What happens when pilots have to fly without an autopilot?
Quick summary
Last week, the report of an incident involving at 737-500 departing from Madrid highlighted how pilots struggled to fly the aircraft manually when the autopilot system became unavailable. According to investigators, the aircraft departed with one autopilot inoperative, leaving another autopilot system still available for use.
Aircraft are designed to fly safely without a full complement of systems. Most systems will have a backup, and the more important ones will have a backup to the backup. In this situation, pilots and engineers consult the minimum equipment list (MEL) to see which systems we are able to legally depart without.
If all minor faults had to be fixed before every departure, the majority of flights would depart late. By using a MEL dispensation, we are able to reduce delays by fixing minor issues at a later date. The aircraft will continue to fly safely until a scheduled maintenance gap when all the defects will be rectified.
In this situation, the aircraft was able to depart with one of the autopilots unserviceable. The thought process being that should the other autopilot fail, the crew is still able to fly the aircraft safely. As it transpired, this didn't happen.
Shortly after takeoff, the other autopilot failed. According to the official report, the pilots had "problems" flying the aircraft, despite the fact that all flight deck instruments required to fly accurately "were available to the crew at all times".
After some fairly erratic flying and two go-arounds, the aircraft landed safely at Getafe airbase.
So what is the autopilot and how do crews normally use it? How often do we fly without it and what could cause an experienced crew to fly like this?
How do pilots use the autopilot?
"Using the autopilot" is a bit of a vague statement that needs to be clarified. Firstly, we need to ask the question, "What does the autopilot actually do?"
On a modern airliner, there are various levels of automation available to the pilots. Some are emergency systems only activated if the safety of the aircraft becomes in doubt, for example, slow-speed protections. Others are used every day, such as the autothrottle. All these systems are designed to reduce pilot workload, allowing us to dedicate more of our mental capacity to other aspects of the flight, particularly during emergency situations.
The autopilot is merely a servant, it does what it's told to do and no more. As a result, the performance of the autopilot is only as good as the instructions the pilots give it. If a crew instructs the autopilot to fly a heading of 090 degrees at 3,000 feet, the autopilot will do this perfectly... until it flies into a mountain. There are no prizes for this.
Read more: How pilots deal with volcanic ash encounters
For the autopilot to do something useful for us, we must give it something to follow. For the majority of the flight, this will involve following the vertical and lateral elements, which we load into the flight management computer (FMC). On Boeing aircraft, such as the 787 Dreamliner, which I fly, these are known as LNAV (lateral navigation) and VNAV (vertical navigation).
With LNAV and VNAV selected, the autopilot flight director system (AFDS) has some guidance to follow. On our screens, this creates a set of crosshairs known as the flight director. Whenever the FMC requires the aircraft to turn or climb, the flight directors move, directing us where we should point the nose of the aircraft.
This gives us three ways, of flying the aircraft.
Automatic flight -- flight director and autopilot
This is the most common way of flying the aircraft and accounts for around 95% of the flying we do. The FMC is programmed at the gate and the vertical and lateral modes required for departure are selected, giving us flight director guidance. Once airborne, we can engage the autopilot as low as 200 feet (on the 787). This instructs the autopilot to follow the flight director.
With the autopilot engaged and following the flight plan, we have a greater mental capacity to manage other tasks. This is particularly important when it comes to non-normal events, such as problems with the engines or other aircraft systems. The autopilot does the dog's work of keeping the wings level whilst we solve whatever problem has presented itself.
Manual flight – flight director
Once airborne, we do not have to engage the autopilot at 200 feet. If the skies are quiet and the departure fairly straightforward, it's quite nice to actually hand fly the aircraft for a little while. With the flight directors still giving us guidance, we can fly the aircraft manually with the assistance of the automatics.
Flying like this, it's imperative that we follow the flight directors accurately. If we don't, there's the potential confusion between what the FMC is expecting us to do and what we're actually making the aircraft do. For example, when descending, there's a chance that the autothrottle may allow us to fly too slow.
If we are flying manually and want to fly differently to what the flight directors are suggesting, we must turn both flight directors off (captain's and first officer's) and fly the aircraft with just "raw data".
Manual flight – raw data
With the flight directors turned off, there is no guidance linking the FMC to our screens. We are left to fly manually using whatever traditional navigational aids we have tuned, such as VOR beacons and an ILS signal.
We rarely fly like this in the aircraft, as the workload required to fly raw data manually is much higher than with the automatics. The pilot flying the aircraft is focussing solely on the flight path and has little capacity to do anything else.
Read more: How pilots deal with radiation and flights over the North Pole
However, when the weather is good and the overall workload is low, or in the simulator, we do practice flying with raw data. This keeps our skills sharp for the unlikely event that we should lose the assistance of our automatics.
Flying on Instruments
If you're like me, whenever I hear this phrase, I still think of that iconic scene in the movie "Airplane". As much as some of us would love to crack out a double-bass and saxophone to while away those long overnight hours, the reality of instrument flying is quite serious.
When learning to fly, we are taught in two stages: VFR, visual flight rules, and IFR, instrument flight rules.
Visual flight rules
VFR flying is done exactly as it sounds, by looking out of the window and flying visually. Student pilots are taught that "pitch and power equal performance". In simple terms, if you set the pitch (nose angle) of the aircraft on a fixed point out of the window and set a particular engine power, you will get the required performance.
