Latest research on Automation Dependency - Regulator please note..
 

Pilot Reliance on Automation Erodes Skills



By ANDY PASZTOR



MILAN (WSJ)-Increasing reliance on cockpit automation appears to be significantly eroding the manual flying skills of many airline pilots, who are then "sometimes not prepared to deal with non-routine situations," according to the researcher behind a sweeping air-safety study released Thursday.

Presented to an international aviation safety conference here by senior Federal Aviation Administration scientist Kathy Abbott, the study's conclusions buttress the idea that a significant percentage of airline pilots rely excessively on computerized cockpit aids. Such adherence to computer-assisted pilotingg-and the confusion that can result when pilots fail to properly keep up with computer changes-increasingly are considered major factors in airliner crashes world-wide.

Prepared by a team of industry, labor, university and government experts, the findings reveal fundamental safety gaps in commercial aviation. And they are likely to prompt stepped-up debate over whether pilot training should be changed to reverse that trend.

Kathy Abbott, the agency' s chief technical advisor for flight deck design and human factors, said the participants analyzed more than 730 incidents, 26 accidents and some 9,100 flights word-wide between 2001 and 2009.

The study found, among other things, that manual flying errors contributed to about 60% of selected accidents and 30% of selected incidents perused by the team. Those statistics "got our interest, as you can imagine," Ms. Abbott said. The errors included inappropriate control inputs by pilots and incorrect responses when trying to recover from aircraft upsets.

Thursday's report summary was long awaited by regulators and airline officials, because it is a way to understand and highlight the hazards of excessive pilot dependence on automation. The group of experts, including aircraft manufacturing executives and pilot-union representatives, isn't expected to release the final document until next year.

But already, the preliminary conclusions are pinpointing problems some pilots have in properly recognizing when autopilots should be engaged or disconnected in certain types of emergencies. Focusing too much on manipulating flight-control computers, according to Ms. Abbott, often "distracts from managing the flight path of the airplane."

The study is intended to update an influential 1996 FAA-sponsored examination of the benefits and drawbacks of cockpit automation. But because automated flight-management systems, navigation aids and autopilots have progressed so dramatically since the 1990's, the latest study is widely expected to set a benchmark. Other groups and organizations are looking to conduct follow-up research based on its findings.

One cross-cutting theme spelled out in the report, according to Ms. Abbott, is that "pilots sometimes abdicate too much responsibility to the automated systems." Part of the reason, she said, are persistent messages pilots receive from airline management and trainers stressing that "automated systems can do the job better than they can." The study also found that in some cases, pilots don't get adequate opportunities to practice hand-flying skills and therefore often don't feel comfortable grabbing control away from sophisticated flight-deck computers in an emergency.

How about scrapping the Multi Pilot License?

Is the info taken from all of the above 730, or "selected" accidents to prove the point.

Apologies for being sceptical.

about time someone made it official.

time to spend the money on people not on boxes

PTH

Howdy. Point well taken, especially when the savings realized are planned and safety computed. In other words, "Let's dumb down the entry level pilot, and skimp on recurrent?" How much of this "New" report was lifted from BOAC's paper?

bear

The problem with automation is that it just keeps getting more complex until it is just too intricate to manage and impossible to make fool proof.
Also just because a technology exists, engineers feel the need to use or incorporate it regardless of the need.
Finally, computers are just stupid and only as good as their designers. Not everything can be foreseen and it is exceptionally difficult to make these systems adaptive.

So in summary, piloting skill will always be paramount, and telling anyone to rely on a computer ahead of their own judgement is a recipe for disaster.

By "piloting", do you mean manipulation, Jabiman - i.e., hand flying? If so, I think it's gone way beyond redeemable in some airlines, (certainly the one that paid my salary for 20 years), where for years now, they've operated under an official policy of maximum use of automation at all times on the line, to the point where you're not even allowed to practise doing a localiser approach if there's a glideslope available.

I'm sorry to say it, but the beancounters have long ago worked out that maintaining a training regime for a large pilot workforce that keeps all those pilots' manipulative skills at an acceptable standard is more expensive than a possible hull loss every 'n' years - and to date, most (stress 'most') have been very, very lucky that when a really serious situation occurs, they've just happened to have a captain at the controls who was once required to do it the old fashioned way and has retained enough of the old skills to get out of it, or a younger captain who has somehow managed to develop those skills despite his company rules.

Unfortunately, as the years progress, fewer and fewer people will those old skills to fall back on are out there - but the beancounters don't care. They (or their PR Departments) will tell the public that safety is their paramount concern, but behind closed doors, they'll 'prove' to you that, given the minuscule percentages involved where multiple failures or a failure and outside conditions will gang up on a crew to leave them without the automatics, it's simply not worth the money to maintain a training program that insists on pilots maintaining manipulative skills beyond a cursory hand flown approach squeezed into an already crowded sim session maybe once or twice a year.

right bearfoil.

point 1. I spoke with a professor at MIT in 1990. He occupied the Boeing Chair. He said the automation would enable pilots with 250 hours to safely operate (note, not pilot) a 747 sized plane.

I spoke to the defense of skilled, experienced pilots. He said they wouldn't be neccessary. I reminded him of failure modes and he ''pooh poohed it".

I know I could do more with a 30 year old DC9 flying the north east corridor shuttles than someone with a brand new airbust 320... I could squeeze in with a ''crowbar'' approach, or be ready for a reroute without busting a fingernail.

