from Wikipedia, the free encyclopedia

An automatic, usually programmable control system is referred to as an autopilot , which automatically steers means of locomotion if desired without people having to intervene in the control while the autopilot is active. Typically, it is a computer that processes environmental information from the vehicles' instruments to determine how to control the vehicle.

In the event of difficulties, visual or acoustic warning signals are given.

Control elements of the autopilot in the Airbus A340


The term "autopilot" is a compound of the two terms " automat " and " pilot ".

The word pilot, which comes from the French “pilote”, has its origin in the Greek term “ pedotta ”, which can be freely translated as helmsman , pilot. Today, however, this definition is no longer in use. The aviation lexicon defines a pilot “as part of the crew on board an aircraft”. “[More precisely] it is the pilot” who controls the aircraft. The Duden defines "pilot as someone who [professionally] controls an airplane", which is also due to the fact that today the term is mainly associated with the aviation industry. However, according to Duden, there are other meanings of this term, e.g. B. racing driver or "someone who drives a bobsleigh". Here, a pilot in sailor language is also used as a pilot called. In the past, the term pilot was often used more broadly to refer to a driver who was driving a vehicle of any kind.

The second term, “automat”, comes from the French “automate”, which in turn comes from the Greek “autómatos”. Translated it means something like "moving by itself, of its own accord". Automata are therefore machines that independently, i.e. automatically, execute predetermined processes.

The American Federal Aviation Administration , FAA for short, defines an autopilot as a "[...] flight control system that automatically addresses the wing, elevator and, in some cases, the vertical stabilizer to the route that was [given by the pilot]." or follow altitude, rate of descent / climb and course ”.

However, its definition is only valid in the field of aviation and even there is not generally valid for all systems. So their definition is not entirely sufficient. If you combine the two terms as defined above, you can define "autopilot" as a system that automatically makes independent decisions based on external influences and information and then a vehicle, be it an aircraft or another type of vehicle Vehicle controls. It can thus influence the position, direction and speed of the vehicle.

application areas

Autopilots are used today in a very wide variety of fields and for different purposes. Above all, this includes relieving the burden on people, which has become increasingly necessary due to increasingly complex systems. Today autopilots are mostly only associated with aviation, and most people are only familiar with autopilots that are used in aircraft . However, other systems and vehicles are already being equipped with an autopilot. Apart from aviation , autopilots, which correspond to the definition above, are also used in space , maritime , automotive and robotics to control vehicles.

Autopilots in the shipping industry

Autopilots on ships take control according to a fixed course, according to a specified route (waypoint list) or, on sailing ships, according to the wind (wind vane control). On larger ships, they are integrated into a comprehensive electronic navigation system (ECDIS) .

Autopilot on a ship

The autopilot calculates the necessary course correction from various data sources. The steered course ( compass course or course over ground ) is measured and compared with the given course over ground. A control signal is generated from the difference, which acts mechanically or hydraulically on the steering system and causes the ship to change course via the rudder blade.

Modern devices determine the course of the ship with a gyro compass or GPS . They also take into account rolling and streaking movements caused by swell and gusts of wind, wind rotations, currents ( drift : difference between course through water and course over ground ) and changes in the earth's magnetic field. Routes can be entered via a waypoint list or by clicking directly on the nautical chart, which the autopilot then automatically follows. When a waypoint is reached, acoustic information is given and the device automatically switches to the next waypoint.

On large ships, the autopilot is an integral part of a comprehensive electronic navigation system ( ECDIS ), which also includes echo sounder , radar , AIS , electronic nautical chart and much more. Due to the advancing miniaturization and the drop in prices of the technology required for this purpose, ECDIS-like systems - mostly consisting of a notebook with navigation software and peripheral devices connected via NMEA interfaces - are increasingly finding their way into smaller ships and even on recreational craft vehicles.

Incorrectly programmed autopilots or autopilots that “get out of hand” can have serious consequences. The latter occurs in particular with systems that are too weakly dimensioned, or with particularly violent waves and winds or with sailing courses in front of the wind , if the automatic control reacts too weakly or too strongly.

Wind vane control

Mechanical self-steering wind vane system in operation
Mechanical wind vane control

The mechanical wind vane control uses a wind vane at the stern of the ship, with which the wind direction is measured relative to the ship. It reacts to both wind changes (change in wind direction) and ship rotations (change of course). This change of direction is mechanically transmitted to the rudder blade of the ship in such a way that the angle between the ship and the apparent wind always remains the same. As a result, the ship travels straight ahead with the wind direction remaining the same. The course only needs to be corrected in the event of major wind changes.

Mechanical wind vane controls are often used when sailing around the world. They are robust and do not require any electrical energy.

Electronic wind vane control

With the electronic wind vane control, the rotation of the wind vane on the top of the mast is fed to the autopilot as an electrical signal. This then calculates the necessary correction from this, taking into account the given and the actual course of the ship.

