3 ms·

Firing pilots because they are going through a divorce, as suggested by the author, seems a little extreme. CFIT (controlled flight into terrain) crashes such

Firing pilots because they are going through a divorce, as suggested by the author, seems a little extreme. CFIT (controlled flight into terrain) crashes such as this - whether accidental or intentional - can be eliminated by technology and hopefully will be soon. Airlines can complain about the cost all they want - the direct loss projection from this single incident is $350 million [1]. The cost of doing nothing is demonstrably greater than the cost of developing and deploying a CFIT prevention system. [1] http://www.bloomberg.com/news/articles/2015-03-27/co-pilot-suicide-leaves-lufthansa-facing-unlimited-liability http://www.bloomberg.com/news/articles/2015-03-27/co-pilot-s...

sokoloff 12y ago
I agree that we can make great strides against unintentional CFIT accidents, but don't think we can rely on tech to eliminate intentional CFITs. Cut the fuel supply to both engines; what's the computer going to do? Now, if you protect against that, what's to prevent that system from stopping a pilot from extinguishing an in-flight fire in a situation where cutting fuel to one engine (or perhaps two/all) is the right response?
- downandout 12y ago
  But you're implying that we can't make the aircraft smart enough to know that there is a fire and that cutting fuel is the right thing to do. I believe we can.
  - sokoloff 12y ago
    How can you ensure that the fire detection circuitry isn't ever mis-wired in the field (such that when engine 1 catches fire that the computer knows that it's engine 1 and not engine 2)? If that happens (and it's certainly the case that engine indicators have been crossed before in the field), the pilot now couldn't shut down the engine with the fire (theorizing that the computer wouldn't allow the discharge of fire bottles and cutoff of fuel to the only engine NOT displaying a fire alarm). Or an engine that's streaming fuel (but not burning). Mid-ocean, you might want to cut the fuel to ensure you make landfall. Substitute other failure scenarios as you need, realizing that any system has cases where the pax die when an alternate system would have saved them, but also vice versa. It's not that we can't get systems that are right way more often than they're wrong, but that's still a long way to go from there to taking control from the pilot(s), IMO. A determined suicidal pilot can find a way. If they can't, then they don't even need to be on the airplane right? Disclaimer: I'm a [small airplane] pilot and well aware that I'm more likely to cause a crash from my own doing than from poor maintenance, but I definitely subscribe to the Boeing philosophy more than the Airbus...
    - 12thr0wit 12y ago
      Pilots really push back against this kind of safety feature. They want 100% control of their vehicles, including the ability to lift the landing gear while on the ground, because the risk of equipment failure that cannot be manually overridden is high.
      - sokoloff 12y ago
        Agreed. It's a very common feeling that "I am better than the automation." Even as an engineer and a fan of statistics, I share that visceral and inherent bias in my ego-laden monkey brain. On your specific example, there's little life-safety risk in inadvertent retraction of the gear. Sure, you're going to make expensive sparks, but it is extremely unlikely that anyone will die or even suffer life-changing injuries. The only likely injuries are from the slides and evacuation process. In contrast, being unable to raise the gear can be a life-safety emergency. Take a piston twin such as the one I fly. An engine failure with the airplane heavy and the gear down is a bona fide emergency. Suppose I take a bird or deer strike just after liftoff that damages the propeller and the "squat switch" (the weight-on-wheels sensor that prevents the gear being raised electrically). The prop damage may force me to shut down that engine at the same time that I can't raise the gear electrically (because the squat switch says we're still on the ground). I may not be able to climb the airplane with the available power on the left side engine while manually cranking up the gear, and I'm heavy, low, and slow. What first seems like a "duh, pilots would have to be idiots for arguing they should be able to raise the gear with weight on the wheels" is actually a tradeoff between a high likelihood of inadvertent switch mis-use with only financial consequences versus a very low likelihood of losing the aircraft an all on board. Of course, people are both bad at math, bad at understanding low probability/high impact events, and very uncomfortable with "costing" human lives such that a tradeoff can be made across financial losses and human lives lost.

sokoloff 12y ago

I agree that we can make great strides against unintentional CFIT accidents, but don't think we can rely on tech to eliminate intentional CFITs. Cut the fuel supply to both engines; what's the computer going to do? Now, if you protect against that, what's to prevent that system from stopping a pilot from extinguishing an in-flight fire in a situation where cutting fuel to one engine (or perhaps two/all) is the right response?

downandout 12y ago
But you're implying that we can't make the aircraft smart enough to know that there is a fire and that cutting fuel is the right thing to do. I believe we can.
- sokoloff 12y ago
  How can you ensure that the fire detection circuitry isn't ever mis-wired in the field (such that when engine 1 catches fire that the computer knows that it's engine 1 and not engine 2)? If that happens (and it's certainly the case that engine indicators have been crossed before in the field), the pilot now couldn't shut down the engine with the fire (theorizing that the computer wouldn't allow the discharge of fire bottles and cutoff of fuel to the only engine NOT displaying a fire alarm). Or an engine that's streaming fuel (but not burning). Mid-ocean, you might want to cut the fuel to ensure you make landfall. Substitute other failure scenarios as you need, realizing that any system has cases where the pax die when an alternate system would have saved them, but also vice versa. It's not that we can't get systems that are right way more often than they're wrong, but that's still a long way to go from there to taking control from the pilot(s), IMO. A determined suicidal pilot can find a way. If they can't, then they don't even need to be on the airplane right? Disclaimer: I'm a [small airplane] pilot and well aware that I'm more likely to cause a crash from my own doing than from poor maintenance, but I definitely subscribe to the Boeing philosophy more than the Airbus...
  - 12thr0wit 12y ago
    Pilots really push back against this kind of safety feature. They want 100% control of their vehicles, including the ability to lift the landing gear while on the ground, because the risk of equipment failure that cannot be manually overridden is high.
    - sokoloff 12y ago
      Agreed. It's a very common feeling that "I am better than the automation." Even as an engineer and a fan of statistics, I share that visceral and inherent bias in my ego-laden monkey brain. On your specific example, there's little life-safety risk in inadvertent retraction of the gear. Sure, you're going to make expensive sparks, but it is extremely unlikely that anyone will die or even suffer life-changing injuries. The only likely injuries are from the slides and evacuation process. In contrast, being unable to raise the gear can be a life-safety emergency. Take a piston twin such as the one I fly. An engine failure with the airplane heavy and the gear down is a bona fide emergency. Suppose I take a bird or deer strike just after liftoff that damages the propeller and the "squat switch" (the weight-on-wheels sensor that prevents the gear being raised electrically). The prop damage may force me to shut down that engine at the same time that I can't raise the gear electrically (because the squat switch says we're still on the ground). I may not be able to climb the airplane with the available power on the left side engine while manually cranking up the gear, and I'm heavy, low, and slow. What first seems like a "duh, pilots would have to be idiots for arguing they should be able to raise the gear with weight on the wheels" is actually a tradeoff between a high likelihood of inadvertent switch mis-use with only financial consequences versus a very low likelihood of losing the aircraft an all on board. Of course, people are both bad at math, bad at understanding low probability/high impact events, and very uncomfortable with "costing" human lives such that a tradeoff can be made across financial losses and human lives lost.