Who really flies the plane? Why do I meditate? Questions and thoughts I had while on an early Saturday morning flight on an Airbus A321.


  1. Fly-by-wire and flight envelope protection grossly simplified for the scope of this post.
  2. I’m not a cognitive scientist and this is a set of casual observations and opinions based on personal experiences.

The plane that I’m on right now is an A321, a member of the A320 series from Airbus. On this early morning flight, most people are snoozing in the quiet cabin. Considering that I was flying at ~800 km/h with two turbofans spinning thousands of times every second, the sense of tranquility is a bit paradoxical. Trying to follow a morning routine that I've been attempting to establish with a friend, I gently close my eyes and try to meditate. Only to be distracted by a nearby passenger.

Boeing 737
Sunrise at Boston Logan airport. On the adjacent gate, there was a Boeing 737, which is a major competitor to Airbus A320.

One notable fact about A320 is that it was the very first commercial airliner with all-digital fly-by-wire with flight envelope protection system. In the early days of aviation, the controllers (steering wheel and pedals equivalent) were directly wired with the control surfaces (literally with mechanical cables), mechanically transmitting the pilot's commands. On a fly-by-wire plane, the computers first analyze the inputs (either from the pilot or the autopilot), determine which actuators and how much to move for adjustments, and then electronically transmit the commands over the wire—hence the term fly-by-wire. This makes airplane safer and lighter.

In the late 1980s, taking advantage of the computerization, Airbus took a step further with A320 and implemented a flight envelope protection system, which lets computers to override the pilot to keep the airplane within its operating limits (under certain conditions, the pilots can retain manual control). This "guardian angel" system would prevent pilot errors and also improve efficiency.

Suppose you are driving and late to an important presentation that you spent all night staying up to prepare. It's raining outside, and the sun is about to rise. Suddnely, out of no where, a scurry of squirrels appears. It's too close to break in time. Instinctively and mistakenly, you quickly turn the steering wheel, perhaps intending to move to the shoulder of the road. However, with the current speed and the road condition, this would cause a devastating rollover. Now—suppose your car has a computer equivalent to flight envelope protection. The computer knows your intent to make a directional change and various parameters of your car and the road. Then, it would simply execute your command while keeping the car within the operating limits (e.g. staying in the lane and not crashing), saving both the squirrels and yourself.

Airbus' automation philosophy is reflected on their control interface. The yoke (control wheel) of the Boeing (pictured: 777) cockpit resembles typical car steering wheel. On the highly automated modern Airbus (pictured: A380), sidesticks and keyboards replace conventional controls.

Just like our pilot does not have to manually adjust specific configurations of the control surfaces, you and I do not need to command specific muscles to contract and relax. Even without any explicit and specific muscle commands, we can make graceful dance moves, effortlessly pick up an object, make a passionate speech while constructing a sequence of complex facial expressions, or quietly type while sipping in juice on a plane. Beyond the motor control, cognition seems to incorporate such intent-to-command-to-control abstraction and operational limits enveloping. Hypothetically, this would decrease the explicit "cognitive load" and also keep our mind in check, under normal condition.

But what if you need to take an action outside of the envelop? A notable case is China Airlines Flight 006 in 1985. During a routine flight from Taipei to Los Angeles, the Boeing 747SP initially experienced a single engine flame out, a relatively minor emergency given that a 747 has 4 engines. In such situation, without the fly-by-wire and flight envelop protection, pilot has to manually compensate the asymmetric thrust. Being preoccupied with the engine failure, the cockpit crews did not properly handle the imbalance. The plane continued to roll and the 747 suddenly plunged 30,000 ft in just ~2.5 minutes. Thankfully, the cockpit crews were able to regain control and land the plane in San Francisco without any fatality. However, the pilot had to pull a ~5.5 G maneuver, which is about the double the designed structural limit and outside of the flight envelope. This sort of maneuver would not be allowed by the A320 flight envelop protection system.

The incident coincided with the dawn of the (commercial) aviation automation, and many concerned aviators and engineers asked the question of "who has the ultimate control?" Airbus simply responded that A320 "never would have fallen out of the air in the first place: the envelope protection would have automatically kept it in level flight" ("Flying the Electric Skies", Science, 1985). This debate is still on-going and exemplifies the phrase "engineering is an art."

I think undertanding how my mind works could help me to live a fuller life and become a better person. Since the source code and blueprint are not available, I try to casually follow relevent research and books on the matter. The autopilot and envelop protection in cognitive and behavior contexts are extensively covered in popular and well-written books like Malcolm Gladwell's "Blink: The Power of Thinking Without Thinking" and Daniel Kahneman's "Thinking, Fast and Slow". However, this further complicated my problem since I was concerned about constructing a faulty model of myself and making adjustments based on that. Even to oneself, to some extent, one's mind is a blackbox.

Engineers who design the autopilot systems have the same issue since cockpit control is a mixture of human and computer with unique advatanges and limitations from both sides—especially a merky one during a sensor failure or an emergency. Since automation is still done under specific conditions and assumptions, having a misunderstanding or a faulty model (Asiana Airlines Flight 214, Air France Flight 447 ; 99% Invisible has a nice episode on Air France Flight 447) could (and actually have) cause fatal disasters.

Some years ago on an A330, a flight back from Almaty, Kazakhstan. The control surfaces (spoilers, flaps, etc.) are configured for landing.

One confusing aspect of Zen is the striking contrast between the lack of explicit explanations and the rigid ritualistic elements. No one has ever provided me an explicit reason why, and my questions are often returned with inexplicable and esoteric answers like "knowing how to hit the drum." Yes, I have read all the research and science out there, but I still don't know "why" one should medidate.

However, with a little bit of practice under my belt, I discovered some interesting things. Unlike the common misconception, meditation is about neither always staying calm nor a simple de-streesing technique. I also feel like that the objective of meditation would not be the acquirement of an explicit understanding of the whole mind, neither fixing all the problems we have. However, I began to think that it might improve perception of myself and the world. Actually, about a month into daily practicing, one manifestation I noticed is the ever so slightly heightend perception of myself—including some cognitive biases and faults. It doesn't provide me with why and how to fix, but knowng the existence of the issue is a good place to start. I still don't know the obective of meditating, but, at the very minimum, it helps me to continously ask questions and examine myself.

Given the complexity of the mind, I doubt I'll ever "understand" my mind. I doubt meditating is a solution to the sort of questions I ask. Even the flight automation, a tiny subset of mind-machine interaction, is still on the table for debate—after decades of research, discussion, and progress. However, I can see myself continuing meditation for the foreseeable future. As the aformentioned 1989 Science article on aviation automation concludes, it's not about finding the "right" answer but keep asking the questions.

Perhaps one day, once in awhile, I might be able to think outside of the envelope.