PDA

View Full Version : Autopilot question


GXER
5th Mar 2009, 07:30
Just wondering (from a non-flyer perspective but having some experience of autopilots on boats) - if you have a route programmed in the A/P that commands a turn to a new heading at some point, do you (professional pilots) check (by reference to compass or other heading sensor) that the turn has been correctly made onto the new heading or do you just leave the automation to get on with it and assume it has done what it should have done?

Northbeach
5th Mar 2009, 15:14
On our navigation displays we have the different fixes connected by the route we are assigned to fly. By looking at the display we are able to verify that the jet is proceeding to the next fix on the flight plan. Often we have to fly a particular course from one fix to the next. They are called jet airways. The heading necessary to fly that particular course will vary due to the winds aloft. Just the way your boat’s heading will vary when you are trying to steer a straight course to the next navigation buoy depending on how fast the tide is running and how strong the wind is. So, yes we do verify the jet is going where it is supposed to, but that verification is probably not checking some specific heading.

As a side note this is one area that does cause a lot of problems. Typically what happens is jet going where it was programmed to go, but that programming is in error so the jet is actually not where air traffic control thinks it should be. When air traffic control asks a pilot: “Where are you going” or “Say you’re heading” those questions causes an adrenalin shot and a mad scramble on the flight deck for the flight plan and uneasy looks down at the flight management system (so I have been told, because I of course have never done anything like that). The computer is doing what it was told, but there was a breakdown between what air traffic assigned the flight and what the autopilot was commanded to do. Fortunately most of these errors are caught and the fix is simple, sometimes not with tragic consequences.

deltahotel
5th Mar 2009, 15:17
The route is programmed into the Flight Management Computer and when the lateral navigation function (Boeing Lnav, Airbus ?) and A/P are engaged the A/P is used to follow the programmed track. The track is shown on the HSI screen (Boeing - magenta line), so it is visually obvious if the a/c is not following the track. If you need to go off the planned track (ATC vectors eg), the heading is selected and the HSI (which has a compass overlay) is monitored to ensure it turns onto the correct heading.

hope this helps

dh

SNS3Guppy
5th Mar 2009, 23:56
The autopilot is never left to it's own devices. We don't simply check the heading during a turn or when a turn is completed, but maintain a watch at all times heading, altitude, airspeed, power settings, etc.

When the autopilot makes a turn, I follow the turn using a manual "heading bug," which is a small marker built into the instrument that is used to keep track of a specific heading. A knob is rotated which moves the "bug" around the instrument bezel or display to highlight a particular heading. When the airplane is turning to the new heading, the bug is manually rotated to highlight the new heading...the pilot flying (and usually the pilot not flying) is verifying not only that the aircraft is turning to the correct heading, but that it's correctly displayed, highlighted, and that it concurs with the flight planned data (or ammended data).

Old Smokey
6th Mar 2009, 15:05
From the very lowest levels of automation (simple Auto-Pilots able to hold a Heading, Airspeed etc.) through to the most modern where FMC/FMS/Auto-Pilots accurately execute Lateral and Vertical Navigation, the output commands are scrupulously checked by flight crew.

The "Automatics" are NEVER left to "do their own thing".

Modern systems are extremely accurate and reliable. The weak link is inevitably the human, and human pilots are well aware of the GIGO concept (Garbage In, Garbage Out). So, we very carefully check and cross-check the human inputs (the GI component), and scrupulously monitor performance output to be alert to GO component.

Note - The human errors may not be the pilots' errors. I've encountered several serious errors with new Navigation Databases, the human at the data supplier has slipped up.

Regards,

Old Smokey

Pugilistic Animus
10th Mar 2009, 17:53
Old Smokey,

I think You said it best with the advice "Trust Nothing; Trust No One":ok: