Moderators: jsumali2, richierich, ua900, PanAm_DC10, hOMSaR

 
sk909
Topic Author
Posts: 244
Joined: Mon Nov 21, 2005 7:38 pm

Automation Catch 22

Tue Jul 23, 2013 1:39 pm

The last decade or 2 decades, airplanes have improved in quality, reliability and flyability. Computers have taken over more and more functions, making pilots more of observers. One advantage of the automation is that planes are flying more efficiently, and are more economical. But has is also started a unforeseeable negative trend?

In the last 2 decades there been a few incidents/accidents, that must raise some questions wether the automation and computerization is really the right way to go? And have pilots been reduced to mere observers and knob turners?

Especially interesting is it to observe all the experts after the accidents, saying: "Fly the plane!". But is it really a possibility? Does the pilots have the skills to do it? Certainly they have the knowledge, but maybe not the skill.

Everybody on this forum, or almost everybody, knows the overall design philosophy of Boeing and Airbus. Boeing with the overall design philosophy, that the plane can do most of the flying, but in case of some kind of emergency, the pilot can fly the plane. Hence Boeing presume the pilot has the skill to fly the plane.
Whereas Airbus has the overall design philosophy that the plane cannot exceed certain design parameters. Giving the pilot limited access to flying the plane, but also theoretically limiting the plane for crashing. But again presuming the pilot has the
Both philosophies have in their own some advantages, BUT also some flaws. But are they both driving the industry in the same direction? As pilots are getting less and less actual stick time, their flying skills are reduced. Hence both Boeing and Airbus philosophies are flawed because they both presume the pilot has the skill/confidence to actually fly the plane.

Looking at the Air France 447 accident and the OZ 214 accident, the overall conclusion is; the pilots didn't fly the plane. For the AF 447 accident, the pilots weren't skilled enough. I am sure had they been asked the question as to how fly the plane without instruments they would have said; 85% throttle, 5 degrees up. But they weren't skilled/confident enough to make that decision. As for the recent OZ 214 accident in SFO, again the pilots weren't skilled enough to make the decision to disconnect the auto-pilot and land the plane, or make a go-around.

After having analyzed 30-40 accidents the last 20-30 years, it seems that pilots are relaying more and more on computers, and have lost the fundamental skills of flying the plane. Is the answer to this problem more computer-control? Or is it a re-skilling of pilots. Because the philosophies of Boeing and Airbus take their off-set in an era of skilled pilots, skilled pilots designing the system. But in reality, the pilots have lost the flying skills because of the automation of the planes. Will this have an impact in the next decades? I do think that we will see more accidents because of mis-communication between planes and humans.

So how do we change this trend? Giving pilots extra time in simulators? Or maybe is the answer to force the pilots to fly the planes. This will make the efficiency go down a little, but isn't it worth it to save human life? Or have we really become a full-blown bean-counter society, only focused on the money, and devaluing human life? I hope not. I would for sure pay an additional $5-$10 to finance the extra cost of having skilled pilots.

Letting pilots fly/land the plane manually every day would ruin the efficiency of planes, making prices sore. But an international agreement stating that pilots will have to fly/land manually on every thursday, when the weather is good (this is of course subjective). The pilot will have to take it down from 30,000 feet and land it; manually. This would allow pilots to accumulate the skill and confidence, they need in case of an emergency, to disconnect the autopilot and fly the plane.

I am of course interested in all of your comments. Lets find a way out of this morass.
Life's for Living!
 
wingman
Posts: 3950
Joined: Thu May 27, 1999 4:25 am

RE: Automation Catch 22

Tue Jul 23, 2013 2:22 pm

My comments from the Economy section:

I'm not sure there is a morass, as you say. It's never been safer to fly and the statistics prove that we live in an era of almost completely safe air travel, measured by fatalities at least. But the AF and OZ crashes certainly highlight your point. It's hard to see how these professionals can seemingly abandon what appear to be the most critical Day 1-100 flight school lessons: airspeed is good and when you stall point the nose down. Based on my limited knowledge, the crews in each incident could have, and should have, applied their skills to achieve the desired result.

I know that's an absurdly simplistic comment but when you place these two incidents side by side against Capt. Sully, the difference in experience and what masterful hand flying skills can do to save a stricken aircraft is stark. Nevertheless, in the great ROI calculator in the sky, how could you possibly implement such a training regimen? It would be beyond the economic ability of almost any carrier to implement and based on the number of incidents per thousand, and the already safe conditions we fly in, does it make any sense? I think the answer is no, but I would definitely be impressed by airline that could state that each of it's pilots undergoes 5 hours of glider flying each month.
 
