In 2003, I assisted the Aviation Safety Council (ASC) – the Taiwanese version of the NTSB’s Aeronautical division – during the China Air 611 accident. Taiwan is an incredible place, very beautiful and full of history. I grew to respect the professionals at the ASC as we shared investigatory techniques. Unfortunately Report ASC-AOR-16-06-001, the TransAsia Airways ATR72, flight GE235, has lost something in the translation from Mandarin (Chinese) to English.
I find the world today relies heavily on written interactions, e.g. texting, e-mails, etc. Verbal communications are almost non-existent; we’re losing the ability to translate our thoughts into understandable intentions. If I capitalize all letters, I’m yelling; if I write something sarcastically witty, it’s viewed as patronizing or insulting. The [digital] pen is more powerful than the sword.
My experience with report writing dictates that there is a ‘method’ to the flow of information ‘madness’. Take the ongoing Malaysian Air MH370 disappearance/spectacle: how much opinion and theory were (and still is) flung against the wall in the hopes of something sticking? It was like watching a confused Durga, unsure of which path to point to. Now, how much fact came to the surface and was followed up on? Not so much. There are two reasons for that: one – way too many experts, and, two – one must follow the facts to analyze, not the other way around.
I’m often critical of the NTSB. However, one thing they do right is the blue cover (aviation) report, the final collection of information for an aviation accident report; the blue cover report lays out the facts before engaging in any type of supposition or analysis; it’s like Hercules Poirot explaining the crime to the suspects. It provides that flow of information without the contamination of opinion, allowing the reader to make the connection. Terms like ‘may have’ or ‘probably’ have no place here; more definitive verbs and adverbs are used.
This is even true with the Blue Cover’s Executive Summary; this is the down-and-dirty about the accident, recapping for the Board Members so they can go right to the probable cause(s) and recommendations; Board Members review accidents taking place in five different modes, so it makes sense. It takes up less than one page and no more than four paragraphs.
Report ASC-AOR-16-06-001, by comparison, commits nine pages to the Executive Summary; this is the equivalent of explaining the entire plot of a book on the jacket. TransAsia Airways, flight 235 lost control, crashing into the Keelung River after striking the Huan-Dong overpass. It lost power shortly after take-off, maneuvered through several downtown buildings before rolling over to port and plunging into the river. It was miraculous anyone onboard survived and multiple building occupants weren’t lost on the way to the accident.
In the Executive Summary, report ASC-AOR-16-06-001 contains numerous assumptions on the part of the investigator, e.g. ‘Had the crew prioritized their actions to stabilize the aircraft …’ or ‘the pilots did not respond to stall warnings in a timely manner’. This takes the report out of context. Many allegations are presumptuous unless evidence is provided to lead the reader there, e.g. following the cockpit voice recorder transcript (CVR-T) later in the report I was unsure as to how these conclusions were made – the factual section starts 16 pages after, while the analysis comes in at 140 pages in.
One unsettling assumption in the report states, ‘the pilot reduces power and then shuts down the wrong engine.’ I read the attached CVR-T several times; unless the conversations were incorrectly translated from Mandarin, there’s no mention of the #1 engine being shut down in error. To further the point, the crew ‘sound’ calm and in control; there is no fighting or lack of situational awareness in the CVR-T transcript.
Now to be fair, the first section of the report is called Findings as a Result of this Investigation; it lays out all the findings in three topics: powerplant, risk and ‘other findings’’. What it doesn’t do is bring the reader to the end of the report; it’s like a mystery’s first chapter describing the crime and who did it. Furthermore, it’s inaccurate and presumptuous; the investigator ‘reads’ the pilots’ collective minds, knowing their thoughts during the harrowing last minutes of flight.
An accident report is supposed to flow from fact to fact to fact, building upon the narrative; the writer is meant to lay out the events, assuming that the readers don’t know the chain of events … because they don’t.
In the Probable Cause section of the Executive Summary, the ninth probable cause states, ‘An intermittent signal discontinuity between the autofeather unit number 2 and the torque sensor may have caused … number 2 autofeathering.’ May have? I never worked the ATR so as I understand it there are two probabilities: One – the automatic takeoff power control system (ATCPS) became unarmed during takeoff roll; and, two – the ATCPS being activated during climb. This is an assumption – not based in fact (may have) – and a leap. This statement is found on page three of the Executive Summary; I cannot reconcile how the assumption was drawn.
Again, it may be due to translation, but a confusing aspect of the report fails to establish who’s who. The acronym ‘PF’ means the ‘pilot flying’, yet never defines which pilot is flying the leg or if the left-seater becomes the PF in the emergency. One pilot is referred to as ‘Captain A’, instead of Captain or First Officer; this may be a cultural misinterpretation on my part, but it doesn’t contribute to understanding the events.
Finally, the report mentions a probable cause that the crew ignored standard operating procedures for engine flameout, while also stating the airline’s training and procedures were inadequate. Again, perhaps cultural, but are they holding the crew accountable for what they aren’t provided?
I know a lot of hard work went into this report; I also don’t know the efforts required to translate the report into English. Unfortunately for me, it raised more questions than answers.
