Autoland — Page 381, Lesson 465

I want to walk you through the autoland system now. This is the part of your training where the aircraft itself takes over the most demanding phase of flight — the approach and landing. Let me start with the introduction, because it sets the stage for why autoland exists at all. The approach and landing manoeuvre is the most difficult one demanded of a pilot. Why? Because it entails control of the aircraft in all three axes simultaneously — that's pitch, roll, and yaw — as well as control of airspeed through engine power changes. So you're juggling three-dimensional attitude and speed all at once. Let me break down exactly what the pilot has to do. First, align the aircraft with the runway centreline. Second, achieve a sink rate of about 2 feet per second before touchdown — that's the vertical descent rate you want at the moment of landing. Third, reduce the airspeed from 1.3VS to about 1.15VS by progressive reduction of engine power. VS here is the stall speed, so 1.3VS is the typical approach speed, and you're bleeding it down to about 1.15VS as you flare. Fourth, level the wings before actual landing, and yaw the aircraft to remove any drift angle — this is what we call the drift "kick off" or de-crabbing. That's the manoeuvre where you align the nose with the runway just before touchdown to cancel out any sideways drift from a crosswind. Now here's the key point. An automatic landing system that takes over from a pilot must be able to provide guidance and control better than that required of the pilot. So the machine has to be at least as good as a skilled human — actually better. Let me give you some history and context. Autopilots have for a long time now been able to fly most of the approach, allowing the pilot to concentrate on navigating the approach correctly. The pilot would then take over at decision height and continue to land manually. But here's an important distinction. Aircraft that are fitted with all the equipment required for a fully automatic landing may, due to lack of required ground equipment or simply for pilot experience requirements, only be allowed to carry out an auto-approach. Essentially all the procedures are carried out as for an autoland, but when decision height is reached the pilot will take over manually. So auto-approach is autoland procedure, but with a manual takeover at decision height. Now let's move to the autoland system itself, starting with its objective. In order to achieve the objective of automatic landing, the operation of an automatic flight control system must be of such a nature that it will do four things. First, not disturb the flight path as a result of an active malfunction. Second, have adequate authority for sufficiently accurate control along the required flight path. Third, warn of a passive failure. And fourth, allow the intended flight manoeuvre to be completed following an active or a passive failure. Let me unpack those terms because they're crucial. An active malfunction is one that actually disturbs the aircraft's path — it actively pushes the aircraft off course. The system must not let that happen. A passive failure is one that doesn't disturb the path but means the system has lost some capability — and the system must warn you about it. And in either case, active or passive, the system must still allow the intended flight manoeuvre to be completed. So the design philosophy is: never let a failure ruin the landing, and always tell the pilot when something's degraded. Now, the requirements. To enable an aeroplane to complete an automatic landing, the autoland system requires three things. A minimum of two independent autopilots capable of following ILS signals — ILS being the Instrument Landing System. Two independent radio altimeters to give accurate height from the ground information. And a Category 3 ILS ground installation at the airport. So you need redundancy in the autopilots, redundancy in the height sensing, and a very high-grade ground facility. Finally, let's touch on autoland status. The number of autopilots required also depends upon the autoland status of the aircraft. These fall into two main categories — and that's where we'll pick up next, because the excerpt cuts off right there. But the core idea to hold onto is this: the more demanding the autoland operation, the more autopilots you need, and the system is categorised accordingly. Let me make sure the picture is complete. You've got the pilot's four tasks — centreline alignment, 2 feet per second sink rate, speed reduction from 1.3VS to 1.15VS, and the de-crab. You've got the four system objectives — no path disturbance from active failure, adequate control authority, passive failure warning, and completion of the manoeuvre after any failure. And you've got the three hardware requirements — two independent autopilots, two independent radio altimeters, and a Cat 3 ILS ground installation. That's the foundation of autoland.