The U.S. National Transportation Safety Board advised ALPA earlier this month that, in response to a petition to the Safety Board that ALPA submitted in January 1993, the NTSB has revised its official report on the Jan. 19, 1988, crash of AVAir Flight 3378. The Fairchild Metro III went down shortly after taking off from Raleigh-Durham (N.C.) International Airport at night and in low visibility. Both pilots and all 10 passengers died in the crash, which the NTSB originally blamed on pilot error.

For the ALPA safety representatives who worked so diligently to investigate this vexing mystery, the AVAir 3378 crash was a classic example of an accident laid unfairly to flight crew error when the evidence is much stronger that a sudden malfunction of a critical aircraft system—in the AVAir 3378 case, the stall avoidance system (SAS)—enabled the crash.

The NTSB originally determined that “the probable cause of this accident was the failure of the flight crew to maintain a proper flight path because of the first officer’s inappropriate instrument scan, the captain’s inadequate monitoring of the flight, and the flight crew’s response to a perceived fault in the airplane’s [SAS].”

Contributing to the accident, said the Safety Board, was “the lack of company response to documented indications of difficulties in the first officer’s piloting and inadequate FAA surveillance of AVAir.”

The way the Safety Board saw it, the first officer simply let the airplane roll too far, failed to increase back pressure to compensate for the steep bank angle, and then let the airplane descend.

The NTSB said that the captain, suffering from a lingering cold and perhaps distracted by his pilot-not-flying duties, failed to monitor the copilot’s flying until it was too late. The crew, said the Safety Board, apparently began to recover, but at too low an altitude to avoid striking the ground.

ALPA investigators found several significant errors in the NTSB’s analysis of the accident and in wreckage evidence and results of flight simulator studies that pointed toward a sudden SAS malfunction.

The greatest failing in the NTSB report and in ALPA’s original submission to the Board, the petition asserted, was both organizations’ poor understanding of the SAS computer’s operation. Not until the FAA tested the SAS to respond to NTSB recommendations, made in the original report, did the parties better understand the system.

ALPA, in its petition to the Safety Board, took issue with not only the NTSB’s statement of probable cause, but with 5 of the Safety Board’s 13 findings.

Responding to ALPA’s petition, the NTSB deleted four of the five findings to which ALPA objected, and added a new finding that the AVAir 3378 pilots experienced a loss of situational awareness.

The Safety Board also revised its statement of the probable cause of this accident to be “a failure of the flight crew to maintain a proper flightpath. Contributing to the accident were the ineffective management and supervision of flight crew training and flight operations, and ineffective FAA surveillance of AVAir.”

McVenes says, "The NTSB's response to ALPA's AVAir petition, combined with the Board's recent findings and recommendations in the Pinnacle 3701 accident report, appear to be a very positive trend. I am pleased to see the Board going beyond traditional 'identify the error' investigations to seek the cultural and organizational factors behind human error. Identifying and eliminating these preconditions has great potential for preventing future accidents."