Autopilot — Page 368, Lesson 442

We're moving into the heart of the autopilot now. I want to walk you through how the autopilot actually decides what to do, and that starts with the concept of Outer Loop Control. Think of it this way: the autopilot doesn't just hold the wings level. It takes in a goal — like "hold this altitude" or "turn to this heading" — and then calculates the control inputs needed to achieve that goal. That's the outer loop. The data that defines these goals is what we call Command Modes or Flight Path (Referenced) Modes. Now, where does the pilot tell the autopilot what the goal is? That's the Mode Control Panel, or MCP. It's located on the glare-shield, right in front of you, and it's the pilot's interface with the entire auto-flight system. On a modern twin jet, the MCP lets you engage an autopilot and select any of the pitch and roll outer loop inputs. And importantly, the selector switches have lights built in to show you which autopilot or command modes are currently selected. Here's a critical limitation you must remember: only one command mode may be engaged in a single channel at any one time. That means one mode in roll, and one mode in pitch. For example, it's impossible for the autopilot to hold speed and maintain altitude at the same time by pitching the aeroplane — because both of those would require the same pitch channel. If the aircraft had an autothrottle, though, the autopilot could hold altitude while the autothrottle maintained speed. So the pitch channel does one job, and the throttle does the other. Now, when we feed these raw data inputs relevant to a particular flight path, we call it 'coupling' or a 'mode of operation'. And there are two other terms you'll hear constantly: 'hold' and 'capture'. For instance, if an aircraft is flying automatically at a selected altitude, it's in the 'altitude hold' mode. The term 'capture' relates mainly to modes involving the selection and interception of beams from ground-based radio navigation aids — like 'glide slope capture'. In some cases, mode switching is automatic. So, to switch from intercepting a beam or a heading to tracking the beam once you reach it, a Beam Sensor is installed. This device senses beam deviation and switches modes automatically when the aircraft flies into the beam. Glide slope capture can also happen automatically — in that case, the pitch control channel switches from 'altitude hold' mode to glide slope track when the aircraft flies into the glide slope beam. So where does all this raw data come from? It's supplied from aircraft sensors — attitude, air data, heading, radio, and so on — to the relevant auto-flight computer. That computer compares the data with the selected values on the MCP and computes the control inputs needed to achieve those selected values. Let me give you the full picture of the outer loop inputs on a modern transport aircraft with automatic landing capability. In the Roll Channel, you have: Heading hold, Heading select, VOR intercept and track, LOC intercept and track (that's localizer), and Inertial Nav or L NAV. In the Pitch Channel, you have: Altitude hold, Speed hold, Mach hold, Vertical Speed, and V NAV. So the key takeaway is this: the MCP is your command centre, the outer loop defines the goal, and the auto-flight computer does the comparing and computing. One mode per channel, and the sensors feed the raw data. That's the architecture of how the autopilot flies the aeroplane.