Tag Archives: CRM

More Questions Asked by Flight Crew Members Regarding AF 447

More questions have been raised by fellow safety minded pilots regarding AF Flight 447. Here are several:

1. Why were they flying through that nasty convective weather?

2. Why did the captain leave the cockpit just prior to the flight’s arrival to this area?

With respect to the questions asked, here is my take.

1. The captain had at least a 2 hour-old satellite pix by the time he started engines. The thunderstorm encounter occurred somewhere after about the second or third hour of flight, meaning that the AF 447 entered the thunderstorm area about four to five hours minimum after the satellite picture was taken. This means that it is very possible that the satellite picture that the captain received did not resemble the convective weather area that the flight eventually encountered enroute. This is assuming of course that the crew got a satellite pix at all out of Rio.  This is a very common pre-flight briefing problem for all airlines operating oceanic flights. Making the problem worse is that Flight Control is not required to resend the latest satellite shots to the crew during the flight.  Why is that? Why do we have rules that not only do not provide important weather information, but do allow the provision of somewhat inaccurate weather information? Why is that?

A. Let’s say that the captain was a cautious fellow and planned well. So he gets up a few hours before the hotel pickup, has a shower, checks his email, gets packed, dressed and meets the crew in the lobby. At the airport or maybe even at the hotel, he sees the satellite picture prepared and sent by AF Flight Control Dispatch Offices.

I have not seen a satellite pix or a weather observation from the time from 4-5 hours before the mishap, but if I was investigating this incident, that is the first place that I would look. Why? Well satellite picture is the information that will inform the captain as to what lies ahead of him and will inform his game plan, sleep strategy, crew switch strategy for the flight.  So my money is on a sat shot that shows widely scattered storms if any at all given to the captain during pre-flight briefings. He doesn’t see a threat at this time and doesn’t see any reason not to use his normal strategy of getting the oceanic clearance, getting the plane out onto the transoceanic tracks and then retiring to get some rest, letting the other crew members handle the routine of the crossing. This will allow him to get some sleep and be fresh when they get close to Europe and have to start their let down at CDG for landing. Maybe this is his normal routine. Based on this supposed scenario, it is a reasonable one.

B. Alternative scenario: The captain is a “show me” kind of guy; unless he sees the lightning flashes and is face to face with a thunderstorm, he doesn’t worry about the metro stuff, because it is all 4-5 hours old data by the time they get out onto the tracks for crossing.

Either way, I do not think that the captain foresaw that 60,000 ft+ inter tropical convergence zone storms in his path would be a real possibility.

Additionally, we might consider this: thunderstorms anywhere rise rapidly in height, developing at 2000-4000 feet per minute, even as much as 6000 feet per minute. Therefore, information such as “the previous flight got through this path” is not really all that valid information for following flights, especially flights following 10 to 15 minutes later.  The fact that another flight transited this area 10-15 minutes prior is certainly and in my opinion most definitely not an endorsement for safe flight in an area of active thunderstorm activity. Thunderstorms by their very nature are unstable. That is  what makes them so dangerous to flight. Thunderstorms violate the very first principle of aerodynamics, which says, “assume a homogeneous airmass.”  For sure a thunderstorm is not a homogeneous airmass!

Thunderstorm development in the inter-tropical convergence zone (ITCZ) is even more dynamic than at other places on the globe. This means that even more caution needs to be exercised in the ITCZ by commercial airline operations. Yet remember that this mishap occurred on a flight planned and operated regularly in the ITCZ by this airline.

C. Third scenario is this. This is the kind of fellow that does his own thing. He does the takeoff, then he does his nap, then he does his arrival. This is the way that he does things and this is his thing.  This is how he does every oceanic flight and he figures the other crew can handle things.  That is just the way he is.

D. Fourth possible scenario: The Air France Dispatch Office in CDG does not consider it their responsibility to update Air France flights with Air France passengers on board of dangerous weather that has developed and is now in the path of an Air France flight. They consider it their responsibility only to get the pre-flight paper work prepared and delivered and then do no further duties other than to stay awake and answer any questions that come up during the flight.


My guess is that AF Dispatch Flight Control Standard Operating Procedure SOP is not to initiate any inflight communications unless directed to do so by “higher authority” or some similar version of that idea. The second place that I would look were I doing the BEA investigation would be the AF Flight Control Log for AF 447, the AF Flight Control SOP and all of the metro data that was available to AF Flight Control before the mishap.

I would further examine just exactly how AF Flight Control does business every day and night with respect to how much info is passed to airborne flights concerning dangerous weather occurring ahead of any AF flight.

Sorry to say this but more than a few commercial flight crew members are like B and C. I am even more sorry to say that more often than not, many flight control offices are like D. They send company flights on flight paths that intersect dangerous weather and do not change plans along the way. My guess is that this most likely remains “the way things are” at many airlines. How about yours? Are you in a position to make some changes at your airline?

