The pilots of Asiana Airlines 214, a Boeing 777 that crashed at San Francisco International Airport in July, told investigators the auto-throttle system on their aircraft malfunctioned. They swear it was properly set prior to beginning their approach too, assuming of course that the system would adjust the engine’s power as necessary to maintain safe flight.
The problem is the auto-throttles didn’t work as expected, the airplane got too low and too slow and the pilots never noticed until it was too late.
The Boeing stalled on short final to SFO’s runway 28 Left and struck a dike near the approach end. The impact tore the aircraft into a number of pieces also killing three and injuring dozens of other passengers. The aircraft was a total write-off.
Asiana pilots on earlier flights had reported a few maintenance write-ups for the same problem — a series of “uncommanded auto-throttle disconnects” — as a potential culprit in the accident.
My question … so what?
Pilots are trained from their earliest student flight experiences to understand the idiosyncrasies of the aircraft they fly whether that’s a Cessna 172, a Beech King Air or a Boeing 777. Before a pilot is allowed to command, they’re expected to understand how each and every system aboard the aircraft functions, as well as tricks of the trade to identify when something’s amiss.
Does the mix of advanced automation make understanding what an airplane will do next more difficult than the old days? Absolutely.
Some of those system complexities, as well as our willingness to accept advanced flight training the way it’s delivered rather than with the level of depth we need to stay out of trouble, however, is one part of the problem pilots face today. Providers can’t cover every unusual situation a crew might encounter on a complex automated airplane, so they train pilots to cope with the most likely scenarios. All the training providers expect the crews they certify to at least be able to maintain control of their aircraft on a nice day unless something catastrophic occurs … and an auto-throttle failure is not a catastrophe.
The unexpected situations though? Those answers are left to the crew’s judgment, experience and knack for synthesizing what they know into an answer for the unknown. That’s a pilot’s real job description anyway.
When you look at the number of aircraft that crews have wrecked around the world in the past five years alone though, like the Asian accident, it seems pretty clear we’re giving too many pilots too much credit for their ability to fly the airplane in a pinch … or even in good weather.
Loss of control — the crew’s inability to safely fly the aircraft when the unexpected occurs — has become the biggest flying threat the industry faces.
The real question is why so many pilots, who appear highly experienced on the surface and should be well suited to command an automated aircraft still let the computers get the best of them.
When an onboard system fails, or begins taking the aircraft somewhere dangerous, the captain, the first officer or even someone along for the ride in the jumpseat, is expected to speak up … loudly if necessary. They’re also expected to recognize the situation early enough to maintain control of the aircraft and overcome the problem as best they can.
Seriously, isn’t that why we have humans in the cockpit … even triple redundancy computers fail, or switch systems off in a sequence no training provider could even think of demonstrating.
Perhaps too many pilots have forgotten how to fly, or maybe too many of them fail to recognize when the automation’s in charge and when it’s the human’s turn. Perhaps it’s also time to finally stop assuming that crews with lots of hours in their logbooks make the best pilots.
But if pilots are unable to recognize the dangers their aircraft’s automation might be leading them in to, or unable to convince that automation that the humans up front can do a better job when necessary, maybe we’ve already taken the first step toward creating that automation-monitor reality we fear so much.
Rob Mark, publisher
Alert 5 says
It’s all about training. This statement says it all: “it seems pretty clear we’re giving too many pilots too much credit for their ability to fly the airplane in a pinch …” Another way to say it, pilots aren’t adequately trained. As for the number of airplanes that have been wrecked over the past 5 years, I wonder, how many of those pilots were ex-military and how many were not. That stat would be revealing as most would agree that the military provides the best flight training.
110A says
@Alert5, I have to disagree with a portion of your reply. The military provides great flight training, but not necessarily “the best” for an airline crew environment. I’ve spent many years flying with both military and civilian pilots at an airline. I can say, without doubt, that one group was not particularly better than any other.
Brent says
Rob,
Great post! I happen to agree. We seem to be going backward – better technology, but crappier pilots.
I have noticed that training vendors for pro pilots tend to assume that you know how to fly the airplane and they focus on using the automation so you are comfortable with it. While that makes complete sense, it doesn’t account for what we are quickly realizing – that knowing when to turn it off and do that ole’ fashion pilot stuff is really the most important part.
