How Autopilot Works

Automatic pilots, or autopilots, are devices for controlling spacecraft, aircraft, watercraft, missiles and vehicles without constant human intervention. Most people associate autopilots with aircraft, so that's what we'll emphasize in this article. The same principles, however, apply to autopilots that control any kind of vessel.

In the world of aircraft, the autopilot is more accurately described as the automatic flight control system (AFCS). An AFCS is part of an aircraft's avionics -- the electronic systems, equipment and devices used to control key systems of the plane and its flight. In addition to flight control systems, avionics include electronics for communications, navigation, collision avoidance and weather. The original use of an AFCS was to provide pilot relief during tedious stages of flight, such as high-altitude cruising. Advanced autopilots can do much more, carrying out even highly precise maneuvers, such as landing an aircraft in conditions of zero visibility.

Although there is great diversity in autopilot systems, most can be classified according to the number of parts, or surfaces, they control. To understand this discussion, it helps to be familiar with the three basic control surfaces that affect an airplane's attitude. The first are the elevators, which are devices on the tail of a plane that control pitch (the swaying of an aircraft around a horizontal axis perpendicular to the direction of motion). The rudder is also located on the tail of a plane. When the rudder is tilted to starboard (right), the aircraft yaws -- twists on a vertical axis -- in that direction. When the rudder is tilted to port (left), the craft yaws in the opposite direction. Finally, ailerons on the rear edge of each wing roll the plane from side to side.

Autopilots can control any or all of these surfaces. A single-axis autopilot manages just one set of controls, usually the ailerons. This simple type of autopilot is known as a "wing leveler" because, by controlling roll, it keeps the aircraft wings on an even keel. A two-axis autopilot manages elevators and ailerons. Finally, a three-axis autopilot manages all three basic control systems: ailerons, elevators and rudder.

What are the basic parts of an autopilot that enable it to exert control over these surfaces? We'll explore the answer to that question in the next section.

­

The heart of a modern automatic flight control system is a computer with several high-speed processors. To gather the intelligence required to control the plane, the processors communicate with sensors located on the major control surfaces. They can also collect data from other airplane systems and equipment, including gyroscopes, accelerometers, altimeters, compasses and airspeed indicators.

The processors in the AFCS then take the input data and, using complex calculations, compare it to a set of control modes. A control mode is a setting entered by the pilot that defines a specific detail of the flight. For example, there is a control mode that defines how an aircraft's altitude will be maintained. There are also control modes that maintain airspeed, heading and flight path.

These calculations determine if the plane is obeying the commands set up in the control modes. The processors then send signals to various servomechanism units. A servomechanism, or servo for short, is a device that provides mechanical control at a distance. One servo exists for each control surface included in the autopilot system. The servos take the computer's instructions and use motors or hydraulics to move the craft's control surfaces, making sure the plane maintains its proper course and attitude.

The above illustration shows how the basic elements of an autopilot system are related. For simplicity, only one control surface -- the rudder -- is shown, although each control surface would have a similar arrangement. Notice that the basic schematic of an autopilot looks like a loop, with sensors sending data to the autopilot computer, which processes the information and transmits signals to the servo, which moves the control surface, which changes the attitude of the plane, which creates a new data set in the sensors, which starts the whole process again. This type of feedback loop is central to the operation of autopilot systems. It's so important that we're going to examine how feedback loops work in the next section.

An autopilot is an example of a control system. Control systems apply an action based on a measurement and almost always have an impact on the value they are measuring. A classic example of a control system is the negative feedback loop that controls the thermostat in your home. Such a loop works like this:

It's called a negative feedback loop because the result of a certain action (the air conditioning unit clicking on) inhibits further performance of that action. All negative feedback loops require a receptor, a control center and an effector. In the example above, the receptor is the thermometer that measures air temperature. The control center is the processor inside the thermostat. And the effector is the air conditioning unit.

Automated flight control systems work the same way. Let's consider the example of a pilot who has activated a single-axis autopilot -- the so-called wing leveler we mentioned earlier.

This loop, shown above in the block diagram, works continuously, many times a second, much more quickly and smoothly than a human pilot could. Two- and three-axis autopilots obey the same principles, employing multiple processors that control multiple surfaces. Some airplanes even have autothrust computers to control engine thrust. Autopilot and autothrust systems can work together to perform very complex maneuvers.

Autopilots can and do fail. A common problem is some kind of servo failure, either because of a bad motor or a bad connection. A position sensor can also fail, resulting in a loss of input data to the autopilot computer. Fortunately, autopilots for manned aircraft are designed as a failsafe -- that is, no failure in the automatic pilot can prevent effective employment of manual override. To override the autopilot, a crew member simply has to disengage the system, either by flipping a power switch or, if that doesn't work, by pulling the autopilot circuit breaker.

Some airplane crashes have been blamed on situations where pilots have failed to disengage the automatic flight control system. The pilots end up fighting the settings that the autopilot is administering, unable to figure out why the plane won't do what they're asking it to do. This is why flight instruction programs stress practicing for just such a scenario. Pilots must know how to use every feature of an AFCS, but they must also know how to turn it off and fly without it. They also have to adhere to a rigorous maintenance schedule to make sure all sensors and servos are in good working order. Any adjustments or fixes in key systems may require that the autopilot be tweaked. For example, a change made to gyro instruments will require realignment of the settings in the autopilot's computer.

Many modern autopilots can receive data from a Global Positioning System (GPS) receiver installed on the aircraft. A GPS receiver can determine a plane's position in space by calculating its distance from three or more satellites in the GPS network. Armed with such positioning information, an autopilot can do more than keep a plane straight and level -- it can execute a flight plan.

Most commercial jets have had such capabilities for a while, but even ­smaller planes are incorporating sophisticated autopilot systems. New Cessna 182s and 206s are leaving the factory with the Garmin G1000 integrated cockpit, which includes a digital electronic autopilot combined with a flight director. The Garmin G1000 delivers essentially all the capabilities and modes of a jet avionics system, bringing true automatic flight control to a new generation of general aviation planes.

Wiley Post could have only dreamed of such technology back in 1933.

For more information about autopilots, check out the links on the next page.