RussianJet
Posts: 5983
Joined: Sat Jul 21, 2007 4:15 am

RE: Automation Catch 22

Tue Jul 23, 2013 2:28 pm

Oh yeah, all pilots do is sit there watching a computer and push the odd button. Everyone knows that.   

From what I understand, most pilots try to hand-fly as much as possible within reason, and as the poster above says - flying has never been safer. Non-issue.
✈ Every strike of the hammer is a blow against the enemy. ✈
 
0newair0
Posts: 417
Joined: Fri Jun 01, 2007 12:21 am

RE: Automation Catch 22

Tue Jul 23, 2013 2:30 pm

Very well thoughtout question. My quick take on it is that it's, i'm going to use a harsh word, absurd to regulate based on "one off" events. If events like the Asiana/Air France/Colgan Air crash happend regularly, it would be a completely different story; however, these crashes are very rare and air travel is the safest form of transportation on the planet. Even safer than walking in some places.

The airlines will learn from these events and they will use these events to better their pilots.

No international regulations need to be made because 4 pilots, for instance, didn't notice that the autothrottles weren't engaged on a 777.
That's not how this works! That's not how any of this works!
 
CosmicCruiser
Posts: 2438
Joined: Tue Feb 22, 2005 3:01 am

RE: Automation Catch 22

Tue Jul 23, 2013 2:51 pm

That's already in place. At my previous airline the training dept said when the situation ALLOWS we recommend disconnecting the A/P & A/T and fly the jet. In high density terminals the automation definitely should be used to allow both crewmembers to focus outside.
 
UALWN
Posts: 2186
Joined: Mon Jun 01, 2009 3:27 pm

RE: Automation Catch 22

Tue Jul 23, 2013 3:21 pm

Quoting wingman (Reply 1):
I know that's an absurdly simplistic comment but when you place these two incidents side by side against Capt. Sully, the difference in experience and what masterful hand flying skills can do to save a stricken aircraft is stark.

In the context of the discussion in the thread it's worth mentioning that US 1549 remained in normal law until the end, with the automation helping Capt. Sully achieve his goal of a clean ditching. Actually, the alpha protection kicked in at an altitude of 150 ft and prevented a stall that could have resulted in a more violent impact with the water.
AT7/111/146/Avro/CRJ/CR9/EMB/ERJ/E75/F50/100/L15/DC9/D10/M8X/717/727/737/747/757/767/777/787/AB6/310/32X/330/340/350/380
 
747megatop
Posts: 1785
Joined: Wed May 23, 2007 8:22 am

RE: Automation Catch 22

Tue Jul 23, 2013 3:28 pm

Quoting sk909 (Thread starter):
I am of course interested in all of your comments

I would say, fix this issue if we start seeing many accidents, right now as others stated flying is the safest form of transport. The reality is that the amount of in built redundencies; training; preventive measures incorporated after an accident; strict enforcement etc. that are there in the aviation industry are not there in any other industry perhaps other than the nuclear power generation industry (not even the health care industry has such measures) making aviation THE safest industry.

On the other hand I would say increase automation to 100% and take humans out of the loop in the vehicle that drives you to the airport  . The drive to and from the airport is more dangerous and hopefully this can be made safer in the near future with the Google Driver commercially available now.
 
sk909
Topic Author
Posts: 244
Joined: Mon Nov 21, 2005 7:38 pm

RE: Automation Catch 22

Tue Jul 23, 2013 4:29 pm

The interesting thing is that this is an issue, raised by experts, although many of you think this is a non-issue. Several reports have been published on this matter. Even major US government institutions have looked at this issue, and raised their concerns.

My points are not based on 2 or 3 accidents, but on a general notion that generally pilot skills are declining, based on unstable approaches, incidents ( that doesn't make it to the media) and major incidents.

Quoting 747megatop (Reply 6):
I would say, fix this issue if we start seeing many accidents

Why not fix the problem before it becomes a problem? This is what we every time, fix the issue when 100, 200 or 500 people die. So as long it is not you, it is ok? Could we have foreseen some of the previous issues, should they have been fixed before people died, or are we ok with burning a few people on the alter?

Quoting 747megatop (Reply 6):
On the other hand I would say increase automation to 100% and take humans out of the loop in the vehicle that drives you to the airport

Absolutely.

Quoting UALWN (Reply 5):
Actually, the alpha protection kicked in at an altitude of 150 ft and prevented a stall that could have resulted in a more violent impact with the water.

