Skills, Training, CRM, Culture... Safety

[Dispatch Dog]

it sets an example to others: intentionally break the rules and people die, and you go to jail for a long time. forget about your license and your career. where you're going you won't need one.

you seem to want to coddle these people. we do enuf of that in our society and look where it's gotten us. hit 'em and hit 'em hard. very hard.

as to the hiding things, sure, if you're talking about real criminals. in which case the answer is videorecording of the cockpits and random screening of the video and voice recordings.

on your last point, the fact that it is common enuf for people like us to even discuss it openly is beyond comprehension. once again, this is a DIRECT result of lax enforcement. money is not the answer. jail time is. lots of it.

look, we put pilots in jail if they drink alcohol a certain amount of time before a flight. why should their superiors who consciously make a decision to place the lives of hundreds in danger to save three f*ckin dollars get off the hook? oh, because they are executives? wall street bitches? because they answer to the shareholders? f*ck the shareholders. the people you really have to answer to are the pax. period. end of story. absolutely nothing to discuss.

we have created this insane culture wherein we have "come" to believe that on-time performance actually means something. i fly a lot as part of my work and honestly, don't give a rat's ass about on-time performance. ANY flight that arrives safely was a good flight. risking lives to leave on time? i do not believe there is a jury on the planet that wouldn't convict.

I have explained badly:

I'm an advocate of serious heavy punishment in cases where people KNOWINGLY deceive or practice incompetence. In the case of a pilot who knows he is too tired, but because of the airlines practices, also knows he may as well resign before if he declares himself unfit to fly. I can't see punishment of the pilot being as productive as prosecution of the airline that KNOWINGLY creates situations that are inherently dangerous, read: criminally negligent.

Coddling is what we do in the airline industry by letting airlines self-regulate to the degree they are at the moment.

Too easy under employment law for a manager to say that they have no evidence, just hearsay, so cannot take direct action against undesirable cockpit performance.

A pilot that makes a mistake and is subsequently honest about it should not necessarily earn punishment depending on what the consequences are. A pilot that knowingly acts to hide that mistake should face disciplinary action. If they are in real terms actively endangering lives, then there is, imo, a strong case for them to be strung up by the balls in a dungeon.

My point is that airlines can easily, and legally evade the higher cost of preemptive action as they do not have 'enough' evidence of poor performance that may lead to an accident. At the moment, people who speak up, are removed from the operation, or in my case have no alternative than to resign having exhausted all possible reporting avenues which included confronting senior management and authorities... the answer was always the same... we don't see any smoking holes as a result of the issues you have raised, we therefore deem it safe!!!!

http://www.fss.aero/accident-reports/dvdfiles/GB/1995-02-25-UK.pdf

This is the report of a 737 that only just made it after the oil caps were left off after routine maintenance. The report basically determined that the airline failed to monitor a safety sensitive procedure and created a situation where untrained personnel were 'forced' to do jobs they were not qualified or trained to do. Were they prosecuted, or even fined... NO! They were later asked if they had rectified this and on trust, the CAA accepted the airlines answer that they had carried out the AAIB recommendations!!!
5 years later, the same airline continued to do the same thing with the same risks in a number of areas of operation.
I don't think engineers should be punished as much as the individuals in the airline that continued to knowingly 'create' these situations fully expecting that the line employees will take do the jail time when something actually goes wrong.

What I am talking about is a controlled way of monitoring practices, procedures and performance in order to preemptively take action, to actually prevent accidents BEFORE they happen. And yes, SERIOUS punishment to those that WILLINGLY fail to act or use deception to cover the fact that airlines are not fully supporting individual employees in regards to safe opts.

[Dispatch Dog]

BTW, Gabriel, great thread!

I hope it doesn't die away or get hijacked before we hear from more professionals.

Agree with the above posts... did the crew of Asiana deliberately ignore procedure, CRM and too what they considered excessive risks? I doubt it.

Did they work in an environment that allowed bravado, cultural issues and questionable ability exist? I would wager that some manager somewhere said something like: we can't afford to do more. Meanwhile, other pilots in Asiana are probably guilty of thinking: I knew this might happen.

That is the culture that still needs changing. Punishing these pilots won't change much.

[3WE]

The "sink rate for an instant" .. well that sounds like another way of saying a broken clock is right twice a day.

...but if it's just that the battery is dead, all you gotta do is pop a new battery in at the right time....

Seriously, if someone had been glancing at the ASI and NOT relying on autothrottles, it might have been an extremely normal landing...

