Black Swans and Checklists

June 25, 2014 by Bruce Landsberg

By HeartThe loss of a Gulfstream IV several weeks ago with all on board is tragic, and the cause seems obvious and yet a mystery. From what we now know or think we know, it appears that the big jet accelerated down the runway and reached rotation speed when the crew belatedly discovered that the flight controls were locked. There was no escape—no way to fly or to stop. They skidded for over half a mile on the remaining runway, into the overrun, through the localizer antenna, breached the airport fence, and down into a ravine where the G-IV broke up and burned.

The mantra about flight controls being ” Free and correct” MUST be done prior to every takeoff in every aircraft. It’s a killer item. So how could a professional crew with so much experience miss this most basic of before-takeoff checks? A friend brought up a key point point—that most factory checklists are absurdly long and too many pilots ignore them or significant parts of them. They are written to prevent lawsuits, not to help pilots prepare for flight.

With my usual caveat about speculation so early in an investigation, part of the answer may be perfectly obvious—complacency and/or distraction. It is present in almost every accident involving experienced pilots. We become complacent because we’ve seen or done this many times before and it’s always worked. One should never get too comfortable in an aircraft, which is never a totally benign environment.

Distraction means not putting first things first. Humans are no good at multi-tasking—it’s amazing that job-seekers still think this is a good buzzword to put on their resumes. In aviation, as in business, deal with the nearest biggest alligator first. If the first one gets you, everything else is irrelevant! Shorter, more relevant checklists perhaps?

Now to the mystery part. The G-IV designers, anticipating that humans make the most basic of errors, added a thrust lever interlock that would prevent engine thrust from being increased beyond taxi speed if the gust lock was engaged. That should have prevented takeoff power from being applied. Was there a “black swan event” (a one in a gazillion chance) that the interlock failed at the same time the crew failed to check the controls? Or, did the gust lock release mechanism fail to release the controls while releasing the thrust lever interlock?

How would a crew know if the interlock failed in routine operation because they always checked the controls and everything worked normally? Suppose a key human factors device became inoperative? And in the one in a gazillion times that the crew failed to verify flight controls “free and correct,” the safety backup would not be there to save them.

A good way to check flight controls: “Box them.” That means to move the yoke or stick all the way left, then pull it all the way aft, then all the way to the right, then full forward and finally back to neutral. Of course, you could go clockwise instead, if you’re of that political persuasion. Full control movement is needed, and please actually look to see that the controls did as commanded. There have been dyslexic mechanics who reversed rigged the system, and that will really mess your mind on takeoff. Let’s pare checklists down to essential items only and make it easy to check—there’s more to learn after this accident.

Bruce Landsberg
President, AOPA Foundation

ASI Online Safety Courses  |  ASI Safety Quiz  |  Support the AOPA Foundation

17 Responses to “Black Swans and Checklists”

  1. Duane Says:

    From what has been reported, the CVR playback verified that nobody on the flight deck checked or reported on the status of the flight controls. And apparently the flight controls weren’t free. It may take some time for the investigation to be completed as far as the “black swan” failure of the interlocks, if any … but if these early reports are fully true, then this is about as clear cut a case of pilot error as can ever happen.

    As we all know, all flight students are taught from the very first flight lesson to always check the flight controls before takeoff, verified free and clear and correctly operable. I have made many piloting errors over the years, but this is one I never made because it is unimaginable. As a pilot, I simply must KNOW beyond any doubt that I can control any aircraft before I will ever start a takeoff roll. And I have never witnessed another pilot NOT do it in many years of flying and riding up front.

    Yet obviously at least some pilots don’t do it.

    Complacency is one thing, but that’s not complacency … that’s just bad piloting, period. It’s not even a near call.

    There are many pilot error-induced aviation accidents that we read about that, and think … “there but for the grace of God go I” – knowing that we all are humanly capable of making the same or a similar error.

    But this failure just seems incredible – as in, unthinkable. Failing to verify the operation of the flight controls before takeoff is the aviation equivalent of playing Russian roulette. Except it’s worse – other lives are depending upon the outcome.

  2. Andy K Says:

    I am of the opinion that any time an aircraft is returned to service immediately after an annual the flight controls should be checked on the ramp after (before ?) engine start. I would do this as well as on the run up area (checklist). No sense going all the way to run-up with the ailerons reversed. It’s happened.
    As I am a student I am open to comments on this. The people I fly with do this every time as the cockpits we have are quite cramped and there’s not a lot of room for six-footers.