Great point. But wasn't it his knowledge, skill and confidence that gave him the ability to do this? He was a skilled instructor, knew the systems of the aircraft, how it worked and could use it. Had it been another pilot, the out come could have been very different. Many experts agree on this, hence skill and confidence played a major role.

Quoting CosmicCruiser (Reply 4):
In high density terminals the automation definitely should be used to allow both crewmembers to focus outside

Absolutely right. But you have many, many many airports that are not in high density terminals.

Quoting CosmicCruiser (Reply 4):
At my previous airline the training dept said when the situation ALLOWS we recommend disconnecting the A/P & A/T and fly the jet.

Sounds very encouraging. But there are many airlines that don't allow this, as autopilot is a lot more economical. Especially LCC are prone to discourage this.

Quoting 0NEWAIR0 (Reply 3):

It is not absurd, unfortunately. As I have stated prior it is not a handful of accidents, but a general tendency that new generation pilots are able to control advanced airplanes, but are not skilled and confident in using stick-and-rudder.
Unfortunately airlines don't always learn themselves. They need a little help. The reason they don't learn is that they try to calculate the cost/risk of accidents. And setting up extra pilot skill is expensive it is not feasible to setup a schema to give the pilots the skills.

Quoting RussianJet (Reply 2):

No pilots don't just sit on their a**. And the fact is, new generation pilots are not trying to get as much stick-and-rudder time.

Quoting wingman (Reply 1):

Sure airline industry is safe. But does "safe" give us the right to relax, sit back and say "hallelujah we are safe". Or is it the right time to foresee what could happen, improve it and be on the forefront? As I have said, I wouldn't mind spending $5-$10 extra to have that extra security.
Life's for Living!
 
747megatop
Posts: 1785
Joined: Wed May 23, 2007 8:22 am

RE: Automation Catch 22

Tue Jul 23, 2013 5:58 pm

Quoting sk909 (Reply 7):
Why not fix the problem before it becomes a problem? This is what we every time, fix the issue when 100, 200 or 500 people die. So as long it is not you, it is ok? Could we have foreseen some of the previous issues, should they have been fixed before people died, or are we ok with burning a few people on the alter?
Quoting sk909 (Reply 7):
The interesting thing is that this is an issue, raised by experts, although many of you think this is a non-issue. Several reports have been published on this matter. Even major US government institutions have looked at this issue, and raised their concerns.

My points are not based on 2 or 3 accidents, but on a general notion that generally pilot skills are declining, based on unstable approaches, incidents ( that doesn't make it to the media) and major incidents.

Definitely, prevention is always better than cure. We should always strive to continously improve things and always stay ahead of the curve thereby improving things before they happen. Reason i said fix it we start seeing many accidents is because we have to strike a balance between "perceived" problems and real problems. Reason i am saying that is a PILOT by definition is someone who can fly the plane and use automations as tools with varying degrees to do his job isn't it?. I'll let the real experts on this forum comment on this as i am no expert. As far as the OZ incident goes if 3 pilots on the flight deck could not spot that the aircraft was well below it's landing speed it beats me and begs the question..what the hell were they doing? If the job description of a pilot has really changed without our knowing and is relegated to that of button pushers as your post suggests then yes, i think there is something fundamentally wrong; but the question is, is that the case? OR are the experts raising questions about the level of automation without manual backup as in the case of Airbus planes versus Boeing's philosophy of providing automation but leaving a manual backup avenue for the pilot to take full manual control and hand fly the plane.

In addition to the automation concern there can be many other problems in a similar vein that can be argued about as a disaster waiting to happen and shouldn't it be fixed? -

- outsourced maintenance. I am sure folks in various airlines will be worried about questionable maintenance practices in other countries where aircrafts are flown to. We know that faulty maintenance was the cause of the Alaska Airlines crash in 2000. How many such faulty maintenance incidents have occurred since then but not resulted in a crash, i don't know (at least one comes to mind though..the recent engine door left open incident in LHR emergency landing).

- over congested air space with over worked air traffic controllers. I am sure near misses are a very regular feature, more regular than we are aware of. I would say fixing this deserves the top most priority over anything else.

- exhausted and over worked pilots; Air India Express 812 crash is a stark reminder that an exhausted & sleepy pilot is the most dangerous thing (i would probably as a passenger have more trust in a computer landing the plane versus a exhausted pilot) in an already complicated & dangerous phase of flight (landing phase). Was exhaustion & disorientation a cause in the OZ crash, only the final report will tell us. Should we fix this? Definitely yes on a priority basis. Does not take a rocket scentist to figure this out.
 
cbphoto
Posts: 1258
Joined: Sat Dec 20, 2003 6:23 am

RE: Automation Catch 22

Tue Jul 23, 2013 7:19 pm

If you look back in time and history, you will see that back before automation, pilot error was a huge problem in fatal crashes, that happened quite often. In the past decade or so, between the advances of automation, CRM training, and pilot training, the skies are safer then ever. We will never have a 100% safety record, and it would absurd to even try and achieve that. However, if we can continue to strive and make the number of incidents or accidents smaller, we will continue to head in the right direction. You can thank automation for that huge leap in the safety record.