[Gabriel]

I want to clarify a few things:

First:
Shutting down the wrong engine? ==> Mistake.
Thinking that the AT would keep the speed when it wouldn't in that config? ==> Mistake.
Not aborting a knowingly unstabilized approach when crossing the stabilized approach criteria, knowing that they were required to abort? ==> NOT mistake.

Second:
I don't think that they knew that what they were "carried a great risk or likelihood of serious injury or death", as TeeVee defined "criminal negligence".
In fact, I don't think that it did carry such a great risk.
There were many ways that this accident could have NOT happened even after intentionally violating the stabilized approach criteria.
- The autothrottle could have kept the speed as they expected (it would have if it was in a different mode).
- The could have reacted a bit faster: They noticed the slow speed 8 seconds before impact, called go-around 4 seconds before impact, and added power 1.5 seconds before impact. If they had called and started the go around say 7 seconds before impact, they would have probably gotten away with it.
- The situation could have deteriorated earlier. The stickshaker could have activated just 100ft higher than it did, and that might have given them enough room to recover (after all, they almost did).
- They could have kept a proper monitoring and control of the glide slope and airspeed, even if they were not stabilized, which is what 3WE always does in the sim.

Third (closely related to the above)

If someone had been glancing at the ASI and NOT relying on autothrottles, it might have been an extremely normal landing...

That's correct. And that leads us to the core concept, I think.

In aviation there are a lot of "overkills" to prevent mistakes. Take the landing gear:
FIRST, the crew has to lower it during the approach.
AFTER that, when they run the landing checklist, they have to check it's down and locked.
AFTER that, when they cross the stabilized approach gate, they have to check that the airplane is configured for landing, which again includes the gear down.
THEN, there is the landing gear warning, that works on a combination of thrust and flaps setting (if it looks like they are set for landing and the gear is up, they sound)
THEN you have the Ground Proximity Warning System, that will activate when you approach the ground and the gear is up.

This combo generated a system where landing gear-up "by mistake" is almost an extinct species.

Now, I bet that if pilots were to disregard the above procedure (three first points above), thus knowing that they are intentionally violating the procedure, and just lower it when they see it fit and didn't include it in the landing checklist, 99.99% of the landings would still happen with the gear down. Surely there will be more "oh Gosh" cases where the pilots are saved by the warnings. No worries, we just lower the gear and on we go.

Now, 99.99% of correct landings would still be several landings per year with the gear up. And most (or all) of the "wrong" ones would not end in serious injury or death either. So, by TeeVee's definition, it would hardly be criminal negligence.

Commercial aviation has not become so safe because it's intrinsically free of problems. On the contrary, it's full of problems. But the system is designed with several layers of safety that, combined, produce the amazing safety record that we have today.

SO: THE PILOT IN COMMAND CAN DEVIATE FROM ANY PRESCRIBED RULE OR PROCEDURE IF HE DEEMS IT NECESSARY FOR THE SAFETY OF THE FLIGHT.

This is the same than:
THE PILOT IN COMMAND SHALL NOT DEVIATE FROM ANY PRESCRIBED RULE OR PROCEDURE, UNLESS HE DEEMS IT NECESSARY FOR THE SAFETY OF THE FLIGHT.

I think that a pilot intentionally not complying with the gear procedure, or with the stabilized approach criteria, will have a hard time explaining how this was for the safety of flight, even if the flight ends in an uneventful landing.

And here is where an additional definition of "criminal negligence" comes to play (I know that definitions vary from one place to another):

"The knowledge that what you are about to do is wrong and carries a great risk or likelihood of serious injury or death, or is a gross deviation from what you are required to do or from what a reasonable person would do."

Now, where we charge them with it or not (for the sake of knowing the truth) is worth debating.

[Evan]

SO: THE PILOT IN COMMAND CAN DEVIATE FROM ANY PRESCRIBED RULE OR PROCEDURE IF HE DEEMS IT NECESSARY FOR THE SAFETY OF THE FLIGHT.

This is the same than:
THE PILOT IN COMMAND SHALL NOT DEVIATE FROM ANY PRESCRIBED RULE OR PROCEDURE, UNLESS HE DEEMS IT NECESSARY FOR THE SAFETY OF THE FLIGHT.

So... A pilot can:

intentionally drop the airspeed lower than V APP – 5 kt while maintaining a vertical speed greater than – 1000 ft/mn beyond the 500' gate if he deems it necessary for the safety of the flight?

I can't find a safety scenario that goes with that.

I get your algebra Gabriel but I don't see the real world application there. Maybe I'm missing your point...

[Gabriel]

So... A pilot can:

intentionally drop the airspeed lower than V APP – 5 kt while maintaining a vertical speed greater than – 1000 ft/mn beyond the 500' gate if he deems it necessary for the safety of the flight?

