Yesterday’s airliner crash in Lexington, Kentucky has resulted in a few friends asking how it might have happened.
It was dark, one hour before sunrise, and hazy (see http://www.airnav.com/airport/KLEX). Runway 26 and Runway 22 start almost right next to each other (see official FAA airport diagram). The short runway, 26 (oriented in magnetic direction 260), would be reached first by an airplane taxiing from the terminal. The 3500′ of runway is very comfortable for a slow piston-powered airplane, tight for a light business jet, and more or less impossible for a fully loaded airliner.
This is not the kind of mistake that two professional pilots would be likely to make. If it hadn’t been dark and hazy, the control tower would probably have noticed the mistake and called to suggest aborting the takeoff.
How can we pilots protect against this kind of error? In airplanes with a heading bug, always set it up for runway heading before leaving the runup area. If you are positioned on a runway, preparing for takeoff, and the HSI is not lined up with the heading bug, this gives an extra opportunity to notice that something is wrong. Of course, the airline crews usually do things this way and it didn’t help the folks in Lexington.
I also make it a habity of looking at the big white numbers painted on the runway as I roll over them.
Apparently this has almost happened before:
http://aviation-safety.net/database/record.php?id=20060827-0
NASA ASRS records show a possible similar scenario when some 13 years ago a twin engine passenger jet inadvertently taxied into position on runway 26 at LEX while being cleared for an immediate runway 22 takeoff. The tower controller noticed this and cancelled the takeoff clearance. According to the report “Possible contributing factors were poor visibility and wx (rain), confusing rwy intxn and twr’s request for an immediate tkof.”
Didn’t they set up their FMS for 22? I think some very bad things happened procedurally in that cockpit.
When you are the first flight out and it’s still dark and you are setting up the aircraft for takeoff and you are running the Taxi and Before Takeoff Checklists and you didn’t get any sleep last night and you see the planes ahead of you departing on 26 and so on and so on. Yes, this may have been a people error, those people are ALL the individuals involved in this particular takeoff, the pilots and ATC. If only one of these people had done something differently, this might not have happened.
Aviation has been an extremely safe way to travel. It gets better everyday. With each incident/accident we learn and procedures are changed, pilots and controllers are trained and new signs are installed so THAT situation will not happen again.
What can I learn from this? When my turn comes, will I do the right thing?
I read that the sole survivor, the first officer, was the pilot flying the aircraft. If he makes it through (he is still in critical condition) I am sure that the human-factors people will want to have a long conversation with him. Is more and more automation necessarily a good thing? Are the procedures that are in place correct and appropriate for the kinds of mistakes humans are likely to make? Maybe some good can come of this in terms of making the system even safer and more robust against human error.
Given that we have GPS units that can tell where you are while driving, etc. is there any reason that some sort of GPS-enabled device w/airport-specific data couldn’t be added to the cockpit that would be able to say, “Warning: this runway cannot accomodate your class of aircraft”?
Obviously it could be much more precise, using airport-specific data, e.g. “Warning, you are now on LEX runway 26 which cannot accomodate this class of aircraft. Please contact the tower to obtain a takeoff slot on LEX runway 22”.
Or, at the very least, even if info about the aircraft couldn’t easily be incorporated into such a device for some reason, a generic GPS gadget w/airport data could still say, “Attention, pilot: you are now prepared for takeoff on LEX runway 26. This runway is not suitable for aircraft weighing over 12,000 pounds. If your aircraft weighs over 12,000 pounds, please contact the tower to obtain a take-off slot on runway 22”.
A more impractical idea: pressure sensors under the tarmac on short runways connected to an alarm w/sound, lights and radio-broadcast to alert the clueless in the cockpit and the tower.
–DC
It already exists
http://www.honeywell.com/sites/aero/Egpws-Home3_CB54AACBB-D557-208D-8CE0-EC44CECAAB3B_HD821C499-7201-CF0B-5533-6EFD92534345.htm
The lone air traffic controller appears to have gotten just 2 hours sleep in the previous 24 hours. I have been sleep-deprived off-and-on for 19 years running (my children are 19, 17, 13 and 4 years), and it really adversely affects one’s reflexes for glancing up frequently to make sure the airplane was pointing in the right direction while doing paperwork (as he reportedly turned to do once he gave the okay to this ill-fated craft).
I am looking forward to going up in Philip’s helicopter on Sat 9/9, first helicopter flight of my life. Closest I’ve gotten to that is a hot-air balloon ride in Arosa CH, but it was tethered to the ground.
I flew Egypt Air in the Sinai in the early 1990s a fair amount, so Philip’s helicopter probably isn’t that risky by comparison. Egypt Air frequently made precipitous drops in altitude when another airplane was sighted.
Should I go for it or chicken out? Four-year-old goes along for ride with dear Mum.
So it does.
According to this Washington Post article, the system costs $18,000 and only a few airlines are using it, none of them commuter airlines.
http://www.washingtonpost.com/wp-dyn/content/article/2006/08/31/AR2006083100959.html
From the article: “A Runway Awareness and Advisory System made by Phoenix-based Honeywell Aerospace uses a mechanical voice to identify the runway by number before takeoff and warns pilots if the runway is too short for their plane.
The system, which can pinpoint a plane’s location using global-positioning systems, also alerts pilots if they are trying to take off from a taxiway instead of a runway.
The software program _ an enhancement to Honeywell’s widely used ground proximity warning system that alerts pilots to mountain peaks ahead _ costs about $18,000 a plane. It was developed in response to Federal Aviation Administration concerns over runway accidents and close calls.”
