"Unless there is a dramatic improvement in airline safety performance between now and the end of 2010, then 2001-10 will become the first decade since the Second World War when global airline accident rates did not show improvement." You will shortly be able to read those words in Flight International's review of global airline safety in the first half of 2009, where you will also find out why airline safety has stopped improving.
Although the text says that safety improvement had previously been continuous since the Second World War, in actual fact it had improved ever since the Wright Brothers. WW2 was mentioned only because airline operations before the war were at the embryonic stage and statistics are hard to come by.
It's only a few days before the magazine comes out, but lets have some intelligent proposals as to why safety is not improving any more. Offers?
on July 15, 2009 5:47 PM | Reply
It has stopped improving probably because most airlines have lost a lot of money during the decade, and the cost-reduction pressures internally, must be simply enormous. The regulatory (safety) authorities need to step up their own efforts - generally they are rather passive and reactive about new incidents.
on July 16, 2009 12:55 PM | Reply
It may be that we are seeing an indirect consequence of the relentless cost reduction pressures that have been created by practically unrestricted competition, increasing charges and passenger taxation. Everywhere (in the airline industry and without) the accountants have replaced the operational people, and the deskilling of individuals (to being pretty much machine minders or data entry clerks) continues apace. People cost money. Skilled and knowledgeable people cost more money. Ergo, leave as much of a process to a programmable machine and just have a couple of people to keep the machines going......(I find this difficult to square with a relentlessly increasing population, but that's another issue).
Where until relatively recently aviation was well endowed with experience and talent at a technical level (both on the ground and in the air) the experience pool is drying up as the technically experienced people retire. In the cockpit, crews are increasingly composed of people who have little experience of traditional flying skills, in "classic" cockpits, and for whom the term aimanship may mean little more than a spelling challenge. Finally, governments increasingly see aviation as target for tax policies which further tighten the belts of airlines and passengers alike.
Many chickens are coming home to roost now. Look at the frightening increase in Loss of Control accidents where the previous big killer -Controlled Flight into Terrain - has been largely contained.
It's not just the accidents that give cause for concern - the rise in the number of incidents (which don't make the headlines) is bad enough, but a rise against a backdrop of the worldwide drop in traffic makes the real picture even worse.
Until passengers accept that they will have to contribute more to the real cost of a safe industry, and fares reflect this, the airlines fight for survival will always be their priority. Governments likewise must realise that aviation is no longer a cash-cow which can be taxed to raise money for social projects. Otherwise, the incident/accident rate will remain at a level which does nobody credit and may be as deadly to the survival of some airlines as economic pressure alone.
on July 16, 2009 11:37 PM | Reply
As a 25 year major airline pilot, most would predict that I will say that management, pay and economics are driving the safety record. I don't think so. Airline flying is changing and the industry is having trouble coping with it. The aircraft capability has sometimes outstripped our ability as pilots to deal with it. I fly a NextGen 737 and I have had copilots show me new display modes and computer capability that I've never seen before (or seen mentioned in training). I think the automation has vastly improved our safety but at the same time, when it starts to degrade, it's very, very confusing to everyone. As these automated aircraft start to age, they get kind of quirky. Sensors stop sensing, little gremlins in the ducting start causing mischief and leave everyone including the new diagnostic computers scratching their heads.
So yes, pay is worse, we work harder and the airline doesn't try as hard to keep the planes placard free. But the guys I work with are pros. They are every bit as good as the guys who came from the f-4s and F-16s in my generation. They just flew in BE-1900s and ATRs. I continue to believe that our industry needs to upgrade the training and the understanding of automation and systems so that we can get a handle on what's going on when the systems start to degrade and fail. Among the most common heard phrases in a highly automated cockpit is "Why is (isn't) it doing that?"
Randy
on July 17, 2009 2:51 AM | Reply
Do not know the numbers, but I suspect spatial disorientation is a contributing factor, and avoiding this requires regular practice in blind manouvering.
I alluded in the AF477 discussion to the point that too little real flying takes place these days.
The Emirates 345 incident pilot made the point that he had the maximum 30 hours flying duty for that month, which in reality is just two flights with three other crew involved, so how much "flying" was involved per member?
Understandably most flying takes place in the simulator these days, but in my opinion there is a fatal flaw with the methodology.