I can't find a safety scenario that goes with that.

I get your algebra Gabriel but I don't see the real world application there. Maybe I'm missing your point...

Yes, you are. Read the line immediately following those that you quoted.

By the way, I'm quite confident that the drop of airspeed below Vapp-5kt was not intentional. But by 500ft they knew that the speed (and other parameters) were not stabilized and that, hence, they were required to go-around. And intentionally they didn't.

[Dispatch Dog]

I want to clarify a few things:

Not aborting a knowingly unstabilized approach when crossing the stabilized approach criteria, knowing that they were required to abort? ==> NOT mistake.

This is what I was implying by getting into the safety culture of the airline:

If (big if) the pilot in question had been advised or even regularly allowed to bust the stabilized approach criteria as it 'always worked out in the end'?
Another aspect of bad safety culture is if he did abort an unstabilized approach and was disciplined or criticized in some way, not for failing to stabilize the approach, but for failing to continue the landing despite an unstabilized approach, the pilot could at least cite undue pressure to ignore set procedures.

I'm fairly sure the pilots would be sacrificed before the airline admits to undue pressure to ignore regulations, and more worryingly, if the pilots feel duty bound to take the full blame to save the company, which is another eastern cultural thing I have seen.

[Gabriel]

This is what I was implying by getting into the safety culture of the airline:

If (big if) the pilot in question had been advised or even regularly allowed to bust the stabilized approach criteria as it 'always worked out in the end'?
Another aspect of bad safety culture is if he did abort an unstabilized approach and was disciplined or criticized in some way, not for failing to stabilize the approach, but for failing to continue the landing despite an unstabilized approach, the pilot could at least cite undue pressure to ignore set procedures.

I'm fairly sure the pilots would be sacrificed before the airline admits to undue pressure to ignore regulations, and more worryingly, if the pilots feel duty bound to take the full blame to save the company, which is another eastern cultural thing I have seen.

Then we are saying the same thing.

Read the few last lines in the first post, starting at "My point, that I have raised before in this forum, is two fold:"

And also a few posts later, where TeeVee proposed charging the pilots with criminal negligence and I proposed charging, instead, the airline or the top managers for the same offense.

[TeeVee]

maybe they need to start using FDR and CVR data BEFORE accidents occur to assure that pilots are doing what they are supposed to be doing. at least during the critical phases of flight.

[Gabriel]

maybe they need to start using FDR and CVR data BEFORE accidents occur to assure that pilots are doing what they are supposed to be doing. at least during the critical phases of flight.

Well, that's pretty much what FOQuA (also FOQA) is about.
Not FDR and CVR because they are sort of tamper proof. But they use the QAR (quick access recorder) which records much more data than the FDR, except it lacks the crash resistance of the black boxes.

[Dispatch Dog]

I want to clarify a few things:

Not aborting a knowingly unstabilized approach when crossing the stabilized approach criteria, knowing that they were required to abort? ==> NOT mistake.

This is what I was implying by getting into the safety culture of the airline:

If (big if) the pilot in question had been advised or even regularly allowed to bust the stabilized approach criteria as it 'always worked out in the end'?
Another aspect of bad safety culture is if he did abort an unstabilized approach and was disciplined or criticized in some way, not for failing to stabilize the approach, but for failing to continue the landing despite an unstabilized approach, the pilot could at least cite undue pressure to ignore set procedures.

I'm fairly sure the pilots would be sacrificed before the airline admits to undue pressure to ignore regulations, and more worryingly, if the pilots feel duty bound to take the full blame to save the company, which is another eastern cultural thing I have seen.

[Gabriel]

Déjà vu, Dispatch Dog?
You've already said this and I've already replied. See posts #37 and #38.

[Evan]

Yes, you are. Read the line immediately following those that you quoted.

By the way, I'm quite confident that the drop of airspeed below Vapp-5kt was not intentional. But by 500ft they knew that the speed (and other parameters) were not stabilized and that, hence, they were required to go-around. And intentionally they didn't.

Sorry, it was late at night. I think I see what you are driving at now.

I think you are suggesting the possibility that the crew were monitoring flight parameters and intended to recover late and therefore continued what they knew was an unstable approach below 500'.

I think we both agree that this was probably due to a case of overconfidence in the automation driving confidence in the safety of continuing this unsafe approach. I think the pilots believed that they were still in an unofficially safe zone for this reason, but this is not and will never be true as long as the automation can fail (for any reason). The pilot is the only contingency for such things and that can never be given up if we want to avoid this kind of accident. After a point, adding manual thrust from flight idle can no longer arrest the descent. Any fault with the automation has to be detected before that point. I would place that point above 500'.

