Sunday, April 1, 2007
Safety Update: Winds of Change
HUMAN ERROR. THAT’S THE COMMON EXPLANATION for helicopter accidents. We’ve all heard the statistics. Human error is the cause of 80 percent or more of all accidents.
In a way, though, the problem is in the statistic itself. It’s been held as conventional wisdom so long — perhaps as long as man has found heavier-than-air vehicles to fly, and crash — that "human error" can be a mental stop sign for an accident investigation. Find the fellow who screwed up and your work is done. Why press on? After all, eight out of 10 accidents are attributable to human error.
Imagine the results if that same logic had been applied to the "technical" side of aviation. Engine failure at one time was a common problem. Did accident investigations stop with a finding of "Well, the engine quit. Happens all the time?"
Of course not. Some pilot, operator, investigator, or engineer wanted to know why the engine failed. Then, human nature being what it is, they wondered how they could prevent another engine from failing in the same way. Then they came up with a fix — a procedural change, an inspection, a fuel improvement, or a design refinement.
Over time, such fixes built upon one another and engines became both safer and more efficient. The same could be said for airframes, rotors, transmissions, flight controls, etc., etc. But not for the human element in the aviation operation. For that, we’ve been satisfied with giving a guy "a few days on the beach" or yanking his (or her) pilot’s or mechanic’s license. That is, when it’s necessary. Often, the human error in flying is taken care of if it results in a crash, because the "cause" is eliminated, i.e. the pilot who committed it is killed.
Such an approach takes care of the "cause" of the problem at hand. It does little to prevent the problem from occurring again, and doing so with worse consequences.
Now many say, "Well, you can engineer hardware, but you can’t engineer a human."
Years of research into human factors and aviation psychology might belie that. Research has established that humans under similar conditions will behave in predictable ways. Perhaps not as predictable as hardware, but predictable enough that the consequences of their behavior (which is to say, errors) can be bound by safeguards.
The human element in aviation also has a unique advantage over the hardware ones. It can warn you of weak points and potential failures in the system long before something breaks with a serious consequence (or even a minor one). A highly skilled and trained human, say a pilot, mechanic, or engineer, intuitively senses where an operation falls in the spectrum from smooth-running to SNAFU, and may quickly craft means of working around its weak points to get the job done. Each human in the system is a treasure trove of data on how it works well and poorly, as investigators invariably learn when they go out and interview friends, colleagues and co-workers of those involved after an accident.
Rotary-wing aviation spends hundreds of millions of dollars every year on collecting and analyzing information on the performance of its hardware and software. How much do we invest in effectively tapping that human stream of data, which is very likely richer and more readily accessible than the bits in a new black box?
The odds are that we will invest more and more in that. One reason is the host of domestic and international efforts to improve rotorcraft safety. Those efforts are one reason Eurocopter is developing an inexpensive cockpit information system, using video recorders, for its aircraft that aren’t now required to have cockpit voice and flight data recorders, and why Bell Helicopter is considering doing the same.
Another reason is that it is simply a more efficient way of doing business.
Standard Aero last year began pursuing a program in its Helicopter Business Unit focused on increasing the safety of its products by reducing the potential for human error. Dubbed "Immune," the program is to focus on what the Winnipeg, Manitoba-based maintenance, repair, and overhaul company calls "the softer human and organizational issues related to engine maintenance."
The program’s objective, according to Standard Aero, "is to actively promote a culture where employees feel safe in reporting potential mistakes."
It launched the effort with a full-time project team of mechanics, managers, engineers, and quality specialists. They worked with Gary Eiff, a Purdue University researcher and top expert on human error in aircraft maintenance. They spent months analyzing Standard Aero’s maintenance, quality, and management systems and processes to identify likely areas for human errors. Late last year, the Helicopter Business Unit began reviewing safety defenses, searching for weaknesses in the system of redundancies that lie at the heart of aviation’s impressive safety record. The plan is to work on strengthening those weak points.
"We want to promote a collaborative culture of transparency and full disclosure between our staff and management," said Standard Aero’s vice president of helicopter programs, Tom Roche, in a prepared statement.