Surely simulation should involve all aspects of a flight, including exhaustion and sleep associated problems.
Flight crews suffer jet lag as much as passengers, so I suggest part of the flying training should take place with the pilots under those constraints.
Maybe a wartime "scramble" scenario once in a while where the crew are dragged into a simulator and given the treatment when they are not at their best like say 3.00am.
on July 17, 2009 4:51 PM | Reply
As each new generation of airliner is born the technology used is becoming more and more intricate with increased emphasis being placed upon software and automated systems. Subsequently, aircrew (as capable as they are) cannot possible be expected to understand what goes on behind the scenes of their displays and controls. The result of this is that aircrew are less and less able to troubleshoot and understand what is happening when a system failure occurs, especially if it results in a cascade effect.
How many times have you used a computer where it doesn't do what you want, when you want it and cry out in exasperation "What are you doing?".
The traditional analysis tools and techniques that we in the Safety Engineering world have used over the previous decades have not moved on and kept pace with current technology. Subsequently, we are rapidly approaching (if in fact we haven't already reached) a critical tipping point where the technology used is beyond our abilities to confidently analyse every perceivable failure mode that is possible.
For many years now it has not been possible to fully exercise a software-based system through all the possible failure modes and hence the techniques used have attempted to concentrate on the most likely or most severe consequences. However, modern systems are now so integrated and complex that it is not practicable in terms of time and cost to fully model and analyse even these more hazardous consequences as fully as we would perhaps like prior to a system entering operational service.
In addition, and perhaps compounding the problem, the traditional certification and safety Standards have not evolved quick enough to keep pace with the ever more complex systems used.
The consequence of all this is that systems and aircraft are being developed and certified as airworthy using standards and techniques that you could argue are just no longer fit for purpose. We therefore cannot guarantee that, over the hundreds of thousands of flying hours of a typical airliner, an undetected 'bug' in the system won't rear it's head and result, through some unexpected chain of events, in a Catastrophic outcome.
And so the answer? Well hopefully a new set of international safety and certification standards (along with new tools and techniques) that are capable of addressing the complexities of modern systems, such as, for example, the long awaited DO-178C for Standard for Airborne Software.
However, for now until we get this then the high levels of flying hours per year mean that accidents due to complex system failures are statistically inevitable; but wouldn't it be good to one day be able to say hand on heart that "We really have done all that is practicable to prevent them"?
on July 17, 2009 10:23 PM | Reply
Let’s look at the most recent accidents and with the exception of USAir in the Hudson the one element that they seem to all have in common is a lack of airmanship. The Colgan Crash was clearly caused by two pilots who reacted very badly to an impending stall and turned what should have been a non-event into a fatal accident. The NTSB information released today on the Continental Crash in Denver shows that the Captain never applied Right Rudder. We don't know what happened with Air France 447 but it’s not impossible to think that the crew mishandled the loss of Air Data information which resulted in a departure from controlled flight. Turkish 1951, Emirates 407, FedEx 80, Spanair 5022 and on and on. We are seeing far too many accidents where pilots (and I'm one at a US Major Airline) are at the root cause. Whether this is because standards have fallen, training isn't properly rigorous, or the quality of applicants has dropped due to the pay and working conditions now prevalent in the Industry (Just look at the Colgan crash) I'm not sure but I strongly suspect all three are related.
In our rush to cut costs training events have been made as short and as brief as possible which reduces the quality coming out the other end. Add to that the lower standards mandated when you have trouble finding qualified applicants because the entry barrier is high and the career field no longer attractive and that mix will create accidents.
on July 18, 2009 4:32 PM | Reply
Money - airlines losing money - training footprints reduced to bare bones. Older planes - maintenance investment reduced, MEL deferrals increasing.
Training - Inadequate training on human interface with regards to identifying and handling automation failures.
Fatigue - more airlines flying to regulatory limits as opposed to using scientific based fatigue mitigation strategies and scheduling.
Morale - pilots are flying distracted with all the uncertainty of their careers getting flushed...
Insufficient investment in good hiring, training and mentoring of pilots at entry end of the industry. Minor league airlines - the Colgans of this world - couldn't give a damn about pilot working conditions and a worthwhile professional career, and set up their crews to fail right from the start.
A case of failed shared responsibility, and arrogant complacency, all around.