Posted: 11/5/2012 12:23:21 PM EDT
|
It all comes down to training and experience.
Once a pilot - now a computer's sidekick Following the Air France Airbus A330 accident in the South Atlantic on 1 June 2009, the French Authority BEA made several recommendations to prevent future accidents of this type. Aviation Week magazine reports that a human factors working group has been set up as part of the BEA investigation. Better training is likely to become a crucial follow-up action as the report is absorbed and discussed across the industry. Talk-fests on the subject are a dime a dozen and will get nowhere until operators realise that the solution is straight forward. And that is the accent on flight simulator training must be reversed from automation back to basic principles of flying an aeroplane on raw data manual flying. Raw data includes flight director turned off; despite what some airline instructors think. Currently 90 percent of simulator training involves automation. In spite of this it has been known for many years that automation addiction forced upon crews by manufacturers and operators leads to complacency and loss of basic instrument flying skills. Besides the spectacular demise of the Air France A330 there have been numerous other jet transport accidents where the cause has been poor instrument flying ability caused in part by over-reliance on automation. Below is an article extract cut and pasted from the latest issue of Curt Lewis & Associates Flight safety website No 147. It should be absorbed by every automation dependant airline pilot. The author makes his point very clearly that operators must grasp the nettle and fix the problem before the inevitable next accident caused by the pilot's lack of basic instrument flying ability. The article is headed Once a pilot, now a computer's sidekick Industry analysts estimated last week that in the next 20 years, the airlines are going to need 466,000 new pilots. When I said to an airline pilot friend that such a job market would make it easy for his son to follow in his footsteps, he smiled. "I think he wants a flying job instead," he replied. I noted that this sounded odd coming from a fellow who just flew a planeload of passengers back from overseas. "I didn't fly," he replied. "The computer flew. I sat in the front office, monitoring systems." "Who flies better," I asked, "you or the computer?" "Oh, the computer," he replied. "No contest, as long as things function. When they stop functioning, it's a different story. Then the computer quits, and I go to work. Provided I still remember how." The dilemma isn't new, but it's being discussed more and more frequently. Pilots don't fly enough. They get rusty, and when they really need to call upon their flying muscles, they find them either atrophied or insufficiently developed in the first place. The symbol of the problem has become Air France's Flight 447, an Airbus dropped by its pilots into the ocean three years ago, according to a French inquiry's final report released last month. When a faulty speed sensor made the autopilot quit, two co-pilots on the flight deck would have needed to hand-fly their Airbus 330, established in cruise at 35,000 feet over the Atlantic, until the captain, who was taking his scheduled nap, returned to the cockpit. Their task was to fly straight and level for two or three minutes on instruments, with no visual reference to the horizon, without reliable airspeed indication, in light turbulence. They couldn't do it. By the time the captain came back, the Airbus had stopped flying and was about a minute from contacting the water. In 1915, Arthur Roy Brown, the flying ace credited with bringing down Manfred von Richthofen, the Red Baron, had his pilot's license issued with six hours of flight time. By comparison, even the least experienced pilot on AF447 had 2,800 hours in his logbook. It isn't that today's pilots train fewer hours; it's that the study of increasingly complex systems and regulations compete for time and emphasis with flying skills. Airmanship and command authority are being boxed in by petty rules, for the comfort of lawyers and bureaucrats rather than to enhance operational efficiency and flight safety. Before leaving the flight deck for his scheduled rest, the captain of Air France's ill-fated Flight 447 was obliged, as part of his briefing, to ask his relief pilot if he had a commercial pilot's license. Why would anyone weigh down the captain's workload with such a query? Would an unlicensed impostor say to the captain: "Crikey, skipper, I didn't know you needed a license for this gig" or would he just lie and say: "Yes, sir." The crew whose fate it was to be flying Flight 447 had the necessary qualifications. The problem was that they had them in their wallets, rather than in their heads. Qualifications in wallets satisfy bureaucracies, but only qualifications in heads ensure the safety of a flight. It was a "Thales"-type speed sensor that iced up as the Airbus was skirting a thunderstorm high above the South Atlantic. Air France, aware of the limitations of the device, had just begun to replace the $3,500 units. It hadn't gotten around to changing it in the ill-fated airplane before it departed Rio de Janeiro for Paris on the night of June 1, 2009. Grounding the entire Airbus fleet until all units were replaced may have cost only a fraction of what the accident, investigation and lawsuits will end up costing Air France, to say nothing of the tragic loss of 228 lives. Some analysts argue, though, that turning all potential flaws into mandatory "no go" items would make air transportation unaffordable. The "Thales" sensors were more susceptible to icing than other designs, but they didn't all ice up, and the planes carrying those that did remained flyable and were landed safely by their Air France crews. So were two other Airbus 330s belonging to Paris-based Air