Tiller steering

With the tiller control, the electrical control signal acts mechanically directly on the tiller via an electrically driven push rod.

Wheel steering

With wheel steering, the electrical control signal acts directly on the steering wheel via an electrically driven toothed belt.

Hydraulic steering

In the hydraulic control, the electrical control signal acts via a hydraulic pump and hydraulic valves on a cylinder which, with oil pressure, moves a push rod that acts directly on the rudder quadrant and thus moves the rudder blade.

In the automobile

Autopilots in motor vehicles are usually called driver assistance systems . An autopilot intervenes in the control of the vehicle by definition. Individual autopilot functions are, for example:

  • Lane departure warning: Keeps the vehicle in the marked lane.
  • Lane change assistant: changes lanes on command of the driver. (But usually don't pay attention to the traffic).
  • Collision avoidance assistant: Warns of impending rear-end collisions and brakes automatically in conjunction with the emergency braking assistant.
  • Side collision avoidance assistant: Dodges sideways if another vehicle comes too close.
  • Brake assistant : recognizes an emergency braking situation based on the speed at which the brake pedal is depressed and then brakes faster and harder than a person does.
  • Cruise control : regulates the vehicle speed.
  • Crosswind Assist: Stabilizes the track in gusty crosswinds.
  • Parking assistant: detects parking spaces and parks lengthways or crossways.

The degree of automation of an autopilot is described by autonomy levels from level 0 to level 5 (see autonomous driving ). For some time now, test vehicles from Google have reached level 4. This means that the vehicles can drive completely autonomously. For safety and legal reasons, there is still a (responsible) driver in the vehicle. There are also efforts in Europe to establish self-driving vehicles . One project that has already been successfully tested in Spain is SARTRE (Safe Road Trains for the Environment). This is a convoy system that automatically connects all vehicles in a network and controls them fully automatically. In the test in Spain, a truck controlled all vehicles behind it, which transmitted all of their measured data to the control vehicle.

Another step in the direction of autonomous control is the Highway Pilot System, which is used in the Freightliner Inspiration Truck . The truck is based on the US production model Freightliner Cascadia Evolution, which is equipped with Detroit Connect (on-board diagnostic and fleet monitoring system) and Highway Pilot technology. The latter includes front radar, a stereo camera and tried-and-tested assistance systems (distance control). Since May 2015, two trucks of this type have had a license for road traffic in Nevada, USA.

Tesla, Inc. provides driver assistance in its vehicles known as "autopilot". These are functions of level 2 autonomy. B. The lane can be changed automatically by tapping the turn signal lever. The competition center is suing Tesla because Tesla is giving the misleading impression that its cars can drive autonomously on German roads.

Space travel

Autopilots have become indispensable in the space industry today. Without them, the research and commercial use of would universe impossible. Autopilots are used in rockets , satellites and probes to control them during take-off, to keep them in stable orbit and to independently explore planets or other objects in the depths of space. One example is the Curiosity rover of NASA , which currently on Mars is in use.


The steadily growing complexity of aircraft and the increasingly dense traffic would require a high degree of concentration and work from the pilot if he were to steer the aircraft manually. “An autopilot can relieve the pilot of his monotonous and tiring task of controlling the aircraft [...]. The pilot is thus free to concentrate on other tasks [...] ”. Apart from precise control of the aircraft, autopilots also take on other tasks and not only support the pilot in normal flight. Modern systems are able to support the crew during landings and take-offs in poor weather conditions, such as strong winds and fog. But also “fully automatic landings in poor visibility up to zero visibility” are among the application areas of the system. Autopilots are also used in modern helicopters. "Controlling a large helicopter [...] requires a lot of work and concentration on the part of the pilots, as they are based on different requirements". The ability to hover in particular places very high demands on helicopter autopilot systems.

historical development

The first systems, named “autopilot”, were developed for the shipping industry before they found their way into airplanes. This began with the invention of the gyro compass by Hermann Anschütz-Kaempfe (in 1904) and Elmer Sperry (in 1908). This technical achievement made it possible to obtain a reliable reference of the course. This was previously difficult to do with the help of a compass on ships built from steel due to the magnetic properties of ferrite . These gyrocompasses were quickly adopted by the Navy. Elmer Sperry then developed the first gyrocompass-controlled automatic control system, which quickly established itself in the shipping industry. However, Sperry's autopilot was "a massive unit that was only practicable in shipping, [...] but quickly established itself as the standard for new shipbuilding" Elmer Sperry's son Lawrence finally brought his father's invention onto the plane. On June 18, 1914, in France, he demonstrated the first system that could keep and stabilize a Curtiss C-2 in level flight without the intervention of a pilot. “This autopilot consisted of four separate gyros, each rotating at 7,000 revolutions per minute. These gyroscopes were set to the zero position of all control surfaces and were mechanically connected to them ”. Even then, this early autopilot was able to make landings without the intervention of the pilot.