In today's world of heavy automation, pilots need to constantly maintain their stick and rudder skills, however, a lot of international companies don't allow that. I know of a number of big, world class airlines that specifies in it's operating specs, that at 400 ft, the pilot must engage the autopilot, and is not allowed to disconnect it until 400 ft on landing, in normal operating conditions. How are pilots ever supposed to maintain their skills, when the autopilot is flying the aircraft?

This brings me to my final topic of culture which is the biggest issue plaguing the industry today. While most of North America and Europe have dealt with the culture issues in the flight deck, there are many parts of this world that have not. Asia, Africa and India are all great examples where the culture is still from the 70's. It's things like, the Captain is a God and should never be questioned, or policies that discourage the use of CRM skills that plague these parts of the world. Talking to a few friends who fly in these parts of the world, they all say it can get very tense and nerve wracking flying for these companies. What do you do, speak up and risk loosing your job, or don't say anything and hope you don't crash? If the culture in these countries could be changed and if the airlines and governments embraced the open culture of CRM, I think you would see the number of crashes decline even more.

Automation today, while not perfect, has saved many many more lives, then it has killed people. That I can say with 100% accuracy!
ETOPS: Engines Turning or Passengers Swimming
 
cuban8
Posts: 253
Joined: Wed Sep 09, 2009 7:17 pm

RE: Automation Catch 22

Tue Jul 23, 2013 8:27 pm

I for one have been arguing on this forum against several other members as I don’t believe flying skills to be the main factor behind these accidents. The main problem is in my opinion the man/machine interface and man/aircraft knowledge which people commonly express as situational awareness.
If we compare the 3 mentioned cases of AF 447, OZ 214 and US 1549 you have actually 3 different starting points:
AF 447 – The pilots were not sure what was going on.
OZ 214 – The pilots thinks they know what is going on.
US 1549 – The pilot’s knows exactly what is going on.
My conclusion is that I believe that both the AF 447 and OZ 214 crash could have been avoided if they actually knew their situation awareness at that point. Therefore I don’t see the lack of flying skills being the root cause, but more likely an additional factor to these accidents.

Quoting sk909 (Reply 7):
No pilots don't just sit on their a**. And the fact is, new generation pilots are not trying to get as much stick-and-rudder time.

The problem is that flight training is a matter of cost today. A pilot’s license is in a way not different from any other license; people try to get it as cheap and as fast as possible. Thereafter, the average pilot tries to fast track to an airline to recover the loan on that pilot training. In addition to that, there are fewer government and military flight schools who has “unlimited” funds. The requirements to become a pilot are lower, the quality of many flight school as well, and basically a person with $100.000 in his pocket can become a pilot as of today.
When business goes to hell, you get rid of three things. Your private jet, your yacht and your mistress..........and most importantly in that order.
~ Russian Billionaire ~
 
User avatar
airmagnac
Posts: 451
Joined: Wed Apr 18, 2012 10:24 pm

RE: Automation Catch 22

Tue Jul 23, 2013 9:02 pm

Quoting sk909 (Thread starter):
have pilots been reduced to mere observers and knob turners?

No, automation has taken over the repetitive, high frequency, short term tasks, and freed the human pilots to do what humans do best : anticipate, plan, prepare the long term actions, ie those decisions that occur over several minutes or hours. There is indeed much less visible, concrete activities like constant fiddling and fine-tuning of knobs (I'd say there was actually **more** "knob-turning" back in the day), and more abstract "thinking" tasks. But it doesn't mean the humans are useless.

Quoting sk909 (Thread starter):

In the last 2 decades there been a few incidents/accidents, that must raise some questions wether the automation and computerization is really the right way to go?

Completely meaningless without a global picture. Which accidents ? Out of how many accidents in total ? How does that compare to the increase in number of flights ? What role did automation play in the accidents ? How does that relate to the increase in the number of automated functions ? ...

Quoting sk909 (Thread starter):
Everybody on this forum, or almost everybody, knows the overall design philosophy of Boeing and Airbus.

