Modern commercial aircraft almost always have some form of autopilot onboard. Up-to-date systems are usually computer-controlled and can be highly sophisticated, but the underlying processes involved have changed little in more than a century. So, how do these systems actually work?
A technology that’s changed little in over a century
In 1933, famous eyepatch-wearing aviator Wiley Post touched down at Floyd Bennett Field in New York, completing the first around the world solo flight in seven days, 18 hours and 49 minutes. His journey had begun in his Lockheed Vega Winnie Mae on July 15th that year, flying nonstop to Berlin in Germany. He continued to the Soviet Union, making stops along the way, before hopping across to Alaska, Canada and finally back to where he started in New York.
This was Post’s second famous flight, having completed an around the world operation with navigator Harold Gatty onboard to help keep him on course and alert. So how did he manage to complete a similar flight with no extra pair of hands onboard? The answer was autopilot, which kept him flying in the right direction while he rested.
Post’s autopilot was not the first time autopilot had been used. Indeed, US aviator Lawrence Sperry created the first successful autopilot back in 1912. Today, autopilot is a common feature on most commercial aircraft. Remarkably, it has changed very little in the past 100 years.
Stay informed: Sign up for our daily and weekly aviation news digests!
How does autopilot work?
While autopilots can be present on everything from ships to cars, naturally, we’ll be focusing on airplane autopilot technology. In an aviation sense, autopilot can be more accurately described as the automatic flight control system (AFCS). Originally designed to provide relief to pilots during the long and often tedious cruise stages of flight, modern AFCS systems can carry out some fairly advanced maneuvers.
Depending on the system, the autopilot will be able to control the elevators, the rudder, the ailerons or even all three. Basic ‘single axis’ autopilots control one piece of equipment, usually the ailerons, to keep the aircraft on an even keel. More advanced two- or three-axis autopilots have access to more of the controls.
Modern autopilots use a computer with a high-speed processor to control the aircraft, but the underlying technology is very much as Sperry designed it in 1912. Using a gyroscope and altitude indicator remains the fundamental basis for modern autopilot, although now they are more accurate.
Other aircraft systems involved in autopilot function include compasses, airspeed indicators and accelerometers, as well as receiving GPS signals to indicate the plane’s position. Armed with all this information, the autopilot is capable of not just keeping a straight and level path, but executing an entire flight plan.
Taking all the data from these various inputs, the autopilot ensures that the aircraft retains the correct pitch, speed, heading and altitude. Slight adjustments are performed by sending signals to servomechanism units, which use either motors or hydraulics to alter the aircraft’s control surfaces.
The crucial element of this activity is that it creates a feedback loop. The pilot sets the control surfaces and engages the autopilot. Based on the data gathered by the autopilot, it makes some adjustments, which changes the data set, and then the process starts again.
How much does the autopilot actually fly?
Autopilots are becoming increasingly intelligent and are capable of performing complex maneuvers. Although they are primarily used only in the cruise stage of flight, they are perfectly capable of landing the plane. Indeed, in situations where visibility is very poor, autoland can be used to bring the plane down. This is only possible at some airports with particular technologies installed and is used very rarely, with pilots landing manually 99% of the time.
Recently, Airbus successfully demonstrated the use of autopilot in a takeoff situation also. One of its A350-1000s took off by itself from Toulouse in December 2019, proving that it is possible. The Autonomous Taxi, Take-Off and Landing (ATTOL) demonstrator project is part of the plane manufacturer’s UpNext program, looking at future technologies for aircraft. However, right now, this is just a demonstration, and all takeoffs are performed manually.
Autopilot can be compared somewhat to cruise control on a car. It’s useful for maintaining a constant speed on the motorway, but it wouldn’t be useful for leaving the garage or trying to park in a space. Although autopilot is a great resource for giving pilots a bit of a break, they are still responsible for the inputs, and still in control of the plane.