Due to the rapid technical development, especially in electronics, in 1947 the first electronic autopilot steered a C-54 Skymaster of the US Air Force across the Atlantic from the USA to England completely autonomously from take-off to landing in Brize Norton . At that time, the British trade press was not really enthusiastic about the performance of the fully automatic flight. It was also clear then that such a system would take a long time to become firmly established. In the October 1947 issue of the Flight and Aircraft Engineer magazine , the autopilot was described as a "glorified electrical alarm clock [] that processed predefined settings [...] when a certain time passed and thus passed on inputs to the autopilot [so that it controls the aircraft] and controlled other systems, such as the landing gear, flaps and engine ”. This example shows impressively that some of the earliest autopilot systems already mastered many things that one would expect from a modern computerized system today, such as autothrottle , autotrim , yaw dampers and automatic landings.

Modern autopilots

Concorde at the start
PA-200 tornado
Northrop Grumman X-47B

From the first purely mechanical to the fully electric autopilot, increasingly fully computerized systems were developed, controlled by complex algorithms on ever faster microprocessors . This process began in the mid-1950s with the first fly-by-wire technology (FBW) in military aircraft, back then only in the USA. In Europe, FBW did not begin until the late 1960s, when Germany, France and England jointly developed the PA-200 Tornado . The Airbus Group benefited from the technical achievements and experience in the development of this multi-purpose fighter aircraft and then developed the Airbus A-320 , one of the first commercial aircraft to be controlled exclusively by FBW - the first was the Concorde .

Due to the increasing computerization and networking, the tasks of the autopilot grew steadily. Modern systems as they are used today even help to reduce noise pollution and save fuel. In addition, the trend is towards a central unit instead of, as before, separate or independent systems. In the past there were separate computers for position control , i.e. H. one per spatial axis, and another for controlling the thrust regulation of the engine. Today, however, it is already common for everything to run on the same processor. Such a central system offers a number of advantages: It saves weight, and the overall system can take on much more complex tasks, as more data is available to it. As a result, autopilots can already operate the aircraft during the entire flight, i. H. including take-off and landing, fly automatically. One example is the Northrop Grumman X-47 , an experimental aircraft developed for the US Navy . However, no manufacturer has yet dared to introduce such a pilotless system in civil aviation.

Some systems, such as those from the manufacturer Airbus, even have the authority to override commands from the pilot, i. H. to make decisions against the pilot himself in emergency situations. Whether this is actually desirable remains questionable, but a computer reacts faster than a human. On the other hand, experts worry that pilots are increasingly being degraded to observers and should only intervene in an emergency, even though computer technology is not yet perfect.


Autopilots are divided into three categories. A distinction is made between how many axes of the aircraft the autopilot can control. A distinction is made between

Rotation axes and control surfaces
  • single-axis autopilot,
  • two-axis autopilot (with or without height preselection),
  • three-axis autopilot and
  • four-axis autopilot.

With the single-axis autopilot, this only controls the rudder in order to control the vertical axis. This is also known as yaw . Only one is fixed rate ( Engl. Heading ) held.

The two-axis autopilot also addresses the elevator to steer the aircraft around its transverse axis (English pitch ). This means that the altitude can also be controlled in flight.

The three-axis autopilot controls all control surfaces of the aircraft in order to steer it in all three axes. The longitudinal axis (English roll ) is added to the other two axes .

Sometimes the term four-axis autopilot appears in the specialist literature . However, an axis in space is not referred to here, but the thrust vector is counted as the fourth axis. Such systems can specifically control the thrust of the engines in order to maintain a set speed or automatically regulate the thrust during landing.

Single axis autopilot

Fig. 2: Autopilot KAP 140, single axis
Fig. 3: Autopilot KAP 140, single-axis, roll mode activated

The simplest form of the autopilot is the single-axis autopilot, also known as a "wing leveler" (in German: "align wings horizontally"). It only controls the rotation around the longitudinal axis (roll axis) - the rolling . He can keep the wings horizontal by controlling the ailerons. The single axis autopilot only has horizontal mode. The aileron can only be used to change direction, but not to change altitude.

The Bendix / King KAP 140 is a widespread autopilot for small aircraft and twin engines. Its model variant as a single-axis autopilot has five control buttons. The KAP 140 is switched on with the AP button (for: Autopilot). The ROL mode (roll mode) is activated, which keeps the wings in a horizontal position. This prevents the aircraft from tipping over to one side. If the pilot has previously set an appropriate speed and the aircraft is trimmed, it will fly in a very stable state.

Fig. 4: Course gyro - at the bottom right of the button (HDG) labeled in red, the course that the autopilot should keep can be turned. At the moment a course of 325 ° has been turned (red line). There is a current course of 295 °. With the lower left rotary button (PUSH, press first, then turn) the course gyro is regularly readjusted according to the course values ​​on the magnetic compass.