I think we both agree that this crew probably misused (and probably misunderstood) the automation, that the a/t failed to engage for this reason, that the a/t had therefore been at idle for too long at too low of an altitude and that they could not arrest their vertical speed and maintain airspeed as they expected to because of this error on their part.

As another anology: In film photography we have what we call 'latitude' when using modern films. In the past exposures had to be very accurate but today's film fomulations have a 'latitude' that allows to you be off by a full stop (sometimes more) and the 'technology' will have your back, not to mention the push/pull capabilities of processing equipment, scanners and post-production. It has made some film photographers a bit lazy on the numbers, and digital now allows pushbutton shooters with no real education in exposure to be professional photographers. Professional photographers rarely resort to 'hand-flying' their exposures these days unless they are going for deliberate effect. The old rules about 'safe' exposure are not widely taught anymore and so the instincts are fading. When I was starting out, these were fundamental things. I like to shoot with vintage cameras that lack any light metering. I can often guess within a half-stop by instinct.

Automation is making aviation safer, just as it is making photography better (technically anyway), but those instincts, the ones based on fundamental rules of safety, are being traded for the confidence that automation creates. Sooner or later, that automation is not going to be there and you have to have the instinct to know when you are in unsafe territory.

I think this crew may have lost that critical instinct.

[Evan]

Let me try to put this into an alternate scenario where the veteran pilot DOES have a good instinct for safety, but, coming from a weak safety culture, is prone to violating the rules governing a safe approach and is also new to the 777:

ATC vectors them to final in such a way that they end up at the FAP high and fast on a visual approach. The experienced Boeing pilot quickly briefs his crew on the quick solution: they will use FLCH to get things back down fast then hand fly the last 1500' with the a/t engaged in SPD mode. He sets the MCP target to accomplish this, the FLCH mode brings the a/t to idle and HOLD as he intended and he then devotes his attention to correcting a deviation from centerline.

At 1500' when the MCP target for FLCH has still not been achieved (and therefore the a/t has not yet reverted to SPD mode), he disconnects the AP and continues to be task focused on the drift. He checks his instruments, airspeed and v/s regularly and is aware of the evolving situation but he knows the a/t will soon correct for this.

-- What he doesn't know is that the a/t is now stuck in HOLD at idle due to a lack of vertical AP guidance. --

But, despite being part of a bad safety culture and being prone to risk, he has a good ingrained instinct for safe piloting. At 500' he realizes that the speed and v/s are trending wrong. He is still expecting the a/t to remedy this at any moment so he continues the approach BUT he also instinctively PLACES HIS HAND ON THE THRUST LEVERS.

A few moments later, when he has not felt the levers move on their own, even though he does not understand what the problem is, he instinctively advances them manually and keeps them there.

He is below the slope, unstable below 500'. Of course he should go around, but he is task-oriented and culturally prone to risk. Nevertheless, because he had the good piloting instinct to put his hand on the levers and ultimately move them, power comes on in time and he pulls everything together and puts it down a bit hard on the numbers. His colleagues are noticeably impressed. He is a damned fine pilot and they have learned a lesson from him: the 777 is a very capable plane and a really good pilot always saves the approach.

How many times does this happen before it all goes wrong? How impressive a lesson is this? Can training overcome this kind of admiration for the pilot who saves the approach after being well outside the missed approach criteria? Would one of the other pilots, later in his career but maybe one without such good old instincts, feel shamed at the idea of abandoning a similar approach when he is in the left seat, and fly it into a seawall...?

Our strongest teacher is experience and reward. Risk often brings the greatest reward. Maybe nothing can overcome that.

[Dispatch Dog]

Déjà vu, Dispatch Dog?
You've already said this and I've already replied. See posts #37 and #38.

Sorry, was posting from a glitching work computer designed by Airbus using Boeing composites and supervised by an over trained Texan.

Yes, we are thinking along the same lines regarding Airline responsibility, or lack thereof. In my experience, there is a common theme in this type of safety culture, whatever the regional cultural attitudes: The bean counters have the last word and feel they can ignore the safety aspect of their decisions because its the job of pilots and engineers to make their asinine decisions work.

The justification?... "we saved $100m last year and nobody crashed".

This only works because of the lack of evidence of near misses, which the pilots and engineers are 'discouraged' from reporting because they know in many circumstances they will be blamed.

I have applauded the QAR for years in a few forums, but it just doesn't work in a blame culture and especially not in a social culture where the individual must take the 'honourable' path.