From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Aviation / Aircraft (Rated C-class)
WikiProject icon This article is within the scope of the Aviation WikiProject. If you would like to participate, please visit the project page, where you can join the project and see lists of open tasks and task forces. To use this banner, please see the full instructions.
C-Class article C  This article has been rated as C-Class on the project's quality scale.
Checklist icon
Note icon
This article has been selected for use on the Aviation Portal.
Taskforce icon
This article is supported by the aircraft project.

The most recent additions to this article adds information but is desperately in need of editing. I'm not expert on this subject someone who's more knowledgable to edit it. Comatose51 02:34, 21 January 2006 (UTC)

Software Science or Control Engineering[edit]

It is not very informative to say that modern autopilots use software to control the aircraft, without specifying exactly what the software does, we might as well claim they work by magic.

It is more accurate to say that modern autopilots are implemented as computer software. As far as concepts are concerned this is mere implementation detail. The methods of autopilot design remains the body of knowledge called control engineering, which makes no assumptions concerning the implementation; the design may be implemented conceptually as analogue circuits, shift registers, computer code or even electro-mechanically. Gordon Vigurs 19:58, 29 May 2006 (UTC)

Scope of Article[edit]

The material has moved on from autopilots, which are basically systems to maintain straight and level flight, to inertial navigation, which serves to navigate specified classes of flightpath by generating autopilot commands. It would appear logical therefore to include ILS, TACAN, LORAN and Doppler navigation as well, but this would wander too far off the point.

In missile parlance, the autopilot serves to remove all the variability in airframe response due to altitude and Mach number, maintains stability with shifting centre of gravity position, and ensures a unity steady state gain. The input to the autopilot is typically a lateral acceleration command which is generated from the guidance (if derived from radiation emitted or reflected from the target) or navigation (if steered to a point in space).

In short, the autopilot is responsible for controlling the handling modes of the aircraft (short period pitch oscillation, and Dutch roll), but control of the trajectory is the responsibility of the navigation.

Pitch Oscillations: There is a difference between dampers and Autopilots. Since you mention "pitch oscillation" it needs to be stated that the system that corrects this is a "damper" or Stability Augmentation system, not an Autopilot which provides pilot relief. Even the sensors (rate gyros, etc) are different and provide not steady-state outputs, but rate signals that are only present when there is movement or acceleration. For example: Pitch stab aug was so critical on the F-4 that the Autopilot could not be engaged unless that channel was functional; no pitch aug no autopilot, and no heading hold, altitude hold, etc. Some pilots would even abort if the pitch aug did not work. Some aircraft like the T-38 only had dampers. The original A-10 not only had no autopilot (just dampers), but no inertial nav system either. MSgtUSAFret (talk) 22:15, 16 August 2008 (UTC)

The sentence beginning 'modern autopilots use software' appears to rename the autopilot 'flight control system' and calls the navigation system the autopilot.

The various categories of 'autopilot' presented are different flight path controllers, all of which potentially employ the same innermost loops (e.g. autopilot proper), and are limited in scope to aviation.

As mentioned above, trajectory control of a missile is governed by a navigation law, typically proportional navigation or some form of line of sight (command to line of sight or beam rider). Furthermore, the autopilot does not necessarily control normal acceleration and bank angle: in skid to turn missiles, they control pitch and yaw lateral acceleration, in some cases they control angle of attack. In a space booster, it is the orientation in space which is controlled in order to steer the vehicle along a pre-defined optimal attitude program. In a satellite, or spacecraft, the autopilot serves primarily to control the orientation in space.

In all cases the 'autopilot' refers to the innermost set of loops, whose purpose is to render the vehicle response reasonably constant in the presence of disturbances and variations in the flight regime (typically altitude and Mach Number) and vehicle mass distribution.