Brent over at iflyblog.com
Ron Rapp says
While I agree that pilots need to get “back to basics”, it’s also worth pointing out that autothrottles, autopilots, and many other systems malfunction on jets every day and 99% of the pilots notice it and handle it without a problem.
Cruise is probably the phase where we’re watching the systems with the least vigilance. Short final is where we’re watching it most closely, with the pilot-not-flying monitoring speed and glidepath almost continuously. That’s what makes the Asiana crash such an anomaly — it occurred during the exact period when something of that nature should have been LEAST likely.
Very odd.
Ron over at Rapp.org :)
Seth Freeman says
I have worked on one side of the window, with lots of military, civilian pilots. I have found that the Military trained pilots are the worst for following published procedures. I think that the automation the computer inputs have made the pilots too overloaded and reliant on the computer to take care of issues. The pilots rarely understand that they are not flying correctly because they have allowed the computer to navigate for them, and they failed to read the published procedure. In the average week of work I have found that at least one military flight will fail to comply with the published procedures and have to be corrected. The foreign civil pilots rarely fail to follow procedures. In the case of Asiana, the company procedures and the cultural issues of captain and first officer or in this case instructor student, may have limited their good judgement.
The need to fly the airplane first should always be point.
Norman says
There has been a raft of comment around this accident, much of it directed at flight training and the status of the autopilot/flight director system. The position of the FD switches may be critical to understanding how this appalling tragedy happened.
Knowing how to remain intimately connected to your aircraft: be tightly within the ‘loop’ during auto-flight are major questions to be addressed during auto-flight – or even manual flight. Current generation Boeing and Airbus types are designed to be flown through the automation, disengage it and the workload can skyrocket. Add to this distraction, momentary confusion as to who is monitoring and guiding trajectory and we have a nasty cocktail of factors brewing a potentially vile outcome. Let’s wait to see what the final accident report has to say.
jim denike says
look forward to reading the transcripts of the final moments of the CVR tapes. Don’t know of any airline operation that doesn’t dictate an immediate go-around if all parameters -airspeed, rate of sink, glide slope visual or ILS – aren’t dead on at 500’AGL or even higher. Where were the standard call outs for the obvious deviations observed?
Too proud to go around. Better that than have the blood of three people on your hands, to say nothing of the loss of a perfectly flyable airplane.
Don Elliott says
They have not “forgotten how to fly”…they never learned in the first place. Now, finally, the FAA has figured it out and will require more “flying the airplane” than being a Master Monitor and pushing buttons and turning knobs. I always had my hands on the throttles on auto land and auto throttle operations. DUH!
George Semel says
The problem I see is a simple one, to much automation, it the quest to cut down on pilot work load, they have machines to do the thinking now. So the pilot just sits there and manages the systems. Great in theory, but its as stupid as driving a convertible with the top down in a Snow storm in Fairbanks in Feb. How much actual hands on physical flying of the airplane these two clowns had in the last 6 months or the last year? And did they even look out side the airplane at all while on Approach? What the heck happened to not having your hand on the throttles during the approach? No doubt this guys got there training at some factory school from some numb nuts that had ink still wet on his or her certificate. These guys broke a perfectly good airplane, but I has to be somebody else’s fault. The Fix is easy, its called fly the aircraft.
Norman says
We have some pointed assumptions and not a few conclusions being offered here. Has the accident investigation been completed, analysed released for publication?
Robert Mark says
Of course the investigation’s not complete Norman. But that doesn’t mean we can’t look at the obvious clues and draw some conclusions, especially since the NTSB has released so much detailed information from the CVR so early on for us to think about.
As someone told me the other day when we were talking about this Asiana accident, “I don’t know what those three pilots were thinking about during the descent to SFO, but I can tell you they weren’t thinking about altitude and airspeed.”
Why of course we don’t know.
Again though to your point, let me ask. Where is the harm to anyone when a bunch of aviation people begin talking about a high-profile accident like this amongst ourselves?
If I sit in front of a CNN camera and state my opinion, that’s my right, isn’t it? You might agree or disagree, but I don’t see anything wrong with an experienced pilot trying to help the mainstream media make sense out of the intricacies of flying while we wait for that final report.
And BTW … when are you coming to ORD? I think it’s my turn for lunch.
Rob