4
$\begingroup$

On this A380 flight after pushback and engine start and before taxi, Capt Juergen Raps and and Sr FO Harald Tschira rely solely on their FID to check flight controls. On a regular flight without the Pilotseye.tv camera crew, they obviously can't see if flight controls respond.

Why isn't a human third-party (ground crew) required to confirm the flight controls' movements? This is obviously shrewder, and adds another safety measure to forestall the possibility that the FID shows a flight control to move when it doesn't.

Improve this question
$\endgroup$
3
  • $\begingroup$ "This is obviously shrewder". Is it? What makes you think that a human is more reliable than some other method? "... adds another safety measure". Is that more useful and cost effective than adding other safety measures, or than adding nothing at all? Additional 'safety' measures aren't free: they have direct costs for implementation and training, and indirect costs for human error. $\endgroup$
    – Pondlife
    Jun 9, 2019 at 6:38
  • $\begingroup$ Big plane like that, don't they have TV cameras aiming at those surfaces? $\endgroup$
    – Harper - Reinstate Monica
    Jun 9, 2019 at 7:04
  • $\begingroup$ They have for landing gears. Can anybody imagine the flight control system pages not reliable! $\endgroup$
    – user40476
    Jun 9, 2019 at 13:59

1 Answer 1

Reset to default
7
$\begingroup$

enter image description here
Same linked video @ 7:03

[A] safety measure to forestall the possibility that the [display] shows a flight control to move when it doesn't.

The display (above) doesn't mimic the input, it shows the positions of the sensors independent of the input signal. Since the position change that is being checked is confirmed to be related to direct input, the data is valid.

In other words, an unexpected result with/without input would be an issue with a sensor and/or surface.

In lieu of digital displays, planes like the earlier 747 variants and Concorde had analog instruments that also read sensors. And with the hydraulics operating on a plane like the 737, the fact that the controls are able to move the full range unjammed, indicates that the control surfaces are responding and are also unjammed.

It gets trickier with a plane like the MD-80 where the controls move tabs instead of surfaces (a tab is a control surface on a control surface). It has resulted in an overrun in 2017. No human [visual] inspection could have detected it as the tab would have still moved as expected, but not the stuck surface, also as expected – tabs need airflow to move the surface.

Improve this answer
$\endgroup$
5
  • $\begingroup$ But that information comes from sensors which themselves may or may not be functioning. It’s also not a true recording of the flight control movements. $\endgroup$
    – Romeo_4808N
    Jun 9, 2019 at 13:41
  • 1
    $\begingroup$ @CarloFelicione: Hence the line before the second image. Input = expected result = valid. Unexpected result with or w/o input would be an issue with the sensor/surface. Unless it's a nefarious sentient sensor watching the input. $\endgroup$
    – user14897
    Jun 9, 2019 at 13:49
  • 2
    $\begingroup$ Exactly, the sensors are an independent system to the actuators. If they're in sync it's because the surface is moving correctly. A human can also screw up. If you move one surface then the other and they operate the wrong way around a human is likely to say "ok" assuming that they were activated in the other order. They could also be asleep, distracted, .... $\endgroup$
    – Dannie
    Jun 9, 2019 at 15:26
  • $\begingroup$ "No human inspection could have detected it as the tab would have still moved as expected" You can inspect it simply by going up to the surface and manually moving it through the range of motion, as is common on smaller planes. The issue is one of cost and time because you need a lift platform to do so on the MD-80. $\endgroup$
    – user71659
    Jun 9, 2019 at 23:06
  • $\begingroup$ Right; clarified it. $\endgroup$
    – user14897
    Jun 9, 2019 at 23:19

You must log in to answer this question.