Oh yeah ? From what I read, a majority of members on this forum cling on to the same old BS about "pilot authority"/"pilot's airplane vs engineer's machine"/... that you mention in your next sentence. It was BS 25 years ago, it's still BS today, but few people seem interested in asking themselves what flight controls are all about, or what exactly is the difference between A and B.
The ONLY difference between the two - yes, the single, one and unique difference - is that on an Airbus, the conversion from attitude target to control surface deflection angle is calculated automatically. Whereas on "older" planes and Boeing FBW, the pilot still does the calculations. That's it, period, done. It's a systematic task, repeated at high frequency, ie requires lots of calculation ressources but little or no variability. A perfect job for a computer, not for a human. That Boeing chose to leave it to the pilot is just a matter of brand continuity.All the stuff about pilot authority is just a load of crap : as the calculations are always the same and allow no variability, the pilot has no authority over them that could be taken away.

Quoting sk909 (Thread starter):
Giving the pilot limited access to flying the plane

Not specific to the Airbus. The limits are physical, they apply to any airplane just the same. A stall is a stall, and you don't stall an airliner on purpose, even a Boeing.

Quoting sk909 (Thread starter):
but also theoretically limiting the plane for crashing

That would be great. But I'm afraid an Airbus can crash just as well as any other airplane.

Quoting sk909 (Thread starter):
Both philosophies have in their own some advantages, BUT also some flaws

No objection there. Humans are niether omniscient nor omnipotent, everything we build has flaws

Quoting sk909 (Thread starter):
they both presume the pilot has the skill/confidence to actually fly the plane.

Of course they do, just as the manufacturer of any tool expects the user to know how to handle the tool properly to avoid hurting himself or others. It's just as true for a hammer as it is for an airplane. The difference is that for the more complicated tools, the operator needs to be provided with the proper instructions, possibly completed by proper training. And the tool itself must be designed properly.
Note the use of the word "proper" - see below.



Quoting sk909 (Thread starter):
the OZ 214 accident, the overall conclusion[...] As for the recent OZ 214 accident in SFO, again the pilots weren't skilled enough to make the decision to disconnect the auto-pilot and land the plane, or make a go-around.

We know next to nothing about the Asiana accident, but you can already state "overall" conclusions ? really ? Or are you just trying to fit reality to your argument ?

Quoting sk909 (Thread starter):
Looking at the Air France 447 accident [...] the overall conclusion is; the pilots didn't fly the plane. For the AF 447 accident, the pilots weren't skilled enough. I am sure had they been asked the question as to how fly the plane without instruments they would have said; 85% throttle, 5 degrees up. But they weren't skilled/confident enough to make that decision

Judging from what you say afterwards, I guess you mean that they realized what situation they were in, but were unable to apply the proper actions (learned as basic airmanship) to correct that situation. And this inabillity was due to their knowledge being eroded by too much automation.
That's not even close to what happened. Due to a confusing initial event, amplified by mis-communication between the crew members, and miscommunication between the airplane and the crew, the pilots never managed to identify what situation they were in. You can have all the stick and rudder skills in the world, if you don't know what your situation is, there is no way you can use those skills to make corrections.
Situational awareness - that's the key. And that's the problem.

Quoting sk909 (Thread starter):

After having analyzed 30-40 accidents the last 20-30 years

Making general conclusions based on such a small sample is not a good idea...

Quoting sk909 (Thread starter):
lost the fundamental skills of flying the plane

You believe the problem is that pilots cannot replace the computers anymore, because they make too many mistakes when they do. I don't think this is the main issue. For sure, it may not be pretty and smooth, but they should be able to dig up their old skills and ultimately manage to fly without coming too close to the edge.

The real issue is more subtle, and has more to do with the transition phase during which control is switched over from automatics to human. This requires 2 steps ;
1) gather the available information, analyze it and identify the need for a manual take-over
2) properly execute the take-over transition
In the case of an automation malfunction, which might occur suddenly, there is also often a degree of surprise, or "startle", involved in the process.
In most cases, the identification is straightforward - there is an alarm that basically yells "take over !". But it happens that crews just shrug and try to reengage the automatics anyway. And sometimes it is just not simple - AF447 is a prime example of a potentially confusing alarm. However, the PF did identify the need to take over ; he stumbled in part 2, where he did not apply the proper procedure(s). And PNF did not correct him.

The problem is not so much automation, nor is it really a lack of "execution" or "action" skills. It is rather a lack of "identification" skills. This has more to do with ergonomy, psychology, sociology, cognitive processes, and properly packaging all that in selection and training courses, than it has to do with any hand-flying of C152s. And as you can guess, it's a tad more complicated to do, too.
Don't get me wrong, hand flying skills are good to have and maintain. I just don't think this is where we should throw ressources if we want to increase safety.