So that is some thoughts on question one. But if you are a good captain and a good communicator, my guess is that you also have a few ideas and I would be very interested in hearing from you and reading your thoughts on this same question.

Now for question two, here is what I think. I do not think that the captain was fully aware of what was ahead of him. I also do not think that the captain expected a failure on the pitot static system due to icing to occur if the flight did go into clouds.

Remember that the pitot static equipment manufacturer, France’s CAA, Airbus and AF all denied that the equipment was faulty, even though there had be several incidents of failure previously of this same equipment in this same scenario and that it had been reported in the industry press.

So that captain may have allowed himself to be informed by the “authorities” in lieu of informing himself through reading industry incident reports.

Again, there are plenty of crew members who allow themselves to be informed this same way, at airlines all over the world, pilots who do the same thing.

Those of us who doubt “authority” are few, probably less than 1/3 of all. My guess is that real doubters are even fewer than 1/3, more like 1/10!

So, here is a third question:  Why didn’t the captain jump back into his seat and take control as soon as he arrived back up front? I know that I would have. You? Sure things were confusing, alarms going off, icing, thunderstorms, panic-what better time for the captain to be in charge of the flight deck and the controls?

I hope that this mishap is eventually re-investigated by some board other than the BEA, a board that knows that they are doing a Safety Investigation and not a Legal Investigation for some future court proceedings.

In the meantime, I am blogging my thoughts to allow local safety managers around the world to think deeper about keeping their operation safe. That is my intent.

Keep in touch.

Air India, Mangalore 22 May 2010, Mishap

All: Does it appear that this operator and the regulator are “coming to the safety party too late? Are they trying to close the corral gate after the horses have already escaped from the barnyard?”

Is this “investigation” going to relearn of all of the lessons already learned by others? Would this investigation be categorized as a “safety investigation,” or more of a futile exercise of re-examing the same mishap over and over again?

Possibly shooting from the hip, but here is what pops out at first blush:

1. No stabilized approach criteria, training or performance
2. No crowned and grooved runway, nor a national policy by the country regulator
3. No wind field data available or provided to the flight crew during convective weather
4. No back up of flight crew with real time metro data from their own flight dispatch office
5. No convective weather wind shear evaluation escape procedures trained, checked or used by the flight crew
6. CRM training appears to be ineffective.
7. Reject landing procedures, training and performance appears to be ineffective.
8. No coordination of convective weather arrival avoidance and holding procedures in effect by ATC

This is Just My Humble Opinion of course. But it seems to me when all this process is completed, nothing will be learned that could not have already been determined ahead of time, before the mishap occurred. This mishap appears to have been preventable. Why are so many organizations around the world unwilling, unable and unprepared to operate with safe standards?


Pilot Error Seen Likely in Fatal India Landing (WSJ, June 21)

Investigators are becoming increasingly convinced that last month’s fatal Air India Express crash was caused by pilots landing too far down a hilltop runway, and belatedly trying to get the Boeing 737 airborne again, according to people familiar with the probe.
Preliminary conclusions about the May 22 accident in India’s southern city of Mangalore, these people said, indicate failures by the pilots to follow basic safety rules during the final approach, compounded by apparent cockpit confusion after touchdown. Only eight of the 166 people aboard survived the early morning crash, in which the twin-engine jet plummeted down a cliff at the end of the strip, broke apart and then caught fire. It was India’s worst aviation disaster in more than a decade, and has prompted a barrage of criticism.
Regulators and international air-safety experts are now focusing on ways to fix persistent shortcomings in India’s pilot-training programs. Indian officials already have proposed tightening rules ranging from acceptable landing procedures to fatigue-prevention schedules for crews. Broader mandates are likely.
Other areas under scrutiny include safeguards when pilots temporarily leave the cockpit for breaks. There also are stricter requirements that all aviators-particularly expatriate pilots-speak good English and are well-versed in air-traffic-control terminology. The captain of the accident plane was a British citizen of Serbian descent, and the co-pilot was an Indian national.
Indian regulators also are considering ways to ensure that pilots are trained and assigned so they fully understand the particular handling characteristics of a designated aircraft model.
“A lot of new and revised safety proposals…will be disclosed next month,” an official with India’s regulator, the Directorate General of Civil Aviation, said Friday. The DGCA official wouldn’t discuss details of the probe except to say that “the possibility of pilot error” on the flight “cannot be ruled out.”
The moves come amid increased domestic and international criticism of alleged gaps in India’s commercial-aviation safety net. A U.S. Federal Aviation Administration team plans to visit the country later this year to make an informal assessment of progress in enhancing safety, according to people familiar with the issue.