Stop calling it PILOT ERROR
A
Emirates Boeing 777 (EK521), Fly Dubai Boeing 737 (FZ981) and Iceland Air B757 (FI315) all shared the same outcome: a poorly executed go-around in a perfectly good airplane ended in an incident or accident and tragic loss of lives. There was no emergency onboard at the time the go-around was initiated and the meteorological conditions were not worse than simulated during training.
So what went wrong? The proverbial
This begs the question, how can experienced airline pilots flying a state of the art machine make unchecked errors leading to catastrophe? Isn’t it the role of automation and pilot monitoring to prevent or catch errors before it is too late?
To understand and prevent PILOT ERROR, we must stop calling it that. Accidents reports in other industries don’t. They refer to similar circumstances as
Training and procedures are in place to prevent mistakes or catch and correct them before they endanger safety. After all, we have two pilots in the flight deck, three if you count the autopilot. That’s redundancy.
Pilots also train in very stressful simulated emergencies, even simulating the incapacitation of the other pilot. Can you imagine the stress level and workload of a pilot who has his colleague unconscious while the aircraft has lost an engine and the weather is so bad, he doesn’t know if he is going to land or have to go around single pilot?!
So how can a normal go-around with all engines, two pilots and an autopilot go so wrong? It all comes down to HUMANS.
Humans are creatures of habits and biases. Pilots are humans, instructors are humans, managers are humans, regulators are humans; and in my opinion, they have built habits and biases into the training of this simple go-around manoeuvre.
Bias #1: TRAIN EMERGENCIES
Because simulating emergencies in the actual airplane could lead to a very dangerous situation, simulators are the obvious tool to train for emergencies. But when do you train for the non-emergency go-around? Certainly not by practicing it flying the line, as you could fly the whole year without having to go around once.
Bias #2: TEST FOR THE WORST
Regulators want to ensure that crews will be proficient in an emergency. For that reason, mandatory items to be checked are focusing on the worst case scenarios. For go-around, the mandatory check calls for manual flying, with one engine inoperative, after reaching minima. Hardly a “normal” manoeuvre.
Bias #3: SIM IS EXPENSIVE
Managers are cost conscious and simulator sessions are expensive. Instructors have to accomplish a lot in the allocated simulator slot. So mandatory items are trained and tested first. A normal go-around is not mandatory and takes time when time is money.
Bias #4: PILOT INCAPACITATION
The dictionary defines INCAPACITATE as “Prevent from functioning in a normal way (example: he was incapacitated by a heart attack)”.
This is where the bias comes in. In the simulator, we train to recognise a pilot incapacitation when he becomes unresponsive, usually with our colleague acting as if he had a heart attack, slouched on his chair. I would argue however that in two of the three incidents mentioned above both the pilot at the controls and the pilot monitoring were temporarily incapacitated at the same time. The problem was not their hearts, it was their brains, due to the startle effect, the lack of understanding as to what was happening, the loss of situational awareness.
Bias #5: AIRCRAFT PERFORMANCES
Worst case scenario, remember? That means we train go-arounds in the simulator in a very heavy aircraft and one engine out. That means very poor climb performances that test the pilot’s ability to fly a precise airspeed and pitch in order to climb.
Sure, but what happens in the airplane when you go around after a long flight, when you are light and all your engines are developing full thrust? The question becomes not if you can climb, but rather if you can control the climb and stay ahead of performances that you are not used to.
Bias #6: DIFFERENT PROFILES
“Minimum, Minimum” says the synthetic voice indicating that without the runway in sight you need to initiate the go-around. “TOGA, Thrust, Pitch, Flaps” and away from the ground you fly. Easy enough and trained in the simulator over and over again.
What about if you go around sooner than 200ft above ground, like 500ft or 1000ft because you are unstable like Iceland Air? What happens then to your airspeed, your level off altitude, your missed approach routing? Should you even press TOGA and add full thrust? For anyone that has looked closely at the missed approach for ILS 07 in Paris Le Bourget (LFPB), the answer is NO. Yet, do you regularly train in the simulator for it or will you discover what really happens on the day?
Bias #7: MY CONTROLS
“My Controls” is the positive exchange of flight controls from the pilot flying to the pilot monitoring when safety is at stake. Airplanes have two pilots and two sets of controls for that very reason so that the pilot monitoring can take over the controls to ensure safety. There lay two biases: The cockpit gradient and the simulator syndrome.
A Captain flying with an inexperienced First Officer will be more alert, ready and willing to take over the controls from the First Officer. Not so if it is the First Officer that needs to override the Captain. That goes against hierarchy, against culture, against experience.
Similarly, pilots train in the simulator in pairs where both need to be proficient at the controls during emergencies. That means the pilot monitoring might refrain from taking over the controls because it’s not his turn to practice or because his colleague is being examined.
Both these biases create hesitation when prompt action would have been paramount.
Bias #8: HUMAN FACTOR
Pilots train hard so they build reflexes, automatic responses, and a mental library that allows them to stay several steps ahead of the aircraft at all times. When the situation becomes so unfamiliar that the pilot cannot fall back on his training, tunnel vision and brain overload lead to the temporary inability to process and react correctly. A go-around should never go so far down this rabbit hole and yet, in the three cases above, it did.
What’s the solution? Recognising we have these biases and training to palliate them. After all, we are all humans.
————————————————————-
About the author: Nick Sabardin is a corporate pilot, instructor and aviation consultant.