Wednesday, July 13, 2016

Reading about airliner incidents and accidents, I have a few humble suggestions for shoring up the problems that are still encountered.


  1. hypoxia
    1. In-cockpit cameras recording pilot's use of oxygen masks when required by rules/regs/laws. It is human not to don the mask, I get it. But it needs to be done. Review the video after every flight (and I do mean every flight) and if somebody didn't do it right then gently reprimand/teach/demote/remind/urge them to do it. Over time we can hopefully train everybody to really truly don the masks when it is right to do so. (Of course, the masks and tubing and valves and all that can have trouble too, so there should be pre-flight checks for them where the pilots actually use them during taxiing.)
    2. The autopilot should be able to bring a plane down to a less depressurized altitude if it detects a bad pressure situation. This feature should possibly be something that ground control can remotely invoke. Yes, I grok the security concerns.
    3. The autopilot should be able to land the plane at an airport if need be: if it doesn't get input from pilots that makes sense (the pilots should be quizzed every 25 minutes with a random question that a non-hypoxia-impaired person could easily answer); if the pressure goes wrong; if told to by ground control.
  2. data
    1. There is apparently an issue where airplanes don't really automatically report in often enough. It is expensive. It can be thwarted by the plane's satellite antenna being blocked (e.g. if things have gotten so bad the plane is inverted). It can be turned off (it should not be able to be turned off ever for pete's sake). So I suggest that a peer-to-peer system be implemented to complement the satellite one. Each plane will more frequently broadcast a minimal squirt of useful telemetry about itself. All other planes that can pick it up will record it. All that data will be automatically uploaded over WiFi as soon as the planes land.

No comments:

Post a Comment