QF32 and information overload

Airbus described the situation the QF32 crew faced after the uncontained engine failure as “dynamic”. That’s accurate insofar as it goes, but rather an understatement.

The obvious and immediate effects, like asymmetric thrust and aerodynamic damage, were dealt with by the autopilot, which remained engaged and the Qantas A380 crew elected to keep it that way until short final approach at Singapore. Tha’t helps, but over time an autopilot can mask the consequences of problems like growing fuel imbalance owing to the pierced left wing tank.

But the biggest psychological problem they had was the plethora of warnings and the mass of systems information provided by the electronic centralised aircraft monitor (ECAM), and the cacophony of associated audible alerts.

If you design a computer-controlled systems monitoring device to tell the pilots when things go wrong, you can’t blame it for providing 54 warnings when 54 problems crop up over a period of about 2min.

But that doesn’t make life easy for the pilots. They have to get their heads around the macro-problem they face very quickly, and avoid being distracted by the micro problems.  

Easier said than done. Each failure on the ECAM has an associated drill that - psychologically – clamours to be performed. 54 drills take a while.

It’s mischievous – but irresistible – to wonder, since this event ended happily, whether Capt Richard Champion de Crespigny who on this flight was undergoing an annual check by another trainer whose checking was in turn being checked, wondered for a millisecond whether this was all a set-up.   

The ECAM design acknowledges that multiple failures can occur, and it prioritises resulting alerts with fire at the top of the list and flight control system failures next. But there is a powerful temptation for the crew to fixate on all that information and inadvertently lose track of its ultimate set of priorities: to aviate, navigate, communicate, in that order.

This crew didn’t make that mistake. It probably helped that the crew was augmented by the check pilots since there was so much to do, but a big crew takes a lot of coordinating.

The full analysis of what happened in this immensely complex event will take a while to complete. It is a unique opportunity for the aviation community to carry out an assessment of human factors engineering as well as the inevitable questions about whether design for systems damage tolerance could be any better.  


4 Responses to QF32 and information overload

  1. David Nicholas 22 November, 2010 at 2:19 pm #

    While there is still a great deal to be learned from this event, and the demands placed upon the two “handling” pilots, it appears fortuitous in the extreme that there was such a lot of collective wisdom and understanding available in the cockpit. Not one, but two supervisory pilots who doubtless did a lot of trouble-shooting and maintained a considered assessment of the technical state of the aircraft, leaving the captain and co-pilot to “mind the shop” while the other two perhaps ensured that nothing was missed through preoccupation with the plethora of warnings and system messages. The other fortuitous circumstances were that this happened in daylight, the crew were rested and the weather good. It doesn’t take a leap of imagination to consider a less favourable outcome had this happened on the westbound trip, max weight, night departure, standard crew and with thunderstorms or cyclonic winds to cope with. If it was going to happen at all, then both Qantas and Airbus must be thanking their lucky stars that fate had dealt the best hand of cards………

  2. Layman 23 November, 2010 at 9:38 am #

    I agree with David. This incident has many positive results~ from RR receiving a largely intact engine to study and work out what went wrong, to Airbus being able to properly interview the pilots and get feedback from a real live event and possible put those learnings into future updates to their software.

    All in all, it’s bad that it has happened, but it will positively contribute to making aviation safer.

  3. Gary from Sydney 23 November, 2010 at 10:33 am #

    Perhaps Airbus will learn how to make a cockpit flight data recorder that lasts longer than two hours – and no I don’t know how long a Boeing recorder goes for – perhaps this is a regulatory issue. The overall impression is that this was not far off a catastrophic incident (which is probably why Qantas suspended A380 ops so quickly).

  4. David Learmount 23 November, 2010 at 10:44 am #

    Qantas has just announced it has completed its checks and is getting its A380 fleet airborne again. Singapore Airlines and Lufthansa removed some of their Trent 900 engines for checks and rectification, but they did not ground their fleets at any time.