The four other modes to be switched on separately are:

  • HDG (Heading; to fly and maintain a course set on the course gyro)
  • NAV (navigation; to fly and hold a course set on the VOR or GPS)
  • APR (approach; works like NAV mode, but is much more sensitive so that a very precise landing course can be flown)
  • REV (reverse course; works like APR mode, but the autopilot reacts to the deflections of the course needle in exactly the opposite way; many autopilots also refer to this mode as BC - back course, reverse course or reverse course)

The four buttons - HDG, NAV, APR, REV - are activated by pressing and deactivated by pressing again. Only one of the four buttons can be activated at a time. If all four buttons are deactivated, the autopilot falls back into ROL mode - provided the autopilot is switched on. The ROL mode is common to all single-axis autopilot modes.

Two-axis autopilot

Fig. 5: Autopilot KAP 140, two-axis, without height preselection

In addition to the roll axis, two-axis autopilots control the pitch axis, i.e. the pitching around the transverse axis. You can also let the aircraft pitch around the transverse axis (eng. Pitching ) - the nose of the aircraft is raised or lowered. This second axis of the autopilot enables altitude control, the vertical mode. To do this, they control a pitch servo and a pitch trim servo.

Single-axis autopilots can only steer the aircraft sideways (lateral navigation), while dual-axis autopilots can steer the aircraft sideways and upwards (vertical navigation).

In a simple version, the two-axis autopilot is a "wing leveler" with the limited ability to correct the deviations around the transverse axis, i.e. the pitching. At the other end of the product range of two-axis autopilots there are very complex designs that are suitable for automatic flight guidance, with the received signals of the radio navigation receivers on board (VOR, NDB, GPS) being evaluated.

In addition to the four buttons for the horizontal modes (HDG, NAV, APR, REV), the KP 140 has a fifth button for the vertical mode - ALT (Altitude). When the ALT key is activated, the autopilot maintains the current flight altitude.

Pressing the UP button initiates a climb at approximately 500 ft / min, which is maintained as long as the button is pressed. Correspondingly, when the DN button (Down) is pressed, a descent takes place.

Fig. 6: Autopilot KAP 140, two-axis, with height preselection

There are also two-axis autopilots with altitude preselection, with which a flight altitude can be specified. The vertical speed at which the aircraft flies to this altitude can also be adjusted. In approach mode, the autopilot can follow the ILS glide path vertically.

The KAP 140 model with height preselection also has a two-part rotary switch for entering the height and the ARM and BARO buttons.

Since the autopilot does not control the thrust lever on these models, the pilot must continue to do this task. During the approach, he must control the correct approach speed. When climbing he must ensure that the autopilot does not steer the aircraft into a stall at too great a rate of climb. The pilot's control and attention are still required.

Even better equipped autopilots offer an altitude warning in addition to the altitude preselection, which gives an acoustic or optical warning signal for the pilot 1000 or 100 ft before reaching a preset flight altitude.

Development of modern autopilots

The modern systems, which are colloquially referred to as autopilot, are actually a multitude of different, partially independently acting systems within an aircraft. All of these systems together are mostly referred to as autopilot by laypeople, but only one of these combined sub-systems is called autopilot. However, given the complexity of the system and the fact that other subsystems also make autonomous decisions that control the aircraft, this fact is not surprising. The architecture of this system is also not uniform, as each manufacturer sets its own priorities and approaches when implementing avionics systems. “There is sometimes some confusion about the interrelationships between the primary flight controller , the autopilot / flight director system ( AFDS ) and the flight management system ( FMS )”. The primary flight control includes the FBW and all other components that are required to control the aircraft. The Auto Flight System ( AFS ), on the other hand, contains all functions for automatic and autonomous flight control. The FMS is responsible for the navigation and flight performance data.

Auto Flight System (AFS)

Figure 7: The Flight Director (F / D) can be switched on separately, while the autopilot in this model can only be switched on together with the F / D (F / DA / P).

The "Avionics System Autopilot" is now a subsystem of the "Auto Flight System" (AFS), which is referred to in the English literature as "Autopilot" or "Flight Director System" (AFDS). In addition to the autopilot, this system also includes other subsystems. Depending on the manufacturer, however, certain systems are also installed outside the AFS or integrated into other systems. Especially in the English-language specialist literature, mostly only the architecture of the manufacturer Boeing is described, which has a slightly different structure than Airbus.

The AFS is the real heart of autonomous flight in modern machines. Individual systems take on various tasks to keep the aircraft in stable flight or to follow a set course. This system is firmly integrated into the fly-by-wire system. The AFS consists of several subsystems, some of which can act as independent units. These can also be switched on or off by the pilots at will in order to hand over partial tasks to the AFS. The pilot can thus initiate a turn with the aircraft and let the AFS regulate the flight altitude. The pilot no longer needs to worry about maintaining the desired altitude, as the computer does this for him.


