Why the Latest Air-Scare Shows How Safe Airline Travel Is, Not How Dangerous

By James Fallows

Yesterday I mentioned a widely circulated web account from an author who felt he had narrowly avoided being part of what would have been history's worst airline disaster. Indeed, that was the headline on Kevin Townsend's post on Medium: "Two Weeks Ago I Almost Died in the Deadliest Plane Crash Ever."

While the episodea sudden if brief descent by a United air crew over the Pacific, to be sure it stayed out of the path of another planemight well have been frightening, the hundreds of passengers on the two planes never faced any danger of a mid-air crash. The quick descent indicated the safety of today's air-travel system, not its brink-of-disaster shoddiness.

Think of the analogy of car airbags. When an airbag goes off in a car, I am sure it scares the bejeezus out of anyone on board (I've never experienced it), plus possibly bruising them and, for infants in the front seat, doing real damage. But that detonation, frightening as it is, is part of a system that has made car travel safer rather than more dangerous. Something similar is true with anti-skid braking systems—they can frighten you, but they help protect you. And the same goes for today's aircraft collision-avoidance technology.

I mention this to introduce a note I have gotten from someone with first-hand experience with "TCAS," the automatic collision-avoidance system that ordered the descent on Kevin Townsend's flight:

I work in the aviation industry as an engineer, and have gotten more and more familiar with FAA requirements for aircraft design. I wish the flying public understood how precisely-engineered each piece of critical avionics must be in order to satisfy FAA regulations for a "safety of life" application

TCAS -- and its successor,TCAS-II -- is one such piece of avionics hardware. By all accounts in the Townsend post, TCAS did its job in resolving airspace issues between what is known in FAA jargon as two cooperative aircraft. And the pilots did what they were supposed to do in taking the correct action.

It is vital that pilots trust their instruments because the avionics driving those instruments (and their design requirements) are engineered with that in mind.

Believe me, I know. I've spent the bulk of my career designing, writing software for, and testing the integration of, avionics. If TCAS says "go up", there are thousands of hours of engineering behind that system making sure that "up" is the right decision.


Let me clarify, too, that I am not picking a fight with Kevin Townsend, with whom I've had a pleasant and mutually respectful exchange of messages. He wrote a post on a genuinely frightening experience withoutas he has pointed out to me in emailthe benefit of subsequent info on how far apart the planes had actually been (at least five miles, probably eight) or other technical analyses of what was going on.

I thought the original headline on his item was an enormous reach ("I almost died"), plus the idea that the planes were in "scraping distance" of each other. But, as he has also pointed out to me, if he had foreseen how widely this would be picked up in "peril in the skies" coverage, he would have been more statesmanlike in telling the story. (Plus, he is doing the Lord's work on the filibuster.) This is very different from a flat-out fake air-peril story last year in the NYT Magazine, and another over-hyped one in the same paper.

Why am I going back to this story? The immediate reason is because Townsend's account has generated another flood of email from newly re-frightened fliers. The larger point is one that Patrick Smith has often emphasized at Ask the Pilot. Something deep and primal in human beings, namely the fear of unnaturally being up in the air, easily spills over into something with no rational basis behind it: namely, the belief that airline travel is riskier than normal life, when it fact it is vastly safer than driving, biking, or walking across the street.


Update: Mark Bernstein, chief scientist of Eastgate software (and one-time guest blogger here), writes about an unforgettable part of Kevin Townsend's account. That was during the zero-G descent when he was "weightless and staring downhill at the thirty-some rows of passengers ahead of me." Bernstein writes:

….back of the envelope suggests that a zero-g descent gets you down 600' in just about 6 sec.  And I expect that one would not pull negative G's in a passenger flight, especially without warning, given any alternative. Surely the writer would have remarked on the experience of negative G's with stuff (and people) flying everywhere.  So that's a boundary. 

In fact, I believe we get to the normal descent rate, 1800'/min, with a second of zero g. 

The point here, again, is not to nit-pick the original account but to underscore the difference between subjective experience of frightening events, and the objective reality of what is going on.

A month or two ago, I was flying into a small airport in the South when, at about 500 feet up on the final approach for landing, there was a very strong wind gust from the right. At the time I "felt," and actually told another pilot I saw after I'd landed, that the gust had "almost upended" the plane. I'm sure if there had been a film it would shown nothing remotely close to that. Probably at most a momentary 45-degree bank to the left, which I'd offset within a second or two. So it is with anything involving the unnatural act of human beings up in the air: our senses tell us one thing, and our minds (when they can act calmly) tell us something else. The calm-mind view of air travel underscores its safety, whatever else our senses may tell us.

This article available online at:

http://www.theatlantic.com/technology/archive/2014/05/why-the-latest-air-scare-shows-how-safe-airline-travel-is-not-how-dangerous/371097/