Lion Air: A Media Muddle


Although they’re relative rarities these days, as pilots, we all react to airline crashes just as we always have. We’re sober-minded and analytical so we don’t even allow ourselves a conversation inside our heads in speculation about potential causes because we’re waiting for the investigators, right?

That’s ridiculous. Of course we speculate, forming snap opinions that we’re willing to share informally, if not across the cyberverse through thinly informed posts on Facebook or Twitter. Most of the grist for such opinions is supplied by the daily press which, on balance, does a credible job of getting the broad basics right.

Last month’s crash of a Lion Air 737 MAX into the Java Sea off Jakarta has proven to be a struggle for the daily press for what I think are three reasons: First, it’s a foreign crash overseen by a foreign safety agency in a country where press freedom is optional. Second, the airplane was a factory-new 737 MAX 8, Boeing’s latest cutting-edge, fly-by-wire automated marvel into which the company has invested everything it has learned over decades of producing ever safer airliners.

Given current accident rates, it’s not unreasonable to think one would never crash. And the fact that one did just two months and 800-hours into its life cycle invites the irresistible subtext that there must be a flaw in the airplane; undetected Gremlins in the software. Grounding orders to follow.

Last, what reporting has emerged has confronted news writers with a subtlety they’re not necessarily equipped to understand, much less explain to casual readers. Specifically, the fact that the accident aircraft had a recent history of unreliable airspeed indications (UAS)—possibly related to pitot or static faults—and, improbably, a simultaneous angle-of-attack sensor fault. (Maybe.)

Pilots would understand that these are two unrelated systems or related only to the extent that they dump data into an air data computer. But at least some news stories suggested that the angle-of-attack sensor was replaced to correct the UAS indications. Add to this the fact that in the first four or five graphs of every news story, reporters paste in the background that Lion Air has had a checkered safety record.

In its Nov. 9 editions, The New York Times made a game attempt to sort out the technical issues, even without a clear understanding that UAS and AoA are different issues. It ginned up the graphics department to illustrate an AoA vane, but missed that part about flow over the wings. What the story desperately needed was a single, well-crafted paragraph like this: “As the ‘angle of attack’—the wings’ angle relative to the wind—increased, airflow over the wings separated. The wings stopped generating lift, and the airplane rolled to one side.” I didn’t write that paragraph, although I might offer something similar. It was written by Matthew Wald, who used to be the Times’ aviation reporter. In a blog, I once compared his artful explanation of the aerodynamic jujitsu of tail stalls to elucidating brain surgery. He told me he sent it to his mother.

The larger point is a question. Actually, two questions. Can a reader of news be expected to parse this level of detail and should news writers attempt to provide it? Yes and yes. These are not difficult concepts to grasp and it’s reasonable to me that both a graduate of Bryn Mawr and CalPoly working in journalism ought to be able to write about them lucidly. There’s no reason a news writer shouldn’t use pitch instead of “tilt” or relative wind instead of “air currents,” as a Popular Mechanics writer did.

And yeah, I know all about deadline pressure, but that’s what second-day stories and next editions are all about. And rewrites. Everything is online these days and you can rewrite until it’s right, just as E.B. White insisted we all should do.

A few words about the 737 MAX. Earlier, I said it was a fly-by-wire wonder. I lied. Or maybe exaggerated. Over at Aviation Week, my friend Fred George flew the MAX 8 and described it as technology originated in the Space Race and updated with more efficient engines, state-of-the-art-avionics and a tarted-up interior, all of it aimed at what airlines most want: lower operating costs, reduced maintenance cycles and capacity for more butts in the seats.

The MAX is not fly-by-wire in the sense that the Airbus 320neo, which it competes against, is. The MAX has fly-by-wire spoilers, but the rest of the control architecture is not much changed from when the airplane was introduced 40 years ago, which is to say hydraulics. Thus, the chance of bugs in the software rendering it uncontrollable seems remote.

Still, it has a peculiar safety feature that the 737 NG lacked. Media outlets have struggled to understand this or put it into perspective. I thought the Seattle Times Dominic Gates did the best job of it. Based on multiple data streams crunched by the flight computer and fed by its AoA system, the airplane will automatically crank in nose-down trim when the stall angle of attack is detected, or nearly. It the pilot works against that with manual pitch-up command, the trim system will continue to cycle nose-down trim, ever increasing the pitch control forces.

The 737 has a jackscrew-actuated trimable stabilizer with a conventional elevator that’s hydraulically actuated. Boeing and the airlines were fully cognizant that runaway electric trim is a risk and have routinely included procedures to counter it—basically it’s to disable the electric trim and revert to manual trim. It’s a simulator training item. Even ahead of conclusions about Lion Air, the FAA published an emergency AD last week requiring MAX operators to add amplified AoA fault indications to their existing runaway trim procedures. Boeing and the FAA appeared to be reacting to initial flight data that showed the Lion Air flight path with cyclical divergence in speed, altitude and vertical speed that might be consistent with the pilots fighting the airplane’s speed trim system, rather than disengaging it, as they may or may not have been trained to do.

To muddy things further, the airplane’s recent history of UAS may or may not be a factor. It’s not impossible that the crew faced two abnormals—the runaway trim and unreliable airspeed. There are procedures for both. (For UAS, it’s establish 4 degrees pitch and set N1 for 70 percent.)

Unconfused enough yet to form an opinion? I don’t often speculate on such things but if I were a betting man—and I’ve been known to dabble at the tables—my bet is that the accident investigation will once again revisit a familiar theme: automation bias. In the name of safety shaped by consistent procedures driven by automation in the cockpit, pilots are increasingly losing their basic feel for putting hands on controls and throttles and actually flying a flyable airplane. A second focus will be the utter perversity of how pilots now have to be trained to keep the automation that’s supposed to help them from killing them instead.