Quoting sk909 (Reply 7):
But wasn't it his knowledge, skill and confidence that gave him the ability to do this? He was a skilled instructor, knew the systems of the aircraft, how it worked and could use it. Had it been another pilot, the out come could have been very different

More critically, he remained calm, assessed his current situation, assessed his options, communicated as needed. It wasn't all perfect either, but he managed to figure out where he was, where he wanted to go, and how to do that. The resulting control inputs ("stick and rudder" skills) were executed properly, with assistance from the control laws, but that's not the most important aspect of the event for me.

Quoting sk909 (Reply 7):
Why not fix the problem before it becomes a problem? This is what we every time, fix the issue when 100, 200 or 500 people die.

I hope you do realize that this is what is done by the industry in most cases...it's just that such changes don't make the news if there hasn't been some dramtic event first.

Quoting 747megatop (Reply 8):
If the job description of a pilot has really changed without our knowing and is relegated to that of button pushers as your post suggests

If it has "changed", I wonder what the job description was "in the good old days" ?  
Quoting 747megatop (Reply 8):
without manual backup as in the case of Airbus planes

Not following here - Airbus control laws have back-up laws which simply require the pilot to take over manually some functions that the automation can no longer accomplish. So IOW it has "manual back up".
My goal as an engineer is to fill my soul with coffee and become immortal
 
User avatar
KarelXWB
Posts: 26968
Joined: Sun Jul 15, 2012 6:13 pm

RE: Automation Catch 22

Tue Jul 23, 2013 9:11 pm

Quoting 747megatop (Reply 8):
OR are the experts raising questions about the level of automation without manual backup as in the case of Airbus planes versus Boeing's philosophy of providing automation but leaving a manual backup avenue for the pilot to take full manual control and hand fly the plane.

Not sure what you are trying to say here, but Airbus planes do have multiple backups, including a mechanical one.

http://www.airbusdriver.net/airbus_fltlaws.htm
What we leave behind is not as important as how we've lived.
 
User avatar
KarelXWB
Posts: 26968
Joined: Sun Jul 15, 2012 6:13 pm

RE: Automation Catch 22

Tue Jul 23, 2013 9:45 pm

Quoting sk909 (Thread starter):
Is the answer to this problem more computer-control?

But is there a problem? There are way, way less crashes than in the old days, partly thanks to automation. I think plane crashes are already down to a minimum. We will never get to 0 because 1) there is gravity and 2) people make mistakes. Even with more training, people will keep making mistakes because that is a human thing to do.

Here, find the 'air crashes' circle:
http://infobeautiful3.s3.amazonaws.c...h_wellcome_collection_fullsize.png
What we leave behind is not as important as how we've lived.
 
hivue
Posts: 2076
Joined: Tue Feb 26, 2013 2:26 am

RE: Automation Catch 22

Tue Jul 23, 2013 9:56 pm

Quoting sk909 (Thread starter):
As for the recent OZ 214 accident in SFO, again the pilots weren't skilled enough to make the decision to disconnect the auto-pilot

The OZ214 crew disconnected the AP at about 1500 ft.

Quoting UALWN (Reply 5):
In the context of the discussion in the thread it's worth mentioning that US 1549 remained in normal law until the end, with the automation helping Capt. Sully achieve his goal of a clean ditching. Actually, the alpha protection kicked in at an altitude of 150 ft and prevented a stall that could have resulted in a more violent impact with the water.

  
Read the BEA's appendix to the NTSB report. They almost beg the NTSB to give due credit to the airplane for helping save the day.
"You're sitting. In a chair. In the SKY!!" ~ Louis C.K.
 
hivue
Posts: 2076
Joined: Tue Feb 26, 2013 2:26 am

RE: Automation Catch 22

Tue Jul 23, 2013 10:09 pm

Quoting 0NEWAIR0 (Reply 3):
No international regulations need to be made because 4 pilots, for instance, didn't notice that the autothrottles weren't engaged on a 777.

There were 3 pilots on the flight deck in OZ214. Current theories about the accident suggest that it wasn't a case of the A/Th not being engaged but rather the crew not understanding how the A/Th worked.
"You're sitting. In a chair. In the SKY!!" ~ Louis C.K.
 
