4 ms·

What does "full and absolute control" mean in the context of a fly-by-wire aircraft? AA587 is blamed on a pilot's inputs driving a structural capability exceed

What does "full and absolute control" mean in the context of a fly-by-wire aircraft? AA587 is blamed on a pilot's inputs driving a structural capability exceedance in the vertical stabilizer. For reasons such as these, all fly-by-wire aircraft will have some measure of envelope protection whereby pilot inputs projected to exceed aircraft capability will be limited. There is always "interpretation" of pilot inputs involved in fly-by-wire systems. The key is in making available all control power to the pilot that will not result in catastrophic structural failure of the aircraft. And the associated challenges are knowing exactly where those boundaries should be set in the design of the control system, as well as the knowledge of the aircraft state to the flight computers.

neuronic 8y ago
Thank you for the extensive response. With that context, I would say "full control" should amount to the ability of a pilot to overrule systems that may send an aircraft nosediving due to faulty sensors. Did the pilots of the Lion Air flight and possibly the Ethiopian Air flight understand what was wrong in-flight and were unable to fix it due to an inability to override aircraft controls? This is what I am getting out of the discussion so far but maybe I am gravely misunderstanding what's happening? So as a passenger without any knowledge of aviation systems my question is merely whether the pilot gets control over the aircraft when he demands it? Fly-by-wire should at least never lead to autopilot suicide, even if it has to ensure flight integrity to a certain degree due to the technical nature of large, modern aircraft.
- m0llusk 8y ago
  We demand this level of conscious awareness and control of 4-5 year old children in our society but machines that carry dozens of lives are not held to such a standard of attention and interaction. This might be something that AI can address, but most current offerings still lack features we consider essential for even very young children.
- ThenAsNow 8y ago
  > Thank you for the extensive response. With that context, I would say "full control" should amount to the ability of a pilot to overrule systems that may send an aircraft nosediving due to faulty sensors. Thanks for reading it :) . As you phrased it specifically, I can't disagree with you. But in these systems, the devil is in the details. For Boeing's MCAS, which is like an add-on piece of functionality to the core control system, it seems like an obvious piece which should have a manual override - and it does. The criticism, as I understand it based on Lion Air, is that the override functionality is not obvious and if you were previously a 737 pilot on non-Max variants, you might not know about this or have practiced its use. However, consider safety systems such as Automatic Ground Collision Avoidance System ("Auto GCAS") or Automatic Air Collision Avoidance System ("Auto ACAS"). These systems are designed as reactions to multiple known cases of human error that have resulted in otherwise under-control aircraft being flown into the ground or other aircraft. They are specifically designed to override a mistaken human. Technically, it's not impossible that pathological failures or spoofing could trigger these automated systems, which would wrest control away from pilots. Should such systems be overridable at all times? Resolving these kinds of issues does involve some judgment and adopting a philosophy prior to the implementation. Airbus and Boeing, for example, have historically differed in their philosophies, each with reasonable rationale. I personally favor the approach Boeing has historically used, with AF447 and QF32 illustrating some of the pitfalls of Airbus' approach. For example, as pointed out by ggreer in https://news.ycombinator.com/item?id=19355402 https://news.ycombinator.com/item?id=19355402 , being susceptible to common mode faults. However, I still fly on either manufacturer's aircraft without hesitation.

neuronic 8y ago

Thank you for the extensive response. With that context, I would say "full control" should amount to the ability of a pilot to overrule systems that may send an aircraft nosediving due to faulty sensors. Did the pilots of the Lion Air flight and possibly the Ethiopian Air flight understand what was wrong in-flight and were unable to fix it due to an inability to override aircraft controls? This is what I am getting out of the discussion so far but maybe I am gravely misunderstanding what's happening? So as a passenger without any knowledge of aviation systems my question is merely whether the pilot gets control over the aircraft when he demands it? Fly-by-wire should at least never lead to autopilot suicide, even if it has to ensure flight integrity to a certain degree due to the technical nature of large, modern aircraft.

m0llusk 8y ago
We demand this level of conscious awareness and control of 4-5 year old children in our society but machines that carry dozens of lives are not held to such a standard of attention and interaction. This might be something that AI can address, but most current offerings still lack features we consider essential for even very young children.
ThenAsNow 8y ago
> Thank you for the extensive response. With that context, I would say "full control" should amount to the ability of a pilot to overrule systems that may send an aircraft nosediving due to faulty sensors. Thanks for reading it :) . As you phrased it specifically, I can't disagree with you. But in these systems, the devil is in the details. For Boeing's MCAS, which is like an add-on piece of functionality to the core control system, it seems like an obvious piece which should have a manual override - and it does. The criticism, as I understand it based on Lion Air, is that the override functionality is not obvious and if you were previously a 737 pilot on non-Max variants, you might not know about this or have practiced its use. However, consider safety systems such as Automatic Ground Collision Avoidance System ("Auto GCAS") or Automatic Air Collision Avoidance System ("Auto ACAS"). These systems are designed as reactions to multiple known cases of human error that have resulted in otherwise under-control aircraft being flown into the ground or other aircraft. They are specifically designed to override a mistaken human. Technically, it's not impossible that pathological failures or spoofing could trigger these automated systems, which would wrest control away from pilots. Should such systems be overridable at all times? Resolving these kinds of issues does involve some judgment and adopting a philosophy prior to the implementation. Airbus and Boeing, for example, have historically differed in their philosophies, each with reasonable rationale. I personally favor the approach Boeing has historically used, with AF447 and QF32 illustrating some of the pitfalls of Airbus' approach. For example, as pointed out by ggreer in https://news.ycombinator.com/item?id=19355402 https://news.ycombinator.com/item?id=19355402 , being susceptible to common mode faults. However, I still fly on either manufacturer's aircraft without hesitation.