I thought by now pilots—especially professional pilots—would know not to troubleshoot equipment problems in the air. Especially significant equipment problems. Everyone should know by now that the best place to troubleshoot is on the ground and when significant failures occur, land the aircraft as soon as possible. Recent events suggest this is a lesson worth repeating.

Reading the preliminary accident report of the Lion Air crash on October 29, 2018, I was shocked that the pilots on the flight before the fatal flight lost significant instruments in flight and yet did not make the decision to land immediately. While the cockpit voice recorder of the fatal flight has been recovered, its contents have not been released. However, the air traffic control tapes are available both for the accident flight and the earlier flight in the same Boeing 737 Max. Investigators are questioning whether the problems experienced on the fatal flight were related to problems that occurred on earlier flights and whether the problems were properly addressed by maintenance when the aircraft was released for service on its fatal journey. For that reason, the preliminary report has more detailed information on earlier flights than one would typically find.

Whatever happened to cause the October 29 Lion Air crash is still the subject of intense investigation and analysis by the Indonesian government, with the assistance of the U.S. NTSB and the accident investigation bodies of Australia and Singapore. That investigation will probably take quite a few more months, if not a year or more. Complex investigations can take a long time to get them right.

The preliminary accident report highlights actions by the earlier flight crew that are worthy of note for all pilots, whether airline, corporate, or GA. According to the report, the Lion Air crew on the October 28 flight (the day before the fatal crash) was advised that maintenance had been performed on the angle-of-attack sensor and that it had been replaced and tested. The flight took off from Denpasar, Indonesia, en route to Jakarta. The report states that while the crew reported a normal takeoff, about two seconds after gear retraction problems began. Within five minutes of takeoff the following was reported:

About two seconds after landing gear retraction, the Takeoff Configuration Warning appeared then extinguished. About 400 feet, the PIC noticed on the primary flight display (PFD) that the IAS [indicated air speed] DISAGREE warning appeared and the stick shaker activated. The FDR [flight data recorder] showed the stick shaker activated during the rotation. Following that indication, the PIC maintained a pitch of 15 degrees and the existing takeoff thrust setting. The stick shaker remained active throughout the flight.

The PIC realized that as soon as the SIC stopped trim input, the aircraft was automatically trimming aircraft nose down. After this happened three times, the captain declared “PAN-PAN to the Denpasar Approach controller due to instrument failure.”  Pan pan is an international term that indicates an urgent situation but not one requiring immediate assistance (as a mayday call would.) The air traffic controller asked if the crew wanted to return to land. At this point, the crew was just five minutes from its departure airport. Although a significant instrument had failed in-flight, the crew chose to attempt to fix it in flight and declined to return to land.

This was clearly not the correct decision and likely violated Indonesian regulations, which—like the Federal Aviation Regulations—require that a pilot-in-command discontinue a flight when an unairworthy condition arises. One of the safety recommendations in the preliminary accident report addresses this very problem, recommending that Lion Air remind its crews:

[Indonesian Civil Aviation Safety Regulations require that] the pilot-in-command shall discontinue the flight when unairworthy mechanical, electrical, or structural conditions occur. The flight from Denpasar to Jakarta experienced stick shaker activation during the takeoff rotation and remained active throughout the flight. This condition is considered an un-airworthy condition and the flight shall not be continued.

So while the pilots decided to continue the flight, believing apparently that the aircraft’s problems had been corrected (although the stick shaker remained on), 16 minutes later the crew again declared pan pan, telling air traffic control that the flight experienced instrument failure involving altitude and autopilot. Once again, the crew determined to keep flying.

Fortunately for this crew and its passengers, the aircraft landed safely in Jakarta an hour later. I know many of you may be shaking your heads thinking this could happen only in a developing country. Or with a low-cost carrier. Or only with a low-cost carrier in a developing country. But I’ve seen this same in-air troubleshooting happen with aircraft and crews of all types, including major airlines in the United States.

The crash that immediately comes to my mind because it happened while I was a Member of the National Transportation Safety Board is the Alaska Airlines accident on Jan. 31, 2000 off the coast of California. The NTSB determined that the probable cause of the accident was “a loss of airplane pitch control resulting from the in-flight failure of the horizontal stabilizer trim system jackscrew assembly’s acme nut threads. The thread failure was caused by excessive wear resulting from Alaska Airlines' insufficient lubrication of the jackscrew assembly.” 

The accident sent the MD-83 crashing into the Pacific Ocean, killing all 5 crewmembers and 83 passengers. For at least 20 minutes before the crash, the crew and company mechanics on the ground attempted to troubleshoot the problem with the horizontal stabilizer. During this troubleshooting, the aircraft overflew airports where an emergency landing could have been made. 

The October 28 Lion Air crew and passengers got lucky that their in-air troubleshooting did not result in disaster. There’s no room in aviation for relying on luck. If there’s an airworthiness problem that requires troubleshooting, find a suitable place to land first.