Heard of Bernard Ziegler, the man who married fly-by-wire technology to the modern Airbus fleet? Probably not, but according to William Langewiesche in his new book Fly By Wire: The Geese, The Glide, The Miracle on the Hudson (Farrar, Straus and Giroux, $24), Ziegler was the invisible hand behind Captain “Sully” Sullenberger’s handling of US Airways Flight 1549 and his life-saving landing on the Hudson last January.
“Like it or not,” Langewieshe writes, “Ziegler reached out across the years and cradled them all the way to the water.”
For a writer of Langewiesche’s precision, that’s a surprising lapse into purple prose.
To explain, Bernard Ziegler, whom Langewiesche interviewed in the course of an extensive lunch near Toulouse, was the prime mover of the fly-by-wire technology that Airbus uses on every one of its modern airliners, including the A320 flown by Sully. And Langewiesche, a pilot himself, believes that it was the robot grip on the controls of Flight 1549, as much as Sully’s, that enabled the successful-and rare-ditching.
Indeed, the final seconds of the flight as described in the book are
white-knuckle reading. Sully’s cool skill and the computers’ authority
come together: “… for the last few seconds of the glide, with
Sullenberger’s stick fully back, the computers intervened and gently
lowered the nose to keep the wings flying.”
However, behind this truly enthralling marriage of man and machine, lies an arcane but critical issue of piloting technique and aviation safety.
In the timing of his book, Langeweische has been caught between two crashes, of Flight 1549 and the disappearance over the Atlantic, in June, of Air France Flight 447. And the lessons of each appear to be contradictory.
Airline pilots familiar with both Airbus and Boeing planes will tell you that they call for a different philosophy in the cockpit. To put it simply-too simply, perhaps-the Airbus computers “protect” a pilot from putting the plane into a situation from which it won’t be able to recover, while Boeing leaves more for the pilot to handle himself.
Lest this begins to sound scary, let me be clear: This difference should normally be of no account in the safety of air travel. Boeing and Airbus design and produce airplanes that are as accident-averse as they can be. Cockpit automation has hugely reduced certain categories of crash, particularly what was called CFIT, controlled flight into terrain.
I know pilots who prefer one system to the other. More traditionally-minded pilots tend to think that Airbus leaves them with too little to do. (They might, to cite a recent episode, believe that the computers fly the plane so well that they can be too easily distracted and miss their appointed airport. I can attest to this myself, having flown an A320 in an Airbus simulator and “landed” it at JFK without mishap, and I am not a pilot. )
But Langeweische is right, I believe, to show how a severely disabled Airbus with barely any power left to work its controls was able to more finely tune its impact on the Hudson than any human could have managed.
The trouble comes when the computers get fed dud data and don’t know that it is dud. It will be some while, if ever, before we fully understand what sent Air France Flight 447 to the deepest bottom of the Atlantic. There is little doubt, however, that the prime suspect is the very same computer-dictated flight management system that served Sully so well.
The Air France plane was an Airbus A330. This plane has a recent record of problems with the gauges that measure airspeed, called pitot tubes. The flight control computers have only one way of knowing how fast the Airbus is flying. That data comes from the pitot tubes.
Flight 447 was at the cruise phase of its route from Rio to Paris, passing through an area of violent storms. The computers were in charge, through the autopilot. The computers believe what they are told about speed, even if it is wrong. We know from 24 bursts of data received via satellite from Flight 447 that its flight management computers were shutting down, like a brain that seizes up when traumatized.
In turbulent weather flying at the right speed is especially critical, leaving little margin for error. The pilots were unable to regain control from the computers before the Airbus fell to the water and broke up.
Langeweische did have time to note this disaster in his book: “Whatever happened was extremely rare” he writes. “Was this the first catastrophic collapse of a civilian fly-by-wire system? It may prove to have been.”
He is exactly right-it is a “may prove to have been.” He is also right to insist that more crashes have been caused by pilots than by computers. Nonetheless, the story of the incredible achievement we call commercial aviation is one of a relentless and incremental drive to reduce accidents to zero. In the course of this, stuff happens. Some stuff defeats the best efforts of man and his works to foresee. And I believe that Air France Flight 447, with its secrets still to be unraveled, will probably turn out in the end to be another costly but consequential further step in reaching for zero.
However, behind this truly enthralling marriage of man and machine, lies an arcane but critical issue of piloting technique and aviation safety.
In the timing of his book, Langeweische has been caught between two crashes, of Flight 1549 and the disappearance over the Atlantic, in June, of Air France Flight 447. And the lessons of each appear to be contradictory.
Airline pilots familiar with both Airbus and Boeing planes will tell you that they call for a different philosophy in the cockpit. To put it simply-too simply, perhaps-the Airbus computers “protect” a pilot from putting the plane into a situation from which it won’t be able to recover, while Boeing leaves more for the pilot to handle himself.
Lest this begins to sound scary, let me be clear: This difference should normally be of no account in the safety of air travel. Boeing and Airbus design and produce airplanes that are as accident-averse as they can be. Cockpit automation has hugely reduced certain categories of crash, particularly what was called CFIT, controlled flight into terrain.
I know pilots who prefer one system to the other. More traditionally-minded pilots tend to think that Airbus leaves them with too little to do. (They might, to cite a recent episode, believe that the computers fly the plane so well that they can be too easily distracted and miss their appointed airport. I can attest to this myself, having flown an A320 in an Airbus simulator and “landed” it at JFK without mishap, and I am not a pilot. )
But Langeweische is right, I believe, to show how a severely disabled Airbus with barely any power left to work its controls was able to more finely tune its impact on the Hudson than any human could have managed.
The trouble comes when the computers get fed dud data and don’t know that it is dud. It will be some while, if ever, before we fully understand what sent Air France Flight 447 to the deepest bottom of the Atlantic. There is little doubt, however, that the prime suspect is the very same computer-dictated flight management system that served Sully so well.
The Air France plane was an Airbus A330. This plane has a recent record of problems with the gauges that measure airspeed, called pitot tubes. The flight control computers have only one way of knowing how fast the Airbus is flying. That data comes from the pitot tubes.
Flight 447 was at the cruise phase of its route from Rio to Paris, passing through an area of violent storms. The computers were in charge, through the autopilot. The computers believe what they are told about speed, even if it is wrong. We know from 24 bursts of data received via satellite from Flight 447 that its flight management computers were shutting down, like a brain that seizes up when traumatized.
In turbulent weather flying at the right speed is especially critical, leaving little margin for error. The pilots were unable to regain control from the computers before the Airbus fell to the water and broke up.
Langeweische did have time to note this disaster in his book: “Whatever happened was extremely rare” he writes. “Was this the first catastrophic collapse of a civilian fly-by-wire system? It may prove to have been.”
He is exactly right-it is a “may prove to have been.” He is also right to insist that more crashes have been caused by pilots than by computers. Nonetheless, the story of the incredible achievement we call commercial aviation is one of a relentless and incremental drive to reduce accidents to zero. In the course of this, stuff happens. Some stuff defeats the best efforts of man and his works to foresee. And I believe that Air France Flight 447, with its secrets still to be unraveled, will probably turn out in the end to be another costly but consequential further step in reaching for zero.
Leave a comment