David Learmount/LONDON
The number of airline accidents worldwide increased by 28% in the 1990s compared with the 1980s, and fatalities rose by 12.5%. A total of 11,950 people died in 480 accidents during the decade. In the 1980s, 10,600 people died in 370 accidents.
But that does not tell the whole story. According to the International Civil Aviation Organisation (ICAO), the number of travellers increased by 32%, the number of flights by 30%, and the average distance flown per journey by 12.5%, so the average risk to passengers fell. Meanwhile, the 1990s saw a gradual improvement in safety rates; passengers at the end of the decade were statistically safer than they were at the beginning, even if the trend is shallow.
But if the 1990s are to be remembered for anything, it will be as a time of reassessment and preparation. A series of powerful incentives took shape globally which will force airlines to improve their safety standards, although the effect of these will be felt only in the next 10 years. It is the worldwide nature of these measures that makes them unique in the history of commercial air transport. All the incentives reflect the stick rather than the carrot approach to persuasion.
Airlines and states with a safety record causing pain for the whole industry will face increasing pressure to improve or they will become pariahs (see 1990s safety milestones box).
Vast differences in safety standards between continents, regions, states and airlines have been apparent for decades, but it took the industry's development into a mass transport medium during the 1970s and 1980s for this to be acknowledged as an issue that had to be tackled. It then took the whole of the 1990s to set in motion a system for controlling safety globally. The next 10 years will prove whether it works.
The 1960s and 1970s saw dramatic improvements in airline safety, largely because jet aircraft, with their far more reliable propulsion systems, were taking over medium- and long-range air transport. Although there was a further safety improvement in the 1980s, the graph was clearly levelling out.
As aviation entered the 1990s, the law of diminishing returns demanded a more focused approach to improving safety. The chosen focus at the beginning of the decade was human factors, because despite the introduction of more modern flight decks, the proportion of accidents attributable to some form of human error remained, as it had always been, between 65% and 80% of all fatal or serious accidents.
Diminishing returns
In the early 1980s, the response was to reduce pilot workload using the new digital automation technology in the Airbus A310 and the Boeing 757/767, but by the 1990s it was clear that this alone was not enough. In 1992, the US Federal Aviation Administration recognised that, 10 years after the introduction of the "glass cockpit" and the digital flight management system (FMS), a review of their strengths and weaknesses was needed.
In 1996, the FAA published a landmark report, The interfaces between flight crews and modern flight deck systems. This revealed that greater automation sometimes increased workload, and the greater complexity of its systems and multi-mode capabilities could produce crew confusion. While making flight and navigation more accurate, the FMS created new types of pilot error, which were a factor in several serious accidents.
The effort to improve flight safety has retained its focus on human factors, but towards the end of the 1990s, the demand was for incident data to be analysed so that future decisions would be "data driven" rather than decided by gut feeling or solved by technology.
That is where the industry stands today, with the focus shifting toward persuading airlines to use customised forms of "flight operations quality assurance" (FOQA). FOQA is a generic term for a system that enables airlines to gather data from aircraft digital flight data recorders or quick access recorders. They can then process the downloaded data with one of several off-the-shelf software tools which recognises operational or mechanical "exceedances", including those that did not lead to incidents, enabling the carrier to build up a picture of where exceedances occur regularly. FOQA can reveal problems ranging from individual pilots' weaknesses to badly designed terminal area approach procedures, enabling the carrier to address the problem precisely, and to eliminate potential accidents.
Meanwhile, engineering and maintenance is a more complex area in which human factors subjected to extensive analysis and working systems have been designed to reduce the likelihood of error escaping notice in a multi-layered system of checks. In the hangar and the flightdeck it has become accepted that it is impossible to completely eliminate error, so all efforts, from equipment design to operational procedures, are designed to be "error tolerant". That is, no single error will lead to danger and if an error is made it will be obvious, and so can be corrected.
The fact is, however, that safety rates have been improving too slowly. With air transport growing so rapidly, the industry has failed to prevent increases in the number of fatal accidents and resulting deaths. It is the number of accidents, not the rate, that affects public perception of the safety of airline travel, according to the International Air Transport Association's director general, Pierre Jeanniot. At IATA's 1997 annual general meeting, Jeanniot said the two greatest threats to a favourable public perception of air travel would be a growth in accident numbers in line with industry expansion, and the failure of ever busier airports to make the beginning and end of air journeys a pleasant experience.
Safety transformation
The transformation of aviation safety from a national to a global issue results largely from the industry's success. As passengers travel ever further in ever larger numbers, they have begun to assume that air transport, no matter where in the world it is provided, is more or less a standard product, with sensible, government-regulated minimum safety standards. Passengers themselves may have not voiced this expectation, but the questions posed during the media inquisitions that now follow all large airline accidents reflect the public's assumption that all airlines should be safe.
That expectation embraces regional and commuter airlines. Although their operations in the more prosperous countries are increasingly being flown by jets, commuter flights are still mainly turboprop- and piston-powered. According to ICAO, propeller-powered aircraft carry about 5% of the world's scheduled traffic, but despite being such a small part of airline operations, they suffered 42% of the fatal accidents in the 1990s (just over 200) and caused 23% of the fatalities (2,830). As ICAO's 1998 annual report observes, they are much less safe than jet flights, but the report does not suggest a reason. It is, however, accepted that it is the relatively unsophisticated environment in which commuter operations take place, more than the fact that they are propeller-driven, which contributes to their accident rate.
There is another factor, however. Aircraft carrying fewer than 30 passengers have traditionally been certificated under less stringent rules than are larger airliners in almost all countries, which means they are allowed to fly with lower equipment redundancy standards. The USA, following several regional airline accidents in 1992 and 1993, gradually began to change its attitude towards certification standards for small regional aircraft. Now, despite protests from the small operators about cost, they are required to be equipped to virtually the same criteria as larger aircraft. The European Joint Aviation Authorities has moved in the same direction.
There are reasons to be optimistic about flight safety in the next 10 years. The world has decided it can no longer tolerate low standards, and hiding behind the quasi-diplomatic immunity that nationality has given in the past will no longer work for airlines abroad. Meanwhile, it is hoped the industry has learned the lesson that new technology can be a good thing, but only if it is designed with human operators in mind.
Source: Flight International