The article would be more compact and accessible if autopilot, guidance and navigation were treated separately.Gordon Vigurs 07:27, 30 May 2006 (UTC)

Further, the Computer system details clearly identify one specific architecture. The detailed implementations in hardware/software can be very different. Surely the way many automatic flight control system contractors go about implementing their system is as varied as the number of aircraft and other flying bodies that are currently or formerly in operation. Handment 21:46, 20 March 2007 (UTC)

It was mentioned that the autopilot can control every phase of flight except for taxi. If a reference is available, please insert it. As far as I know, to date, autopilots cannot control the aircraft down the center line of a runway. In some cases the AP may be available on the ground but in this case the pilot's manual will indicate that the AP can be engaged once the a/c has reached a defined altitude or a period of time after Weight On Wheels = false. In some cases the AP control laws do not allow engagement but the AP may be ARMED on the ground and will engage at a defined altitude or a period of time after Weight On Wheels = false. The reason for this is safety. Therefore, the AP can control ascent, level flight (cruise), descent, approach and landing. The Takeoff phase, as far as the control laws are concerned, ends when the AP is engaged. [BE]

It is mentioned that the AP is a part of the FMS in the write-up. The AP is separate code from the FMS (or FMGS). [BE]

As far as the trajectory, navigation and autopilot go, there is in most flight control systems a function called the Flight Director, or FD. The FD can take inputs from the guidance panel pilot settings to the (ie heading, altitude, speed, climb-modes, etc) or the FMS (or FMGS). The Flight Management System or Flight Management Guidance System (also separate from AP and FD) has numerous "jobs", one is to work with the radios to manage navigation by ground stations (eg VORs) and/or GPS. Most modern aircraft, commercial or military, are considered highly integrated as each of these systems is critically dependant on the other but as Gordon Vigurs mentions, they should be treated separately. Short story longer, the control laws for AP, FD, and the FMS are separate as should be the descriptions. Mention of the other system can be made as a point of reference only. [BE] —Preceding unsigned comment added by (talk) 15:59, 16 June 2008 (UTC)


An anonymous user from the address (talk · contribs) recently added the following note to the CAT IIIc entry in the Aviation Autopilot Categories of Landing section:

(Please look for Globalhawk, which might be the 1st to have such capabilty as of 2006)

"Globalhawk" presumably refers to the RQ-4 Global Hawk UAV. I have reverted the addition, as it was unreferenced and not written in an encyclopedic tone, but if someone with more knowledge about the subject than me could dig up a reference, the mention should presumably be included in some form. —Ilmari Karonen (talk) 01:03, 7 September 2006 (UTC)

Aviation Autopilot Categories of Landing[edit]

The descriptions of categories CAT III a/b/c don't match those in the Instrument Landing System-Article. I assume that they are both describing the same thing? 21:39, 3 October 2006 (UTC)

"It is usually a triple-channel system or dual-dual system." As far as I can see, the article doesn't explain what this means. Can someone provide a reference or an explanation? (talk) 15:11, 13 August 2008 (UTC)

Should this section be in the autopilot article? It really belongs in ILS. --Drpixie (talk) 02:50, 16 June 2014 (UTC)


This article doesn't credit any sources. It looked like a good building block for me to go from on a university essay I'm writing, but the lack of sources means I can't chase up what's said and verify for myself, and quite frankly referencing wikipedia for a piece of coursework isn't a great idea... —Preceding unsigned comment added by (talk) 14:09, 15 November 2007 (UTC)

Autopilots and TCAS[edit]

I know of no situation where an autopilot will take action in a TCAS RA situation, and have removed this line from the article. The first step is invariably to disconnect the autopilot and flight directors if fitted. —Preceding unsigned comment added by (talk) 00:10, 27 August 2008 (UTC)

Self-locating INS?[edit]

The statement "IRU's are completely self-contained and use gravity and earth rotation to determine their initial position (earth rate)." doesn't ring true to me. I'm happy that IRUs can self-align (to find North and Up) without external inputs, but believe that they all need an external Initial Position - either in the form of a manual IP insert, or from some alternative navaid like GPS. I'm inclined to trim this suspect detail out rather than to expand (since it's a long way off-topic), but thought I'd better give others a chance to comment first since my detailed interest in IN design ended almost twenty years ago. Zeusfaber (talk) 21:50, 20 February 2009 (UTC)

