I have posted this elsewhere without response so I thought I would stick my neck out here because a huge number of words have been written recently regarding airline accidents that have involved pilot handling issues.
Many have blamed reduced crew experience and poor training standards driven by cost saving measures as the cause of these errors. Earlier generations of pilots have found some recent failures to control speed and attitude unbelievable, given the presence of two or more licensed pilots on the flight deck. Others have highlighted the loss of handling skills caused by the extensive use of automatics.
Both viewpoints predict a deteriorating future for safety unless we require better pilot training and experience levels and use automatics less.
May I suggest it helps to stand back and look at the full picture that is today’s airline scene and not argue from the particular towards the general, which I was brought up to believe shows lousy logic.
I do have suggestions to improve the situation for both standpoints but first may I explain where I come from, the better for you to understand my thoughts.
Logically every single airliner accident to date has been caused by human error.
There are no exceptions to this because only humans design, build, test, certificate, maintain and operate aircraft and airfields. Human errors or poor performance can clearly be involved in any of these areas. Some may ask what about lightning and birds? My reply is that we know the sky contains those and we choose to fly regardless. Thus aviation safety is about mitigating the effects of inevitable human errors or choices.
Human error can occur in the air and on the ground. The solution to mitigating the effects of errors made on the ground is to involve more people and time in checking and testing processes until the system is effectively free of the results of errors. Note, not free of errors but free of bad results that flow from the errors.
At this point some will doubtless be thinking I am ignoring costs and not in the real world because we will never have enough people and time on the ground. Please bear with me if you feel that because I first wish to establish the ideal system. Costs are a factor that cause us to compromise the design of our system for safe aviation but we should first establish the ideal.
The situation in the air is very different to that on the ground. The number of people that can be involved in detecting and countering human errors is small and limited, plus time available can be very short. Piloting an aircraft involves two very different functions - handling and operating. Handling is about controlling speed, height, attitude and direction of flight. Operating is about everything else from making decisions about ATC, weather, diversions, emergencies, what your company wants or expects, criminal behaviour and so on. The list is literally endless.
Handling errors (where I started at the beginning) tend to be the critical ones but are actually the easiest to eliminate in the long term. Therefore my first conclusion is that in the end all handling should be automatic because automatics do not make mistakes. They sometimes fail but the rate at which such failures occur (and matter) is purely a design spec issue involving redundancy levels. Please note I am not making a case for eliminating pilots only for automatic handing. Pilots are vital for the operation of aircraft but not for handling.
I formed this view back in the mid 1960s when as a safety pilot watching a BLEU trials Comet doing perfect automatic landings in a very strong crosswind, I had to say to the boffin crew that if the system dumped I did not have the skill to do that and would need to choose a different runway for landing.
Since that day I have been convinced that automatics could handle an aircraft better than me and the only issue was whether they could be made reliable enough for certification. As we know BEA was certificated to autoland its Trident passengers back in the 1970s. After 30 plus years of autolands with operators all over the world, has anybody ever heard of an accident during an autoland? What about manual landings over the same period?
In the long term I also believe that as automatics degrade in flight due to, say, loss of air-data input, they should not disengage and give control of handling back to the pilot. Rather they should keep themselves engaged and do what pilots are taught to do in that situation - hold attitude and power setting in this case - while telling the pilots what has happened.
In the short term we are stuck with what we have so what should happen now? Personally I do not think any of the accidents that have given rise to the current debate would have happened to airline pilots if those pilots were also current on small GA types and they routinely practised stalling and unusual positions.
Now costs. In the end the extent to which we deliberately compromise safety by driving down costs is properly down to the likes of the FAA and EASA. It is their job to provide an adequate level of protection for the fare-paying passengers. If they choose to allow two tired technicians to work on their own in the hangar at 0400 then that is their decision and we cannot expect all MROs to determine themselves that such standards are unacceptable.
Similarly if they choose to approve the licensing of pilots of limited meaningful handling currency then that is their shout. My own view is that while there are still aircraft out there that need pilots to handle them, either routinely or in emergency, then recent history shows we have cut costs just a little too far. However despite this an airliner seat is an amazingly safe place to be.
In the end extra safety costs exta money. However do we really believe that (all) current ticket prices could not stand a 10% hike providing that the money did go towards safety? This would be far from impossible to arrange – but that is another topic.