Quora.com is a question-and-answer site where content is written and edited by its community of users. Occasionally we syndicate content from the site if we think it will interest TPG readers. This article originally appeared on Quora.com in response to the question How Does a Plane Land on Autopilot? and was written by Bruno Gilissen, an A330/A350 pilot.
For an automatic landing, the airplane needs a minimum amount of systems to be operational. For the Airbus that’s in broad brushstrokes, like this:
- Obviously at least one autopilot (one is enough and note that an autothrust is not necessary, you can do an autoland with manual thrust. I actually just did one like that last night in the simulator.)
- A minimum on instruments and announcement capability for the aircraft to notify the pilots if something goes wrong, like a flight warning computer, an attitude indicator or a flight mode annunciator; also air data and/or inertial reference systems that drive things like speed and attitude indications.
- A minimum on flight controls that ensure the autopilot can guide the plane according to specifications, like some flight computers (for the fly by wire on Airbus), rudder trim, yaw damper, at least two out of three hydraulic systems, flaps and slats.
- At least one radio altimeter. The aircraft uses radio height and not inaccurate barometric height to guide the flight controls to ever smoother and smaller corrections lower by the ground. Furthermore, the radio height is used to determine when the aircraft has to initiate the flare.
- At least two ILS or GLS receivers. This relates to the type of guidance the airplane receives from ground based and/or GPS signals. GLS is rare and I’ll refer to the ILS system a bit further later. If the airplane doesn’t get guidance from these signals an autoland is impossible.
- Finally, if an automatic rollout and braking is planned, as well as a functional braking system and nose wheel steering.
These items should give you a hint at how an autoland works:
- The airplane gets signals from outside,
- Uses its own instruments to determine what corrections to make and fly the plane perfectly,
- And fine tunes this with ever smaller corrections as it gets lower to the ground.
- Finally, at a certain exact (radio) height above the ground, it initiates the flare and touchdown.
- After touchdown the automatic braking and roll out happens, based again on both the outside signal and the plane’s instrumentation.
- All along, it warns the pilots if something goes wrong so they can take over if needed, which — granted — can be challenging if it happens very low to the ground. The “No Flare” is the tricky one that usually comes up in simulator training and almost everyone hits the ground hard before they realized it or have time to react to it.
The signal that the airplane receives from the ground based station (or GPS if GLS is used) is paramount in assuring all happens safely. 99.9% + of autolands are guided in with an ILS or Instrument Landing System. In fact, in my career, I still have to fly the first GLS approach (or even MLS, both are different systems that can be used), but I understand it may be more common in the USA.
These ILS signals are typically assessed by the airline before it approves its pilots to perform an autoland at a specific runway. The quality of the ILS beam and the effect of the terrain profile before the runway should be checked. A little hill or valley before the runway will influence the radio altitude and therefore will mess with the guidance.
The airports have to abide by certain equipment regulations if an autoland is to be performed in low visibility conditions. Though it’s not directly related to an autoland, as you can do one in perfectly fine weather as well. It just almost never happens because pilots love to land manually. They are only forced into an autoland if they can’t see enough themselves to land manually.
Airport equipment regarding this has to do with power backup systems and signs to notify ground personnel to keep clear of areas where antenna interference can happen. Depending on the visibility there could also be equipment specifications regarding lights so the pilots can see where to land and get off the runway. However, I must state again, as these things are often confused, that those requirements have more to do with low visibility conditions than with autoland itself.
The same confusion between autoland and cat 2/3 operations (which is the more technical term in the industry for low visibility operations) leads people to think that an autoland has to be prepared by the pilots before passing 1,000 feet in approach.
That’s not so: a pilot can decide at the last minute to let the plane autoland as long as he has assured himself beforehand that the airline has done an assessment about that runway (typically that’s written somewhere in an airport briefing of the airline), the airplane displays the capability to autoland, the pilot can see the runway and is ready to take over at any time (since the airport hasn’t gone through all it’s procedures to safeguard the ILS signal quality if weather doesn’t dictate that).