I find it interesting that the media are making a big deal about the sleep -deprived controller. He did everything he was supposed to do. This just reeks of a cascade of cockpit errors, starting with the crew powering up the wrong aircraft (a ground crew member informed them of their error.) The 1st officer may have been at the controls, but the captain taxied and lined up the aircraft. No compass check? No questioning why there were no lights? No questioning where the distance markers were? (I’m pretty sure 26 doesn’t have them.) Sure, we could have more people watching from the tower and more GPS-enabled warnings, but at the end of the day you still need a crew that pays attention to detail and follows procedures. The most sobering part is that things like this happen more often than we’d care to think about, just without the tragic results.
qcohen,
While it may seem logical to develop another avionic gizmo to correct for every possible mishap, such knee-jerk reactions can be counterproductive. The $18,000 system adds weight, maintenance hours, consumes power, occupies panel space, require initial and recurrent pilot training, and, if malfunctioning, could cause confusion that may lead to an accident. Even in this case, I wonder what other complicated systems the pilots were programming and monitoring that morning that distracted them from the relatively simple task of taxiing to the runway. The runway number is painted on the runway in huge numbers, and the pilots have a compass, directional gyroscope, and moving-map GPS, all of which them which direction they are pointing, which is related to the runway number. Life has risks and dangers, and, fortunately, this one is incredibly rare.
(as an aside, my conjecture is that the recent outage of the lights on the main runway was a contributing factor, as it may have blunted the impact that the absence of lights had as a visual cue that they were making an error)
suzanne,
The pilots are responsible for flying the plane. Though the controller may act as a backup and warn of mistakes time-permitting, it’s not their job to put point the plane in the right direction. It is true that “safety is everybody’s job,” and obviously a controller would speak out if he or she saw it, but, at some point, it’s a little like saying “The overworked professors failed to wake my son and drag him to class.” Many smaller airports have no controllers at all at night, and pilots are generally capable of finding the proper runway. Obviously, this accident provides the controllers’ union with an unprecedented opportunity to discuss staffing levels while sustaining the public’s overstatement of the (vital) role of controllers.
I’m an air traffic controller and a private/muliengine rated pilot with 550+ hours. While I prefer to wait for the NTSB report determining the cause/associated factors to this accident, I’m compelled to wonder prematurely. When equipped with lights, runways are generally well lit. “Instrument runways” are always better lighted than runways used for VFR considering the requirement for approach lighting, etc. My focus is “on the lighting” because I can’t imagine why 2 commercial pilots would taxi on/into a dark void, while attempting to takeoff and apparently “not be alarmed” by the lack of visual clues, runway lighting clues typically provide. Notwithstanding “the intrigue”, as we speculate, I hope we realize that’s all we’re doing AND IF the findings of the NTSB prove suspect, the excitement of speculation becomes more justified, in our search for truth. (smile)
Cleveland,
There have been many issues with the lights at LEX lately: I heard that there were 4 pages of NOTAMS about the various outages. This in addition to the construction and change in taxi procedures likely created a confusing environment.
One of the pilots apparently did make a comment about the lighting during rollout, but, after significant construction and with a huge glob of NOTAMS about the lighting, things may have been confusing.
Question for Philip: Are NOTAMS one of the weakest links in aviation from a human factors perspective? Much study has gone into cockpit design, checklists, pilot-controller communication and procedures, and mnemonics like ARROW, PARE, and GUMPS. Yet NOTAMS seem to be a cascade of sometimes lengthy, haphazard sections of all-caps text with strange abbreviations.
While the rest of aviation seems to be “present information in an easily digestible manner,” the mantra for NOTAMS seems to be “you are responsible, let’s see how difficult we can make it for you.” All in all, it seems a little like the housemate who leaves your telephone messages in the refrigerator drip pan or between the sofa cushions. The TFR’s that pop up around stadiums and the President’s constant fundraisers after briefing but during a 6-hour flight are an infamous example.
Question: I’ve learned to set the heading bug to the wind direction – helps you with the wind corrections while taxiing, and provides a simple mental picture where you can expect the wind to come from during take-off. (Should check the windsock, obviously).
So, heading bug to runway heading, or wind direction?
The FMS may have been set up for the proper runway, but it wouldn’t have warned them about using the wrong one (which it is capable of doing) because the departure point for both runways are so close together.
Like most accidents, this one was a long chain of events which combined to claim a lot of people’s lives.
In my experience, the CRJ is not a well loved airplane by those who fly it. It seems to be in best form as the 601 bizjet they originally intended it to be rather than a makeshift airliner.
Fab: I’ve never heard about this wind direction idea. A typical IFR departure clearance from the Tower is “fly runway heading”, which is why virtually all airline pilots default to this.
K: I would concur that NOTAMs are almost useless in the text printout form that we get from DUATS and that most professional pilots receive. If things are at all complicated, I much prefer to deal with a human briefer at Flight Service (very capable folks indeed up here in the Northeast and also in Alaska; I’ve had mixed results in the Southwest and Southern California) who uses his or her judgement to sort through the morass. It would be nice if airplanes had Internet connections (and if the U.S. had a wireless Internet for that matter) and enough smarts to present NOTAMs at the relevant times (10 miles from the destination airport, for example, you’d be reminded of a crosswind runway closure NOTAM). That is how I would do it. I met a wise pilot at KMYF who said that he didn’t like to fly his noisy bumpy Mooney for more than 2-3 hours per day because “every hour that you’re in the plane, your IQ is falling.” At the end of a six-hour flight, how is a pilot supposed to remember something he or she read about eight hours before?