Depending on the system that has been integrated, a distinction is made between several types of autopilot (AP). These are categorized according to how many axes they control. The simplest AP regulate the flight attitude only around the roll axis by controlling the ailerons. "These simple systems are often called" wing levelers "". One level up there is AP with two axes, here in addition to the roll axis, the yaw axis is added to the control loop of the AP. These systems can therefore maintain and follow a fixed course. The last category of AP are systems that can control the aircraft around all three axes. This is where the pitch axis comes in, which determines the altitude, as well as the rate of climb and descent. These systems are used in all modern airliners as they enable autonomous landing.

The AP is an essential avionics system as it precisely holds the aircraft in a stable flight position. The system consists of two control loops. The inner loop, inner loop is responsible for a stable flight. The AP computer receives data, in this example from the height sensor. If there is a deviation from the desired flight altitude, the computer controls the actuators of the respective control surface; for example (see picture) the elevator. "Feedback from the actuator ensures that the servomotors reach and hold the desired position". The movement of the respective control surfaces changes the position of the aircraft, which in turn is recorded by the respective sensor, aerodynamic feedback, and passed on again to the AP computer. Manual inputs by the pilot are sent directly to the AP computer and thus overwrite the current system operations, so the pilot can intervene at any time. A schematic structure of the inner control loop can be seen in the figure, but this is not generally valid and depends on the specialist literature used. An example can be seen at Civil Avionics Systems by Ian Moir and Allan Seabridge. The manual control directly accesses the control surfaces past the AP computers. This inner control loop is the same for all three axes, only the controlled surfaces and sensors are different. The second loop is the outer loop . This generates the commands for the inner control loop. Thus, the outer loop is not responsible for a stable flight attitude, but generates the commands that are necessary to control the aircraft in such a way that it follows a desired course or executes the desired maneuver. The necessary calculations that are necessary for this are generated by the Flight Director ( FD ). The AP controller then receives the data from the respective sensors and compares them with the desired ones, as in this example the course. The computer receives data from the current course and compares it with the desired one. If a course error can be recognized, the FD calculates which maneuver is necessary to correct it. The commands required for the maneuver are then forwarded to the AP computer via the controller. At this point, the inner control loop takes over all further commands and addresses the required actuators for the respective control surfaces. In addition to the course, the rate of descent or climb and altitude can also be communicated to the system. These allow the system to keep the aircraft stable in the air around all three axes.

Flight director

Primary Flight Display (PFD) of a Boeing 737 from the next generation series with the flight director as a purple cross

"The Flight Director (FD) is the brain of the autopilot system". Most autopilots can guarantee a stable attitude; however, if other factors such as navigation, wind and course come into play, more complex calculations are necessary. "The FD and AP are designed in such a way that they work very closely together, but it is possible to use the FD without connecting the AP and vice versa". The FD and autopilot are viewed as separate systems, but both systems can be viewed as a single system. When the autopilot is switched off and the FD is activated, symbols on their Primary Flight Display (PFD) show the pilots how to manually control the aircraft in order to follow a desired flight route or maneuver. "It creates a simple, interpretable instruction for the pilot". The desired position of the desired flight attitude is shown schematically on the PFD, usually as a large cross in red or another eye-catching color. So you only have to steer the aircraft so that it is above the position shown. The pilot has to steer the aircraft manually, but this does not make flying any easier, since "the FD has to be told what should happen and the latter then shows how to fly". The FAA therefore recommends deactivating the FD and flying it by instrument flight, as this creates less workload. If both systems are in an aircraft, it is also possible, depending on the system, to activate the AP without FD.

Stability Augmentation System

In modern AP systems, an additional system for a better stable flight attitude is always integrated, the "Stability Augmentation System". This is actually a combination of two autonomous systems, the autotrim system and the yaw damper. Where there used to be two physically separate systems, these are now an integral part of the autopilot.

Autotrim system
Trim tabs

"In order to maintain a flight condition [...] over a longer period of time, there must always be a balance of forces and moments on the aircraft". This equilibrium has to be readjusted continuously during the flight, as the center of gravity changes due to fuel consumption. In order not to constantly compensate for this with manual control inputs, there is the trim . Here, small control surfaces - so-called trim rudders or trim tabs - are angled on the respective rudder. These generate an aerodynamic force in order to restore the desired balance of forces. In order not to have the pilot do this manually over the entire flight time, there is the autotrim system. "The Autotrim system is able to automatically [...] make adjustments to the pitch trim in order to keep the aircraft at the desired altitude [...]". This system is in turn steered and checked by the autopilot as soon as the computer recognizes that a constant deflection of the control surfaces is necessary in order to maintain a desired attitude. This controls the trim tabs to compensate for this and return the control surfaces to their neutral position. This is desirable in order to reduce drag. Aside from trimming via trim tabs, there is also the Airbus weight trimming method. In this case, no trim tabs are adjusted to create a balance of forces, but fuel is pumped from one tank to another in order to shift the aircraft's center of gravity in flight. This has an advantage because there is no additional drag from the trim tabs. This increases the range, but reduces the longitudinal stability , as you take fuel from the wing tanks.