Caraibes Atlantique. Only Flight 447 fell into the ocean. One disaster is one too many, of course, but it was no more an inevitable consequence than it would be for a blown tire to flip a car. Airspeed is crucial to flight. Too fast and the plane can break up; too slow and it can fall out of the sky. When airspeed indicators become unreliable, the computerized systems - autopilot and auto-throttles - quit. On the Airbus, this is announced by the aural warning of a cavalry charge, the computer's way of calling the human pilot to the rescue. Aviation is full of pithy sayings. One is that an airspeed sensor has no backup except airmanship. Losing airspeed readings can range from a non-event to a dire emergency depending on the pilot's skill and additional circumstances. The autopilot quitting on AF447, as it was designed to do after losing reliable airspeed indication, could and should have been a non-event. It left an airworthy aircraft flying straight and level in light turbulence. All Flight 447 needed was a pilot to fly it - or just let the plane fly itself, which is what planes trimmed for cruise flight tend to do in stable air, especially if their wings are kept level - but, as the cockpit voice recorder revealed, there were no pilots on the flight deck. There were two systems managers being confronted by a system that suddenly had become unmanageable. Real pilots would have disregarded the rebellious computers going viral with flashing lights, cavalry charges, buzzers and bells, huffily announcing all the things they stopped doing for the humans aboard or required the humans to do for them. They would have let the computers crash and concentrated on flying the plane. The systems managers stopped flying and crashed with their computers. This isn't how the French inquiry puts it, needless to say. I wouldn't put it this way in an inquiry myself. I'm exaggerating to make the point that our technology may be getting ahead of itself. If so, we may hire 466,000 systems managers of an unmanageable system in the next 20 years. Once a pilot, now a computer |
|
The automation can be a great thing, provided you run it well, and know when to step down a level or two. We do unreliable airspeed, and a raw data approach in training, and have for years. I still try and hand fly some, but my airplane is not the best hand flown machine, and the other pilots work load goes up quite a bit, when you hand fly.
It will be interesting, as we introduce the A-320 series into our fleet, how the powers that be choose to run the training. Since I'll probably get stuck on the Bus. A good video, for everyone who flies an automated airplane. Children of the magenta |
|
I flew in Alaska and have over 1800 landings on "other than paved" runways before ever touching my first jet...Which happened over the weekend BTW.
For that I am truly thankful because I'll always be able to learn systems. I've already learned to fly in the toughest place on earth. |
| Stick and rudder skills are dying. Now it's autopilot on when the gear hits the wells, GPS direct to, coupled approach. I am not surprised at all when it hits the fan planes depart the skies missing wings and such. Many pilots have forgotten how to fly. It seems most bad wrecks these days are rookie mistakes. |
|
Quoted:
A strong foundation in the basics/fundamentals is a good place to start. Unfortunately the ab-initio programs favored by the euros skips this. It starts early. Look at all the flight schools running G1000 planes so the guys that graduate haven't even seen a steam gauge. |
|
Quoted:
Quoted:
A strong foundation in the basics/fundamentals is a good place to start. Unfortunately the ab-initio programs favored by the euros skips this. It starts early. Look at all the flight schools running G1000 planes so the guys that graduate haven't even seen a steam gauge. I lobbied my the flight school owner to NOT allow primary students to fly in the G1000 for that very reason. |
|
Quoted:
Quoted:
Quoted:
A strong foundation in the basics/fundamentals is a good place to start. Unfortunately the ab-initio programs favored by the euros skips this. It starts early. Look at all the flight schools running G1000 planes so the guys that graduate haven't even seen a steam gauge. I lobbied my the flight school owner to NOT allow primary students to fly in the G1000 for that very reason. Our club runs Warriors with steam gauges and a 430 in them for instrument students. That seems like the best plan to me. Now if they can just pull the ADF so I don't have to do NDB approaches. 
|
|
Quoted:
Quoted:
I don't think that the type of gauge makes the difference. The problem is that what we require a certified pilot to do at PVT/COM/ATP stages of training hasn't changed appreciably from 1930s. What would you change? Requirement for acro. More flight time. Neither is going to happen. GA was hanging on by a thread for the last 10 years. Its dead as fried chicken with the current election results. |
|
Quoted:
Stick and rudder skills are dying. Now it's autopilot on when the gear hits the wells, GPS direct to, coupled approach. I am not surprised at all when it hits the fan planes depart the skies missing wings and such. Many pilots have forgotten how to fly. It seems most bad wrecks these days are rookie mistakes. Very true. In the C-17 RTU we were yelled at for not engaging AT and ATS after 200'. That jet and all of its systems definitely spoiled the crap out of me. 60-1 descent rules, holding patterns, and pitch/power settings were all taken care of by the jet or its mission computer. The HUD was phenomenal as was the auto trim and fly by wire systems. Now that I'm going through T-6 PIT, relearning the stuff I forgot from UPT is definitely a challenge. There is something to be said about stick and rudder flying. It is a skill that is easily forgotten, but every pilot should possess and remain proficient at it. |