0newair0
Posts: 417
Joined: Fri Jun 01, 2007 12:21 am

RE: Automation Catch 22

Tue Jul 23, 2013 10:23 pm

Quoting hivue (Reply 15):

I don't believe that theory because at least 2 of those 3 pilots had several hundred and several thousand hours in type at the time. I highly doubt they didn't understand how the auto throttles worked. Pilots will say the darndest things when they're trying to spread "pilot error" blame around.
That's not how this works! That's not how any of this works!
 
futureualpilot
Posts: 2406
Joined: Thu May 25, 2000 10:52 am

RE: Automation Catch 22

Tue Jul 23, 2013 10:33 pm

This is an outstanding video on automation usage. It is obviously from several years ago but the lessons ring true today. Automation can be great but you must know not only how to use it, but when. I see this being a challenge for some who do not have that solid foundation that is basic aircraft control. Don't drop the airplane to fly something else. Always, always fly the airplane first.


http://www.youtube.com/watch?v=h3kREPMzMLk


This is a good article on the pitfalls of automation and how it relates to human factors.

http://www.washingtonpost.com/politi...-11e2-bb32-725c8351a69e_story.html


I find both relevant to the discussion at hand.
Life is better when you surf.
 
747megatop
Posts: 1785
Joined: Wed May 23, 2007 8:22 am

RE: Automation Catch 22

Wed Jul 24, 2013 4:42 am

Quoting KarelXWB (Reply 12):
Not sure what you are trying to say here, but Airbus planes do have multiple backups, including a mechanical one.

I could have sworn that I have read somewhere that Boeing allows the pilot to fully override onboard computer and take control but Airbus does not; but I could be wrong. If I am wrong then my bad, pls ignore my comment.
 
User avatar
zeke
Posts: 15098
Joined: Thu Dec 14, 2006 1:42 pm

RE: Automation Catch 22

Wed Jul 24, 2013 6:04 am

Quoting KarelXWB (Reply 12):
Not sure what you are trying to say here, but Airbus planes do have multiple backups, including a mechanical one.

The mechanical backup is just a mechanical control loop backup, without hydraulics, you have no control unless you are in a newer aircraft where they have electro-hydro actuators (eg A380/787/A350), i.e. a control surface actuator with its own inbuilt hydraulic system.

Quoting 747megatop (Reply 18):
I could have sworn that I have read somewhere that Boeing allows the pilot to fully override onboard computer and take control but Airbus does not; but I could be wrong. If I am wrong then my bad, pls ignore my comment.

That is what the marketing blurrb says, in reality you need to be able to have the presence of mind mid way through a situation where you think you need to override the computer, then put your attention up to the overhead panel, find the switch for the FBW, and turn it off. Never been used in the real world as far as I know. Overriding the computers in any cases means exceeding the design parameters, which may result in a bigger problem.
Human rights lawyers are "ambulance chasers of the very worst kind.'" - Sky News
 
sk909
Topic Author
Posts: 244
Joined: Mon Nov 21, 2005 7:38 pm

RE: Automation Catch 22

Wed Jul 24, 2013 7:24 am

Quoting airmagnac (Reply 11):
No, automation has taken over the repetitive, high frequency, short term tasks, and freed the human pilots to do what humans do best : anticipate, plan, prepare the long term actions

Yes, true that automation has taken over the repetitive workload, but it has dulled the pilots. It has also reduced the pilots to observers, controlling the computer, a task humans are very bad at. Essentially I don't think that automation is a bad thing, but we need to get the pilot into the flying loop. Since FAA, NASA and other aviation organizations have all raised their concerns about this matter, I don't think we can neglect the problem.

Quoting airmagnac (Reply 11):
The ONLY difference between the two

Great observation. And sounds very true. And if true, than same same but different

Quoting airmagnac (Reply 11):
The limits are physical, they apply to any airplane just the same.

The limits are not physical. They are setup to basically not allow the pilot to do dumb things, and to over stress the plane. The planes can go beyond the limits.

Quoting airmagnac (Reply 11):
Of course they do, just as the manufacturer of any tool

Sure that is the assumption. The problem, is that airmanship skills are declining, raising the question; do pilots have the necessary skills the planes are build by? Take your hammer, a low tech hammer is easy to handle, but a high tech might not, hence more skill, knowledge and education is needed.

Quoting airmagnac (Reply 11):
We know next to nothing about the Asiana accident, but you can already state "overall" conclusions ?

No I cannot state the conclusion. But the fact is that asian pilots, especially Korean pilots, are avoiding manual flying. The plane had no engine malfunction, hence they flew a perfectly working plane into the ground. It flew to slow, which indicates the pilots weren't watching their airspeed. Is this an effect of pilots trusting the computer more and more, forgetting the basic airmanship; airspeed is number 1?