I know I am right. You have to have the skills and experience....sure automation can help...I love it when it does the paperwork for proper pay and ''duty rigs" etc.

I also remind people about the film/novel: "Failsafe". Or, "collosus, the forbin project". People can't keep up with machines and if they fail, you can get in real trouble. Machines can help us with information, but I want a HUMAN BEING on the button, or yoke to still have the option to NOT PUSH IT.

Good pilots, beautiful flight attendants, moon landings, great sitcom theme songs...all things of the past!

PTH

While I'm at it, I may as well demand Pure jet, no Propellors. Colgan should be enough to get the ball rolling, Eh?

(And I'm starting to look at giant Fans as Ducted "Propellor" propelled a/c).

bear

bearfoil

40 years ago we were on the verge of the Boeing 2707. Gadgets weren't part of it so much as power and swing swept wings.

we aren't any faster now, but more gadgets.

I think we took a wrong turn at albuquerque.

you in wisconsin bearfoil????? I love it there.

pth

More like 55 years ago, and a sweet a/c. Many Legacies furloughed their senior Captains when taking delivery of the 707, thinking they were washed up, and Propellor heads. If they still had them on board by 1970, they were re-winged, (the Captains), and typed on Fatty. Their Gray Hair inspired "Confidence" in the Pax, as they climbed, (or walked) on board this "HUGE version of the 707". (:ugh:) Money for danger, the name of the game imo.

Nah, California, Petaluma. Wisconsin? c-c-c-cold, for a Calikid, not that there's anything wrong with Wisconsin. It's habitable....about five months out of the year. Mosquitoes the size of sparrows, but fireflies, so it's a wash.

bearfoil...I meant the Boeing 2707 that never flew...it was Boeing's SST and named the TWO seven oh seven...swing wings.. 300 pax...2000mph. and where did that go?????nowhere...but now we have planes that go slow with lots of gee whiz buttons

Pet-A-Lama...I know that place...near Gnoss field?

yep, Gnoss. Spaced on the SST, my bad. How "soon" we forget. Like politicians, sometimes a/c designers and salesmen love the attraction of "complicated". If "fly by wire" is supplanted by "fly by signal", well.......bound to happen. "Wait, microwaves to spark controls", "there's the ticket".....sheesh.

take care, bear

Wonder what happened to the screaming crowd that keeps insisting that manual flying belongs in the sim.....Funny how the industry recognizes that pilot skills are deteriorating due a lack of flying manual.

protectthehornet, I concur with you sir! I fly old 727's and can do a whole lot more and faster with her then with a fully automated fluff.

But sure, don't fly manual, that belongs in the sim, right?? :E :E

Another problem is that on some aircraft there is no absolute reversion to manual. There is still automation playing around at the edges. Added to all this there is a known lack of unusual attitude recovery training.

My thoughts
 

I have recently done training for pilots from turboprop upgrading to jet in the sim. When confronted with a sitiuation where they were getting high on approach, their immediate reaction was to disconnect the A/P and push the nose down. Who cares about flaps right? There was no need to disconect at the time as they weren't that high, and well within the abilities of the automation. Having automation helps a lot especially for the PM, who has to monitor the flying and do the R/T. Anyhow pushing the nose down was couterproductive for that particular sitiuation leading to the fact that they were not sure how to use automation to its fullest and also had less than the required basic flying skills.

On the other hand another gentleman doing an upgrade from the widebody,when confronted with a sitiuation where the A/P and A/T suddenly "failed", was all over the place because according to him, he wasn't prepared. Obviously not a Scout in his past life.

I advocate the use of as much automation as possible and the thorough understanding of its abilities and limitations, but the very least you need as a PILOT are your BASIC FLYING skills and I train that way.

You really need to explain that philosophy to me. What is the difference between flying manual versus flying on automation when it comes to the PM "monitoring"???The only thing the PM will have to do extra, is make selections on the MCP as requested by the PF, the "monitoring" doesn't change one bit, nor do the rest of his responsibilities....

trent892b . . .
 

Unless I'm faced with CAT-II or CAT-IIIb arrival, I disconnect whether I'm high or low, or on profile! Why? ...Because after 12hrs in cruise I want to do something, I want to hand fly...and feel like a pilot. Get it? :ooh:

Just an ancient here....

Isn't "flying" about always being 'ahead' of the airplane?
Isn't the effect of some of todays automation, that it not only allows you to take your hands off the controls, but also that you end up about fifty miles 'behind' the aircraft, and when some does go not according to plan, you have an awful lot of 'catching-up' to do?

"Now why did it do that?" more often than not does not seem to be a joke any more...

CJ

If it is then it shouldn't be, because that's not what it was designed to do. Automation is there for one thing and one thing only - and that is to assist the pilot by taking care of some operational tasks, usually menial and repetitive, that can be handled by a machine. The idea that pilots can take their focus from the safe handling of the aircraft by allowing the automation to run ahead of them is a fallacy.

The problem we perenially run into with this subject is that over the years it has become hard to separate the safety aspect of integrating automation into the pilot's workflow from the political aspect ("Automation is responsible for de-skilling our jobs", "The beancounters demand automation at all times", "The engineers and beancounters are trying to make us obsolete" and "I want to feel like a pilot damnit, not a computer operator!"). The former is of paramount importance in aviation, the latter is at best unhelpful and at worst a dangerous distraction.