical investigation is complete, the UK believes, the legal system is better able to do as it sees fit with information derived.
Ken Smart, head of the AAIB, did not hide dissatisfaction during the early stages of the Concorde inquiry. At the November 2000 International Air Accident Investigators' Seminar in Shannon, Ireland, Smart railed against the delay in establishing facts caused by the French judiciary's intervention. At the time he described the situation at the accident site, and in the hangars where wreckage was gathered, as one in which the judiciary, not being specialists, did not know what might be significant. So, he said, judicial officers waited until a technical investigator showed an interest in an object, then they registered it as potential evidence and removed it to preserve its integrity and/or for forensic examination. This prevented the technical investigation from proceeding until the judiciary had finished with any significant piece of the crashed aircraft. In response, Smart's counterparts in the Bureau D'Enquetes Accidents (BEA) were phlegmatic. They did not attempt to explain or justify this procedure, presumably because it was accepted that the judiciary had a right to act as it did.
The AAIB has also recorded the fact that, in two areas, it would have carried out the investigation differently, relying more on testing than on scientific theory. This, however, is a minor issue compared with the enforced restriction of access to evidence, and would not materially have affected the determination of primary causes of the accident. That is a difference of investigatory style, and is to be expected.
Meanwhile, appended to the BEA's report is the AAIB's list of what it believes are transgressions by the French judicio-technical process against the provisions of the International Civil Aviation Organisation's Annex 13 to the Chicago Convention, to which France is a signatory. This documents the standards and recommended practices for accident investigation worldwide. The AAIB's "observations" cite withholding of evidence from an agency that, under Annex 13, had a right to examine it, particularly in this case where the aircraft was designed and certificated by both countries. Even where the UK was the design and certification lead, like the engines, the AAIB was not allowed to examine the evidence.
The crucial issue is what accident investigation is deemed to be for. It should be its task to find out what happened with the aim of preventing it happening again. Investigators are not tasked with according blame, and may report a "probable cause". Courts have to rule guilty or innocent, not on probabilities. By their nature, judicial processes are careful and pedestrian, because verdicts can be reversed on procedural technicalities. This is incompatible with technical investigations where safety is at stake, because in the early stages of an investigation, a crucial task is to find out if anything clearly went wrong so that operators of the type can be warned to inspect specific components or change a procedure.
Any national accident investigation system that prevents the early determination of potential causes puts travellers at unnecessary risk by slowing the process of discovering potentially dangerous faults. In Concorde's case, this situation was avoided by grounding the entire small fleet of an elite aircraft type. If a Boeing or Airbus type, of which thousands are operated worldwide, had crashed, grounding would not be so easy, and delay in accessing evidence might put passengers at risk.
A legal attitude toward accident investigation is bad for flight safety. A case in point was the Silk Air Boeing 737 crash in Indonesia. The investigator refused to give a probable cause. He ruled, in effect, that there would be no "verdict" because there was no legal proof.
ICAO's recent authorisation as a safety auditor for state oversight standards is the beginning of a global acceptance by states that once signed, the Chicago convention cannot then be ignored. ICAO should now be authorised to have a presence at investigations to ensure that signatory states base their investigation procedures on Annex 13.