Yaw Damper

The second system for improving stability is the yaw damper or yaw damper . The only task of this system is to dampen the so-called Dutch roll vibration , which occurs particularly in large aircraft. This is generated by gusts that hit the aircraft from the side. Since the tail unit offers a large contact surface, a torque arises around the vertical axis. The subsequent chain of events then creates the “Dutch Roll” vibration. This results in a sine wave around the vertical axis. Every aircraft has its own “Dutch Roll” frequency. If there were no compensation to counteract this, the aircraft would continue to oscillate and also lose altitude. This vibration behavior is particularly uncomfortable for passengers and requires a lot of work from the pilot to compensate for it. The yaw damper thus takes on this task and deflects the rudder in such a way that such a vibration does not even occur. In order to detect this oscillation, the computer receives the yaw rate signals from the aircraft. These are filtered through a band-pass filter in order to identify the aircraft-specific Dutch roll frequency. This also makes it possible to distinguish a curve flight from it. As soon as the computer recognizes that there is a Dutch roll oscillation, it activates the servo motors of the rudder to counteract this. Most of the time, the Yaw Damper is partially or completely integrated in the AP computer. This does not necessarily have to be the case, with the A320 the Yaw Damper is a stand-alone system that is outside the AP system.

Auto throttle system

Another important system of autonomous flight control is the “Auto Throttle System” (ATS) or “Auto Thrust System”. This system controls the engines in such a way that they generate the necessary thrust at any time, which is required for the respective flight position. The specifications for this are generated by the autopilot and passed on to the engines. The ATS offers a multitude of advantages and relieves the pilots, especially during take-offs and landings, where they no longer have to worry about engine performance. Two different systems are still in use today. The classic ATS can still be found in some older machines. The engine is not controlled directly by the AP, but rather it controls a servo motor that mechanically adjusts the thrust levers in the cockpit. The thrust levers thus establish the connection to the respective controls of the engine. These monitor and control all engine processes in order to achieve the required performance. These controllers are highly complex, fine-mechanical computers connected to an electronic controller. Due to their complexity and with increasing digitization, especially through pure FBW systems, these are slowly disappearing. "Modern engines [...] are [today] equipped with a digital electronic engine control system FADEC (full authority digital engine control)". These systems no longer work with the help of mechanical interfaces, but purely digitally. "Your tasks go far beyond the control [...] and always ensure optimal performance or optimal thrust with maximum economy". When using the FADEC system, however, the pilots have no opportunity to intervene in the operating behavior of the engines. Only the input variable of the AP and the thrust lever are used here. The fuel flow is then regulated in a closed control loop via the respective sensors within the engine, which provide data such as temperature distribution, pressure, exhaust gas temperature and speed. Very high demands are placed on these systems because they have to withstand the harsh environment on the engine. Depending on the engine and structure, these even have to withstand temperatures between −60 ° C and 120 ° C. Failure safety is also very important to them, since failure would lead to total engine failure.

Flight Management System

The Flight Management System (FMS) is responsible for an important part of the flight, the navigation. The FMS makes it easier for pilots to plan their flight route and uses several sensors to calculate the current position. In previous systems, the pilot had to enter all the waypoints of his planned flight manually. However, this repeatedly leads to incorrect entries and thus to course deviations. Coupled with an imprecise position determination, this was not advantageous. As early as 1976, visionaries of the avionics industry dreamed of a “master navigation system” that could autonomously navigate the aircraft through all phases of the flight. From 1982 the FMS was introduced. The FMS contains a database with waypoints and procedures that are required to plan a flight route. The pilot then only selects the points that he would like to fly to and thus creates his flight plan. "The computer then calculates the distance and course for each of these points which are on the flight route". With the help of all information "[...] the FMS provides precise navigation between each pair of waypoints during the flight and provides further information about the flight in real time, such as speed over ground, distance, expected flight time, kerosene consumption and maximum time in the air". However, the FMS is not only there to facilitate planning, but also forwards all the necessary information to the AP and the autothrottle system. Here you can see the actual task of the FMS for the AFS. It calculates the necessary course and the altitude that the autopilot must hold in order to follow the desired flight route exactly. As soon as a waypoint has been reached, the AP is informed of a new course which it then follows. Modern systems allow autonomous navigation around all three axes and adaptation of the flight speed. These systems can perform very precise calculations and allow a waypoint to be reached in a very small time window of approx. ± 6 s. They can also control the engines to compensate for any delays. The FMS not only controls navigation, but also sets the necessary receivers in the aircraft to the respective frequencies for communication and beacon frequencies. The FMS consists entirely of two redundant computers that perform all the calculations. With the A320, the FMS is still an independent system with its own computer. But the more fly-by-wire systems and digitalization advances in aircraft, the more individual systems merge. In the A330 and A340 family, the AP and FMS can be found in one system. And in the latest generation, like the A380, all AP systems are housed in the FMS computer and form a whole system.