  3. Johnny nWhite Says:

    There is two very important parts to being an Aviator that is not being taught and are really discouraged by too much emphasis on using a Check List as a “Do List”
    1. Airmanship seems to be a lost art. There are certain things a Pilot should just know. Are you flying a carburetor or fuel injection? Is it cold or hot outside? Precip and what kind. You can’t fly and airplane safely using a checklist or decision tree to cope with every possible issue. Long checklist used as do lists make way more mistakes than good Flows, Habits and Airmanship. Skip on line on a “DO LIST” and it might be fatal. Check lists are called that to check that you the Pilot have done what you should have and need only the very most important items that are critical too flight.
    2. Habits are of utmost importance. Example: Get in cockpit then first thing is seat adjustment and seat belt, set trim to Take Off band. Everytime without fail and let no distraction keep you from it. Get the picture?

  4. Luca Says:

    We can see two points in this story. The first is obviously the human error that caused the mishap. There is very little we can do, except stressing the importance of checking the controls before flight!
    The second point, mentioned but not fully developed in the article, is the absurdity of long check lists. You could take your time to follow through a long check list before takeoff, but what about a 2 pages ‘downwind’ checklist, or a 3 pages ‘emergency’ checklist? No way the pilot could use it…. And this is especially true of complex airplanes: the list grows longer as the airplane flies faster. One could fly a ten miles downwind, or else find a different approach.
    And this is one of the best lessons I ever got in my flying experience. When starting to fly the Siai Marchetti SF260, I was confronted with exactly this kind of ‘to avoid lawsuits’ checklist. My instructor, an oldtimer on the SF260 and experienced aerobatic master, opened up his briefcase and pulled out his ‘secret’ SF260 checklist: one page was covering every phase of flight.
    It was aptly named ‘To avoid death’….
    It only included the items that could kill you (flight controls, fuel, gear….)
    It left out obvious things that any pilot should know (engine settings, alternators, attitude, altitude, non essential instruments….)
    Of course the original checklist was always available, for legal reasons and to be able to pull it out in case of doubt. But this has never been the case!
    This shortened checklist has allowed me, and countless other pilots who have had the benefits of his instruction, to enjoy a much safer and pleasurable flight.
    After a few years I progressed to own an Aerostar, which had the same problematic checklist. Here I did the job myself, to re-write the checklist in order to get it down to one page. The process itself of thinking about every single item in the original checklist, in order to decide whether to leave it or to strike it off, can give you an immense insight into the significance of every action you as a pilot will take during flight, and eventually make you a better and safer pilot.
    After about 1,200 hours in the Aerostar I can testify to the effectiveness of the procedure.
    From my personal experience, I do think this process could benefit any pilot out there.

  5. Chris Says:

    I concur with the above, I like abbreviated checklists. I also believe that a very high percentage of aviation accidents are due to failure of the pilot or crew to check the basics. None of us want to be involved in an accident, yet I still see pilots depart with just enough fuel to go around the pattern once, or without checking a thing prior to departure. Ugh.

  6. Manuel Says:

    The black swan can take many forms. A rumor has circulated in the pilot community that the Gulfstream owner, Lewis Katz, was usually in a great hurry. Observers at the FBO report that the aircraft departed in a rush. Of course, none of this suspends the flight crew’s responsibility to check freedom of the flight controls.

  7. Dan Says:

    I too have NEVER (to my recollection) failed to verify “flight controls free and correct”. I would have to admit to taking other (stupid) short cuts, but never that one.

    I would like to address Bruce’s comment about how checklists seem to be more for the attorneys than the pilots; and I agree. This is an issue that pervades our entire society and I believe it needs to be addressed. Who among us have actually READ that two page small print form we have to sign every time we go to the doctors office? Very few. If we refused to sign we wouldn’t get treatment, so what’s the point? The point is to protect the doctors & healthcare facilities from lawsuits. Who here reads the end user license agreement prior to installing software? Almost no one, because if we want to use it we have to accept, so what’s the point?

    It’s probably the same with aircraft checklists. I fly a Cirrus so I can’t speak from personal experience as I don’t believe the Cirrus pre-takeoff checklist is burdensome. But I confess that my preflight checklist (prior to starting engine) is HUGELY abbreviated compared to the published list. I have no experience flying something sexy like a G-IV but I have to imagine the checklists in that thing must be pages and pages in length — mostly to keep someone out of court, or more specifically, to protect them once they are in court. Instead we should adopt a policy of personal responsibility and accountability rather than blaming our mistakes (like failing to check “free and correct”) on some design flaw from a manufacturer.