For example, setting the top of the cockpit panel just below the horizon and a certain power setting (for that aircraft type), the aircraft will fly level at a set speed. If you want to climb (or turn, or descend), you set a new pitch and power and it will give you the required performance.
All this is done by looking out of the window, using the horizon as your reference. By having a point to fix your vision on, our bodies are able to compute everything going on around us. We can look ahead to the horizon and see that the ground is below us and the sky is above us. If we're making a turn, we can see that the horizon is at an angle. The senses our body is experiencing are backed up by what our eyes are seeing.
However, what happens when we have to fly through clouds and are unable to see the horizon? Very quickly we become spatially disorientated. Try walking across a field with your eyes open. Then stand on one leg and turn around. Pretty easy. Now do the same with your eyes closed. Chances are you had a major wobble and had to open your eyes to stabilize yourself.
With commercial flying, the weather is rarely benign enough to allow us to fly visually. More often than not, we have to fly through clouds where it's akin to walking with your eyes closed. When flying in clouds, you may as well be in a room with no windows. All you can see ahead is white (or black if it's night time).
What we need is a system to allow us to fly safely whilst we are unable to see out of the window. This is what the instruments are for.
Instrument flight rules
Using data sourced from outside the aircraft, the flight instruments allow us to fly safely even when we can not see outside. The classic suite of instruments includes the attitude indicator (AI), airspeed indicator (ASI), altimeter, vertical speed indicator (VSI) and heading indicator (HI). On light aircraft, you will also have a turn coordinator.
Sensors on the outside of the aircraft detect the speed of the airflow and also the pressure of the air. These are then used to display the airspeed of the aircraft (ASI), the altitude (altimeter) and also how quickly it is climbing (VSI). The basics of these systems are the same whether it's on a two-seat Cessna or a 600 seat A380.
The four most important instruments have traditionally been set up in a "T" formation in the cockpit. The most important dial, the attitude indicator sits top and center. If your pitch is sensible and your wings are level, 90% of the time your flight path is safe.
Read more: How a flying cadet graduates to a fully qualified pilot
From here, the airspeed indicator sits to the left and the altimeter to the right. At the bottom of the 'T' is the heading indicator, creating a convenient flow.
Using the attitude indicator as the center-point, pilots are trained to constantly scan their instruments in a methodical fashion around the ''T", always coming back to the AI.
AI, airspeed, AI, altimeter, AI, heading, AI, repeat. Using the basic principles of "pitch plus power equals performance", the T scan focuses on the pitch and then constantly checks that it is giving the desired performance. If you look at the image of the 787 primary flight display earlier in this article, you'll notice that the information is in exactly the same place, just in digital format.
If the airspeed is too slow, or the altimeter too high, it's most likely because the pitch is wrong. This can be corrected as the eyes come back to the AI.
As all flying instructors tell their students, "trust your instruments". However, in order to be able to successfully trust your instruments, pilots need to be aware of the sensations that may cause their body to disagree with what the instruments are telling them.
Sensory Illusions
Our bodies are complex structures, using a range of external and internal stimuli to work out what's going on around us. Our vestibular system allows us to sense where we are in space, gives us a sense of direction and also gives us our balance. When we are unable to see where we are because of clouds (or because we're walking with our eyes closed), the vestibular system becomes an important part of telling our bodies what is going on.
The problem with the vestibular system is that it is designed to work in a 1G environment. When in an aircraft, the difference in sensations experienced between the vestibular system and the eyes can lead to both the feeling of sickness ad spatial disorientation. This can lead to several illusions.
The leans
The most common illusion experienced by pilots, "the leans" gives a false sense of roll attitude. It is caused when a slow turning roll is initiated, below the sensory threshold for the inner ear. The pilot will feel that the wings are level when in fact they are angled to one side. If the pilot corrects this too abruptly, the inner ear will sense this and give the feeling that the aircraft is turning… even though now it is not.
Read more: Brace for impact! How the landing gear on the 787 Dreamliner works
This sensation may cause a pilot to lean their body towards what they feel the vertical is. If this illusion is not rectified, it could lead to the pilot rolling the aircraft to try and rectify this false sensation of already being in a turn. The more they turn, the worse the effect becomes.
Linear accelerations
The part of the ear that detects linear accelerations is also the same part that detects the attitude of the body. This can be described simply as the "pushed back in your seat" effect. When we advance the engines to take-off power, we all get that feeling of being pushed back in our seats.
However, when pitching the nose up and our bodies are angle back slightly, we get the same sensation. The danger here is that a linear acceleration may trick the mind into thinking that the nose is pitching up. To counter this, the pilot could push forward on the control column. This just makes things worse.
As the nose is pushed forwards, there is a linear acceleration, strengthening that sensation. To counter it, the pilot pushes forwards more. Before they know it, they have entered into a steep dive towards the ground.
The only way to stop these illusions affecting the way we fly the aircraft is to trust what the instruments are telling us. If we feel as if the nose is pitching up but the instruments say the nose is level, then the nose is level.
Bottom line
Modern airliners are complex machines that require a great deal of capacity from the pilots, especially when non-normal situations develop. To help us fly the aircraft safely, various levels of automatic flight are available to us. How much of this we use will depend on the situation at hand.
The instruments in the flight deck enable us to fly safely in all weather conditions, but dangers still lurk. We are all still human and, as a result, are susceptible to the effects of sensory illusions. The only way to overcome these is to trust our instruments and fly the aircraft accordingly.