Flight Envelope Protection

All systems in the AP must of course be monitored, and if abnormal behavior occurs, this must be made known to the pilots so that they can intervene. The Flight Envelope Protection System takes on this task. This system is an integral part of the FBW. It ensures a safe flight in the border areas of the aircraft. This ensures that no structural damage occurs due to excessive acceleration forces. This increases safety during the flight, since errors in a computer ideally cannot lead to a failure of a system or structure.

Landing with autopilot

A landing with autopilot on the runway and the subsequent taxiing on the runway center line is known as a CAT III landing or Autoland. A CAT III landing requires an appropriately equipped and licensed aircraft, a trained and licensed crew and a correspondingly equipped and licensed airfield. With the exception of certain aircraft types, CAT III landings may only be flown with the autopilot due to its approximately four times higher reaction speed. Landings according to CAT IIIa and IIIb are currently possible. In addition to braking on the track, CAT IIIc also includes rolling off the track.

Standard activities of the autopilot

When the aircraft has climbed to the desired altitude after take-off, it goes into horizontal cruise flight.

As long as the aircraft moves in a straight line under constant internal and external conditions (weight distribution, earth's atmosphere , etc.), the flight altitude remains constant. However, just by consuming fuel, the airplane becomes lighter and begins to climb. The barometric altitude measuring system will therefore soon detect a deviation from the preselected altitude. The pitch channel , which regulates the angle of attack , sends a signal to the elevator to adjust the difference until the altitude specified as the setpoint is reached again. Since the lift and therefore the angle of attack can be reduced by the loss of weight and thus the resistance decreases, the flight speed increases, which is why the speed channel (Auto Throttle Computer) readjusts the engine power so that the preselected altitude is maintained at the preselected (optimized) target speed becomes.

The flight direction is regulated via the taxiway . Let us assume that the pilot sets 315 ° as the course, i.e. exactly to the northwest . If the external conditions change, such as the wind direction, the aircraft will drift out of the calculated course if there is no counter-steering. A compass system now measures the deviation from the preselected course and sends a signal to the ailerons to compensate for this  - the aircraft tilts a little (turns sideways around the longitudinal axis). The rudder works like a false keel and the aircraft also turns around the vertical axis until the course is 315 ° again. The rolling channel then moves back to a middle position. During the sideways movement, however, the aircraft had a higher resistance and thus took its nose down - whereupon the pitch channel responded immediately and steered the nose up again. This correction had also caused additional resistance and reduced the speed, which is why the speed controller had to increase the engine output again.

In addition to these standard routines, there are a large number of other control functions that intercept undesired movements and make the flight more comfortable for passengers. The pilots, on the other hand, can devote themselves to their activities in demanding flight phases - such as before landing or when planning changes by air traffic control - without having to constantly readjust the aircraft.


A modern autopilot reads the position from an inertial navigation system (INS) and, if it can be received, from several rotary radio beacons (so-called VOR stations) and, increasingly, from GPS signals. Before the GPS era, only the INS was available, especially over the oceans. As the flight duration progressed, a positional error accumulated in the INS. To reduce errors, the INS had a carousel system that rotated every minute, so that the errors were divided into different direction vectors and thus canceled each other out. The measurement errors of the gyroscopes, which falsify the position data more and more, are called drift. This error is due to the physical characteristics of the system, regardless of whether it is mechanical gyroscopes or laser gyroscopes.

The discrepancies between the two are resolved using digital data processing and a 6-dimensional Kalman filter . The six dimensions are longitude, latitude, altitude, bank angle (roll), pitch angle, and yaw angle.

On many transoceanic flight routes, the INS must guarantee a certain position accuracy (English performance factor ). Therefore, the size of the possible position error is also monitored during flight. The longer the flight, the greater the error that accumulates in the system. Near land, the aircraft position can then be updated with additional information from radio navigation systems (VOR, VOR / DME), outside the range of VORs, the position can be corrected and updated via GPS. The primary system for determining position remains the INS, since it is the safest as an independent on-board system and only depends on its own system (software, power supply). The INS can become more and more imprecise, but it cannot be switched off from the outside, while VOR or GPS can be switched off by their operator.

The INS usually consists of three IRUs (i.e. the gyroscopes), which only require gravity and the earth's rotation as input variables. The initial position of the aircraft must be entered into the system by the pilot.

Computer system

The hardware of an autopilot for a typical large aircraft consists of five 80386 CPUs, with each CPU on its own circuit board . The 80386 CPU is inexpensive, has a sophisticated design and has been thoroughly tested. A real virtual machine can be implemented on the 80386. Newer versions are even radiation-resistant and additionally reinforced for use in aviation. The very old design of the 80386 is deliberately used, as it is reliable and its software behavior has been extensively tested and described.

The customer's operating system provides a virtual machine for each process . The software of the autopilot therefore always controls the electronics of the computer indirectly and never directly, instead it controls the software simulation that runs on the 80386 CPU.