Autopilots capable of landing a plane[edit]

When did the first autopilot capable of successfully landing a plane come into use? (talk) 20:28, 15 April 2009 (UTC)

See Autoland. Rwessel (talk) 01:56, 31 October 2009 (UTC)

Autopilot invention date (1912/1911)[edit]

After reverting the edit by, it appears that there were several different dates involved. The original work on *aircraft* autopilots appears to have been done by Sperry in 1912, and publically demonstrated in 1914. 1911 was the year in which the US Navy accepted a *marine* gyroscopic gyrocompass/autopilot (, and development of that dates several years further back (patents seem to date back to at least 1908 - #1,242,065, for example). Rwessel (talk) 04:09, 16 November 2010 (UTC)

Gyrocompass bases on a quite different physical phenomenon than gyroscope, as explained in the article. I doubt if gyrocompass could be used for a marine autopilot, as (in its early forms) it reacted very slowly. The gyro_scope_ was used for marine (even submarine) autopilot much earlier than 1911, if I remember correctly around 1890s? --Kubanczyk (talk) 00:00, 9 January 2013 (UTC)

Modern Autopilot Image[edit]

I have a WP:COI with Honeywell Aerospace, who has donated several images under a Creative Commons license. I added an image of a modern autopilot system they contributed. Since the image is not clearly distinguishable as a Honeywell product and I did not mention them in the caption, I didn't feel this was a COI edit and went a head and put it in. Corporate Minion 22:08, 5 September 2012 (UTC)

Source of the term "George"[edit]

The cite states one possible etymology for "George" was George Debeeson.... "one of the key contributors to its development.[1]" This seems dubious.

That same cite article contains a section written by George Debeeson's son, Max. In that article, the son writes, "Whether George's invention was incorporated into the popular mechanism of the day or not is uncertain."

Prior to World War II, I can find no references to the term "George" as an autopilot. After the war, the term is quite common. It is highly doubtful pilots of that era would be aware of an individual who's son isn't certain of his father's contribution to a device invented and commercialized by the well known Elmer and Lawrence Sperry of Sperry Corporation. Why not, "Let Sperry Do It?"

According to a number of World War II pilots, the term was coined by the RAF. There are three theories: 1) "Let George Do It," referred to the aircraft's owner, King George VI (reigned 1936:1952); 2. The device is essentially a gyroscope. During World War II, the military phonetic for G (Gyro) was George (it is now Golf)<^ "The International Phonetic Alphabet for Radio Communications". Communications Specialists. Retrieved 2009-02-27. ></ref>; 3) A popular phrase from the first half of the twentieth century and the 1940 George Formby film, To Hell with Hitler ("Let George Do It" was the original title). The phrase means when something has to be done, let somebody else do it. From 1934 to 1945, George Formby was the top comedian in British cinema and in 1939, was the number one film star of all genres.<^ "FORMBY IS POPULAR ACTOR.". The Mercury (Hobart, Tas. : 1860 - 1954) (Hobart, Tas.: National Library of Australia): p. 5. 25 February 1939. Retrieved 24 April 2012.></ref>

I would propose that at minimum, these alternative explanations be included.Jjlaughlin (talk) 07:43, 22 November 2012 (UTC)

Fixed. I've simply removed the dubious statement, as it is clear it is incorrectly sourced. It has minimal relevancy but introduces a great risk of misinforming the readers. --Kubanczyk (talk) 23:52, 8 January 2013 (UTC)

Removed picture of B737-800 autopilot, picture is clearly a screenshot from a flight simulator program.[edit]

I have removed from this article the alleged picture ( of the autopilot from a B737-800 as the picture was clearly a screenshot from a flight simulator program that was then cropped. As an encyclopedic article, I believe this did not constitute an accurate real-life representation of a real B737-800's autopilot interface.

As an aside, I am unfamiliar with nominating images for deletion. As the picture in question is not in use anywhere else (as of writing) and is clearly not a photograph of an actual B737-800's autopilot, perhaps it would be prudent to nominate this image for deletion? King Arthur6687 (talk) 02:33, 29 September 2014 (UTC)