Quoting airmagnac (Reply 11):
unable to apply the proper actions (learned as basic airmanship)

Basic airmanship training tells, if you don't know what is going on or in this case what the computer is doing, disable the A/P and fly the plane. Fly the plane! Why don't they fly the plane? That is a skill, a skill pilots are forgetting. For what ever reason.

Quoting airmagnac (Reply 11):
Making general conclusions based on such a small sample is not a good idea...

Well it is not based on only those accident, but on incidents, reports, articles, interviews with pilots. The accidents was what triggered me, to research and asking the question.

Quoting airmagnac (Reply 11):
You believe the problem is that pilots cannot replace the computers anymore, because they make too many mistakes when they do. I don't think this is the main issue. For sure, it may not be pretty and smooth, but they should be able to dig up their old skills and ultimately manage to fly without coming too close to the edge.

The problem is not whether a pilot would be able to dig out the old skills. Because sure they have a set of skills, but if you don't use these skills, you get less confident. Just like riding a bike. Riding a bike is not difficult, but applying advanced skills is. The more skilled a biker the better reaction you have, hence better at avoiding accidents. The same applies to pilots. Use your skills, the more confident you are.

Quoting airmagnac (Reply 11):
It is rather a lack of "identification" skills.

The problem seems to be the lack of pilot/machine interface. The computer can do all the calculations, corrections, flying, but if it doesn't tell the pilot what is going on, the pilot has a hard time. And precious time is wasted. Take the AF 447 accident, the computer kept spewing messages, that didn't make sense. Had it come up and said, pitot tube iced, the pilot could have taken the right actions. That is essentially the major problem. Which again point at the pilot; fly the plane.

Quoting airmagnac (Reply 11):
I hope you do realize that this is what is done by the industry in most cases...it's just that such changes don't make the news if there hasn't been some dramtic event first.

I do know. Just as in any other industry. No accidents, no publicity.

Quoting hivue (Reply 15):
There were 3 pilots on the flight deck in OZ214. Current theories about the accident suggest that it wasn't a case of the A/Th not being engaged but rather the crew not understanding how the A/Th worked.

Fly the plane. Basic airmanship tells you: watch the airspeed. Airspeed is number 1. But lets wait for the final report. But no matter how it turns out, it points at a lack of airmanship.
Life's for Living!
 
sk909
Topic Author
Posts: 244
Joined: Mon Nov 21, 2005 7:38 pm

RE: Automation Catch 22

Wed Jul 24, 2013 7:26 am

Quoting 747megatop (Reply 18):
I could have sworn that I have read somewhere that Boeing allows the pilot to fully override onboard computer and take control but Airbus does not; but I could be wrong. If I am wrong then my bad, pls ignore my comment.

Moi aussi, but as Airmagnac told us, it ain't so. Or so he states.
Life's for Living!
 
sk909
Topic Author
Posts: 244
Joined: Mon Nov 21, 2005 7:38 pm

RE: Automation Catch 22

Wed Jul 24, 2013 7:35 am

Quoting cuban8 (Reply 10):
The main problem is in my opinion the man/machine interface and man/aircraft knowledge which people commonly express as situational awareness.

Very true. But it comes down to who decides, man or machine? If man can't comprehend what the machines is doing, should man take over? I say yes. In the case of AF 447, man did not get what the machine was telling. Although the F/O did disconnect the A/P, he did not handle the situation correctly. He was overwhelmed by information from the computer.
Sure the right thing is to enhance the man/machine interface, but it comes down to man taking control. Especially in a situation that is fast turning bad.
Life's for Living!

Popular Searches On Airliners.net

Top Photos of Last:   24 Hours  •  48 Hours  •  7 Days  •  30 Days  •  180 Days  •  365 Days  •  All Time

Military Aircraft Every type from fighters to helicopters from air forces around the globe

Classic Airliners Props and jets from the good old days

Flight Decks Views from inside the cockpit

Aircraft Cabins Passenger cabin shots showing seat arrangements as well as cargo aircraft interior

Cargo Aircraft Pictures of great freighter aircraft

Government Aircraft Aircraft flying government officials

Helicopters Our large helicopter section. Both military and civil versions

Blimps / Airships Everything from the Goodyear blimp to the Zeppelin

Night Photos Beautiful shots taken while the sun is below the horizon

Accidents Accident, incident and crash related photos

Air to Air Photos taken by airborne photographers of airborne aircraft

Special Paint Schemes Aircraft painted in beautiful and original liveries

Airport Overviews Airport overviews from the air or ground

Tails and Winglets Tail and Winglet closeups with beautiful airline logos