Most serious faulty software operations lead to a system crash of the respective CPU.

Most of the time, there is a low priority process running on each CPU that continuously tests the computer. In principle, every process in the autopilot runs in identical form as a copy at three or more points in different CPUs. The system then decides which of the results to accept. The mean value is adopted, with extremely deviating values ​​being discarded.

With some autopilots, a different design (English design diversity) is added as an additional safety feature. Not only do critical software processes run on different computers, but each computer runs software that was created by different development teams, as it is not very likely that different development teams will make the same mistake. Due to the increase in the complexity of the software and the rising costs for the software, however, many development companies are moving away from this safety precaution through diversity.

State of the art and outlook

With modern “fly-by-wire” systems (Tornado, Airbus, F-16, Eurofighter) the boundaries between the individual systems become blurred. Modern combat aircraft, which are designed to be unstable because of their agility, can no longer be flown without computer support. In addition, the implemented flight control systems enforce compliance with the flight envelope . This largely prevents the physical limits (aerodynamics and structural loads) from being exceeded. The implementation of unmanned aircraft is already a reality today. However, these systems fail because of their still quite limited application possibilities. Special maneuvers (reconnaissance, taking over individual flight phases such as "cruise" or even the fully automatic landing) are already state of the art with the appropriate peripherals.

Autopilots in military aviation

In military aviation, autopilots are used today, which enable interception from an uncontrolled flight condition at the push of a button, as in the Eurofighter Typhoon . This can help prevent some loss of people and material and also increases the chance of survival in a combat situation .

"Automatic terrain tracking" (more often called "Terrain Following Radar" (TFR)) enables low and deep flight.

See also

Individual evidence

  1. a b [1]
  2. Lexicon of Aviation, 3rd edition: Niels Klußman, Armin Malik, Springer Verlag 2012
  3. a b [2]
  4. Dudens
  5. ^ Vocabulary ( Memento from January 18, 2015 in the Internet Archive )
  6. [3]
  7. a b c d e f g h Advanced Avionics : Federal Aviation Administration, US Dep of Transportation 2009, (SG-2), translated from English
  8. a b c d e f g h i j k l m n o p The autopilot, a short introduction: Nikolai Reed, University of Würzburg 2013, PDF , (revised with the consent of the author)
  9. The SARTRE Project ( Memento from November 27, 2010 in the Internet Archive )
  10. World premiere on the US highway: Daimler Trucks brings the first autonomous truck onto public roads  ( page can no longer be accessed , search in web archivesInfo: The link was automatically marked as defective. Please check the link according to the instructions and then remove this notice.@1@ 2Template: Dead Link /  
  11. Your Autopilot has arrived - Tesla Motors Germany . In: . October 14, 2015. Retrieved July 6, 2016.
  12. Because of “Autopilot” advertising: Competition headquarters sued Tesla . ISSN  0174-4909 ( [accessed October 29, 2019]).
  13. a b c Introduction to Avionics System, 2nd edition: Collinson, Kluwer, RPG Academic Publisher 2003
  14. a b c d Aircraft Systems, Mechanical, electrical, and avionics subsystems integration, 3rd edition: Ian Moir, Allan Seabridge, Wiley 2011
  15. ^ A b Aviation History magazine: William Scheck, Aviation History magazine 2003, published online in 2007 , translated from English
  16. ^ Autopilot Systems, An Investigation of the C4I Methodologies Used in Autopilot Systems: Matthew C. Posid (2007)
  17. Radio broadcast England 1947, ( Memento from January 18, 2015 in the Internet Archive )
  18. a b Flight and Aircraft Engineer No 2024 Vol LII, edition of October 9, 1947: G. Geoffrey Smith, London, translated from the English, PDF
  19. Gerald Taufretter: The helplessness of the pilots . In: Der Spiegel . No. 31 , 2009, p. 106-118 ( online - 27 July 2009 ).
  20. a b c d e f g h Civil Avionics System: Ian Moir, Allan Seabridge, Professional Eng. Publishing UK 2003, p. 273.
  21. ^ A b Principles of Avionics Avionics Communications 3rd edition: Albert Helfrick, 2004
  22. a b c d e f g Airplane technology, 5th edition: Klaus Engmann, Vogel-Verlag 2005
  23. a b c Instrument Flying Handbook : FAA, Federal Aviation Administration, US Dep. of Transportation. 2012 (translated from English)
  24. a b D. Briere, P. Traverse: Airbus A320 / A330 / A340 electrical flight controls - A family of fault-tolerant systems . In: IEEE (Ed.): FTCS-23 The Twenty-Third International Symposium on Fault-Tolerant Computing . 1993, ISSN  0731-3071 , doi : 10.1109 / FTCS.1993.627364 (English, ).
  25. see also English Wikipedia

Web links

Commons : Autopilots  - collection of images, videos and audio files
Wiktionary: Autopilot  - explanations of meanings, word origins, synonyms, translations