This much can be said with confidence: Attitude instruments do not fail often. But about twice a year their failures do lead to accidents. In instrument meteorological conditions (IMC), the airplane kill-rate for attitude director indicator (ADI) failure is about 75 percent. Straight after takeoff, it’s probably more like 100 percent.
A simple step might yield a great advance in safety: Reposition the standby ADI so that it is right next to the primary ADI, preferably just above it. The concept is dubbed twinning, and it sticks the standby ADI right in the pilot’s primary field of view.
It takes only seconds for a heavily laden jet to reach an unrecoverable attitude while rolling into a post-takeoff departure turn. That situation alone points to the advantage of twinning. Ideally (and inexpensively) both pilots should have their standby instruments–the so-called "peanut gyros"–in closest proximity to their main attitude reference. The matching roll rates would be reassuring to pilots, who are aware that the paired instruments take their attitude reference from different sources. The arrangement can also mitigate a delayed reaction to an upset. The pilot not flying–preoccupied with radio calls, punching buttons for the flight management system, and whatnot–may only look up when he feels the Gs start to come on. And that may be too late.
To be sure, the source data feeding the primary ADI can be switched from, say, a failed or faulty inertial reference unit to another. But for a pilot in extremis, manual switching of the ADI source data is likely to be the last thing occurring to him. Moreover, once a dynamic upset is under way, cross-checking the ever-changing attitude on three instruments–the two pilots’ primary ADIs and the standby indicator located typically between them, near the center of the instrument panel–becomes difficult.
It seems straightforward for the pilot flying to compare the display on his ADI with that of the pilot not flying and then with the standby ADI. If his indicator is out of step with the other two, well, there’s the culprit. A comparator warning, usually triggered when the two primary ADIs fall out of synch by a few degrees, should further "incentivize" the cross-checking. If the pilot flying recognizes that his ADI is faulty, control quickly can be handed to the copilot.
Yet it didn’t happen that way in two strikingly similar accidents. In both cases the captains flew their faulty ADIs right into the ground or the water. Despite cautionary utterances by other crewmembers in the cockpit, the pilots remained fixated on their faulty ADIs. The standby instrument was just outside that tunnel of intense focus, and quite out of mind.
The Dec. 22, 1999, crash of a Korean Air Lines B747 freighter shortly after takeoff from London’s Stansted airport is the most recent case in point (September 2003). It recalls the 1978 loss of an Air India B747 with all 213 aboard shortly after takeoff from Bombay. Following a right turn, the Air India aircraft rolled left, beyond 90 degrees, and crashed into shallow water about two miles offshore. The probable cause was attributed to a malfunctioning ADI, which led to the captain’s complete loss of situational awareness. Both accidents involved night flights. Both aircraft rolled left after takeoff.
In these two cases, once an upset was under way, there was no real chance to cross-compare the ever-changing attitudes on the three instruments across the cockpit. EICAS (engine indication and crew alerting system) warnings, often associated with ADI failures, can be a clincher distraction. Even if an OFF flag is present on the ADI display, it can remain unseen–not to mention that ADIs have been known to freeze-fail without OFF flags. Furthermore, there is a known reticence for captains to relinquish control once the situation has gone beyond the marginally precarious and verges on disastrous.
Without twinning, the pilot may only become aware that something is wrong once the airspeed starts winding up and the ambient noise level increases. At that point, a successful recovery at low altitude is highly unlikely.
Further assurance of maintaining proper attitude can come from dedicated power sources. In the last five minutes of the 1998 tragedy of Swissair Flight 111, with a raging electrically stoked fire spreading from the attic space in the cockpit, the two pilots had no attitude reference at all–not even a turn needle or turn coordinator to fall back on. They saw inky blackness on the outside and black primary flight displays [PDFs] inside the cockpit. The standby ADI, even if it had been working, was located near the bottom center of the instrument panel. It was poorly lit and even more difficult to see while wearing smoke masks.
The Swissair jet’s standby was lost because it had been hooked to a vulnerable hot battery bus. When it failed, the powerless gyro started to spin down. Therein lies a vital object lesson in redundancy: Standby instruments need their own batteries, so that a loss of generated power doesn’t take down the sole fallback attitude reference.
The Transportation Safety Board of Canada, which investigated the crash, also commented on the standby instruments’ positioning in a safety advisory issued Sept. 28, 2001. Notably, the TSB said standby instruments should be in the "normal line of vision," lest the pilot become disoriented.
Regulations do not dictate where standby instruments must be located. One veteran airline pilot groused that standby instruments are spread across the cockpit, varying from aircraft model to aircraft model. But another pilot objects that twinning the primary and standby ADI can degrade the traditional, selective radial scan of instruments in the T-panel layout of the modern cockpit. The counter argument is that the traditional "T" is being increasingly "invaded" by all manner of other gauges, switches and subpanels. Juxtaposing a little 3.125-inch (80-mm) standby ADI in the near vicinity of the main instrument is simply a matter of design.
The twinning concept might be even more important for smaller aircraft. Members of the Oklahoma State University basketball team killed in the Jan. 27, 2001, nighttime crash of a Beech 200 in Colorado might be alive today if the aircraft had been equipped with twinned standby indicators. The pilot lost power to the primary ADI and could have used the remaining ADI on the copilot’s side. (The copilot in this case was not qualified on the B200 and basically was along for the ride in an airplane certified for single-pilot operation.)
Again, looking across the cockpit at night, with a massive electrical system failure, is just asking for the kind of "graveyard spiral" the ground radar captured of this airplane’s last moments–suggesting spatial disorientation.
A 1985 paper by John Dow is prophetically instructive. Now retired, Dow at the time was an official in the Federal Aviation Administration’s small airplane directorate. He argued that loss of attitude information can be worse than loss of an engine.
"Attitude reference, in my opinion, is more critical than power in a single-engine aircraft," Dow wrote. "An aircraft may successfully land without power after a controlled descent through IMC, [but] an aircraft cannot successfully land…after in-flight separation of the airframe [due to extreme G forces].
"Loss of power dictates an emergency condition with some time available to recover or make an emergency landing, [while] loss of attitude reference easily may proceed to an irrecoverable condition in a short period of time. Further, engine power loss is readily detectable, [while] vacuum source [ADI] failure may not be readily detectable until total loss of control…or ground contact occurs."
With twinning, any ADI fault is far more likely to be noticed immediately because the standby instrument is in the pilot’s direct view. There may be no worse feeling that having a bad case of vertigo–and no reassuring gyro to stare at.
It may be a challenge to adopt twinning as a standard. But the record suggests it would save lives.
The Concept Defined
The philosophy of closely juxtaposing critical flight instruments is based on the premise that they are essentially there to back each other up. Examples include the altimeter and standby altimeter, and also the main attitude director indicator (ADI) and its backup/standby attitude instrument.
The Transportation Safety Board of Canada on standby instruments:
"When pilots are forced to use standby instruments, they must be able to quickly adjust their instrument cross-check scan. The transition to standby instruments, especially in adverse circumstances, could be significantly hampered by having the instruments positioned away from the normal line of vision and by not having them in a standard grouping layout. The result could be disorientation of the flight crew and loss of control of the aircraft."
This comment comes from TSB Aviation Safety Advisory A010042-1, Sept. 28, 2002.
David Evans may be reached by e-mail at [email protected].