, , , , ,

I am an old United States Air Force pilot. One of my totally unintentional mistakes while in the Air Force pilot training was scoring perfectly on an extremely difficult academic examination. The problem was that the official bell curve was pegged to the highest score and the lowest one, which put the failing grade so high that half of my fellow pilots failed. We were given lots of tests and if a guy failed two times, including the retest of the failed one, he was out of flight school. It was unfair, but because of my super high score quite a few student pilots, about 50, were instantly under extreme pressure of being eliminated. So they hated me. I mention this sad story because what I am about to recommend will sound stupid to some people, including a lot of current airline pilots.

Today the final report on the 2009 Air France crash was reported in CNN —  “The Brazil-to-France flight plunged into the Atlantic Ocean on June 1, 2009, killing all 228 people on board.” Their headlineFinal Air France crash report says pilots failed to react swiftly – is totally wrong. The pilots reacted instantly, but they did exactly the wrong thing in the circumstances. The real problem is why a pilot with 800 flying hours in that type of airplane did exactly the wrong thing and continue doing it for almost four minutes. It was behavior which he could have changed in one tenth of a second that would have totally stabilized the problem in less than five seconds. The airplane was flying straight and level at cruising altitude and would have continued to do so if the pilot hadn’t pulled the control stick to the extreme back position. He held it there until the plane crashed. Every airplane flying will crash within four minutes if you hold the control stick in that position. So the question becomes, why did a normally intelligent person with plenty of flying experience not recognize a stall, which was caused by his holding the stick full back, and do the right thing, which was simply to push the stick forward. In fact if he had let loose of the stick the airplane would probably have righted itself and returned to approximately level flight because nearly all airplanes are designed to be dynamically stable.

The reason he crashed a perfectly flyable airplane was because he didn’t have any what we used to call “seat of the pants” flying experience. That experience is very easy to acquire, and in terms of time and monetary expense almost free. The solution to the problem is to have all airline pilots fly a light single engine, perhaps even fabric, airplane through a primary flight training program. Since these people are already experienced pilots this training doesn’t have to involve a great number of flying hours. After an hour of dual instruction in stalls, spins and flying in an airplane without any advanced instruments these pilots would fly only an hour per year in these types of planes. But it wouldn’t be straight and level flying, instead it would be doing out of control types of maneuvers. These are easy to cope with if you know how, but as has been proven in recent times modern high performance pilots are not experienced in these types of low speed maneuvers. They may have been exposed to them in flight simulators but that doesn’t give a pilot the feel of moving through the air and controlling the airplane in three dimensions while encountering unexpected turbulent air. The reason a jet pilot doesn’t know how to cope with low speed maneuvers is because he never flies slow, and in the rare case that he does he should know instinctively what to do. The problem is that what to do is exactly the opposite of what the Air France pilots did.

If I was an ultimate administrator, which I most certainly am not, I would require all airline pilots to have two separate training hours flying spins, stalls and other unusual maneuvers every year. That simple requirement would put an end to stupid mistakes like the one which killed all 228 people on that Air France flight, and many more. Also, it would be lots of fun for the pilots, and it is easy to do.

The recovery from stalls and spins and instrument failure should be a reflex for every pilot, even an airliner pilot.

Some alternate possibilities. — There are some technical fixes which would be relatively low tech but which might prevent or at least help to cope with desperate situations. A pitot-static tube is essentially a hole on a tube pointed directly into the wind and a second hole facing to the side of the wind which change pressure relative to each other as the airplane moves through the wind. These two holes permit measuring airspeed, altitude and vertical speed, but unfortunately the tubes can occasionally become clogged with debris such as birds or ice. This is typically prevented by having an electric heating element built into the pitot tube. If the heater had been turned on in the Air France flight it is unlikely there would have been an ice buildup and the automatic pilot would have worked perfectly.

An alternate method would be to place a flexible surface on a forward facing portion of the airplane. The impact pressure of the wind would be measured and several of these sensors could be arranged on the nose and measure the relative direction of the wind as well as its relative speed. The advantage of this method might be a natural resistance to icing and an alternate form of measurement which could be used for backup reliability.

Another alternative would be to design a panic button for the pilot to hit when he had run out of options. Clearly the Air France pilot needed help and this might have been the answer. When the panic button was hit the airplane would respond with a very low tech solution — for a short time a stabilizing system would take control of the airplane and bring it to a horizontal straight and level condition, with five degrees nose up, at a full power setting; it would maintain this condition for ten seconds unless the pilot turned it off.

This device would be based on a very reliable gyroscope and would be independent of all other factors. It would bring the airplane into a controlled condition and set it into a slight climb. After stability had been achieved it would accept other simple commands such as direction, cruise power settings. None of this requires the pitot-static assembly to operate. Over the course of this minute the radar altimeter, the previous known altitude, and air pressure altimeter would be checked, verified and compared. Ground based radar could sometimes also be used at this time. This is flying at its dumbest but also at its most reliable, at least in the very short term. While the airplane is in this frozen condition the pilots could figure out what change from this controlled locked-in condition was needed.