  8. Gennaro Avolio Says:

    It has been have noted the requirement is to verify “free and correct” operation of the controls when doing the check. Note that in many aircraft the control surfaces are not visible from the cockpit. This has no bearing if the controls are locked but could help to explain why the control check gets skipped, especially if the thrust lever are free.

  9. Bruce Landsberg Says:

    Gennaro….

    Good comment and you are exactly right. I believe, and perhaps someone could inform us, that there is some cockpit display that shows control deflection

  10. Gennaro Avolio Says:

    Bruce…..
    To add a comment to design engineers; can your design be partially operated? I fly an airplane with mechanical control locks. (Jet Provost) I tried to engage the control locks after the throttle was advanced. It was not possible. This airplane has no nose wheel steering so it would be possible to taxi with controls locked. If the nose wheel steering is ruder pedal controlled that should be a clue to locked controls.

  11. Manuel Says:

    Talk to GIV pilots. There is ample tribal knowledge about engaging and disengaging the control lock with and without engines running, and hydraulic pressure up, not bled down.

    Probably the factory is scrambling to understand the situation, develop warning flags to prevent apparent control lock disengagement per red handle position, but the ability to advance the power levers to achieve takeoff thrust… and yet have tan elevator that was locked pitch down.

    And surely, GIV pilots in initial and recurrent training will hear much about freedom of controls… as Bruce points out.

  12. Duane Says:

    Regarding Genraro’s comment about the flight controls not necessarily being visible from the cockpit to verify correct deflections, the obvious solution is to have another member of the flight crew or ground crew observe the control surfaces from either outside the aircraft or from elsewhere inside the aircraft prior to engine start. Any aircraft that is large enough that the control surfaces (especially ailerons) are not fully visible from the flight deck also requires a minimum two-person flight crew.

    And as Bruce also points out, modern avionics in larger turbine aircraft also provide electronic indications of the deflections. There is no excuse for not verifying free and correct prior to takeoff (if not as part of the pre-engine start flight inspection) In my practice, I always check for free and correct flight controls both during both the preflight inspection and as a final pre-takeoff check.

  13. Bonanza Babe Says:

    I’ll never condemn a dead pilot because of what he might or might not have done right before he took his last breath.

    But I will always scrutinize and try to learn from what he did so that I can be a better pilot.

    One thing I’ve learned, though – no matter the cause, complacency on someone’s part is always part of an accident chain.

  14. Laurie Says:

    In reply to Andy, I check the controls during the pre-flight. I cannot see the elevators when I am inside the plane, so I could not possibly verify that they were working. It is also easier to verify that the aileron controls are correct during pre-flight. My DPE, who also flies gliders, went one step farther and applied pressure to the control surfaces while I moved the controls. This was to establish the integrity of the connection; it is apparently common practice in gliders. Of course, I also check during the runup, in case something in the cockpit has jammed the stick or the rudders.

  15. Dave Conrad Says:

    Bruce,
    I flew both the G-IV and the G-550 for many years. Both airplanes require that the “gust lock” be taken off before engine start and hydraulic pressure is applied to the flight controls. Both airplanes have a rear view mirror in the cockpit side window to see the ailerons & spoilers on the wing. Only the G-550 has a “flight controls” synoptic page displaying flight control position… the G-IV does not. In the G-IV, the only way to confirm flight controls free and “correct” (elevator & rudder that you can’t see from the cockpit) is use a ground crew member after engine start and prior to taxi out. A great example of this are the military demonstration teams at an airshow. As far as the throttle interlock on the G-IV in this case, I’ll be waiting for the NTSB report. One thing is got sure… the crew did not taxi out with the intent to crash an airplane. Be careful out there.

  16. Bruce Landsberg Says:

    Dave…..

    Thanks much for the clarification. Rear view mirrors on a Gulfstream – who knew? As noted, there will be much to learn from this one – albeit with a very high price tag!

  17. Ron Rapp Says:

    It’s more of a “side-view” mirror. It’s typically attached to the pillar which sits between the front and side window. You can see a photo I took of the wing of a G-IV via that mirror here:

    http://www.rapp.org/temp/mirror.JPG

Leave a Reply

*