PPRuNe Forums - View Single Post - Air Asia Indonesia Lost Contact from Surabaya to Singapore
Old 17th Feb 2015, 11:58
  #3248 (permalink)  
Ian W
 
Join Date: Dec 2006
Location: Florida and wherever my laptop is
Posts: 1,350
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by SLFandProud
Oh absolutely, I quite agree; the point I was trying to make is that this is a completely normal and natural evolution, and as depressing as it may be if you were once a 'sky god' there's no point tilting at that particular windmill.

It happens in /every/ industry.

Hell, the software engineers of my generation will moan that the job was more skilled when we hand rolled assembly language using nothing but a HEX editor, and that all this modern nonsense like automatic garbage collection and strict type checking means that any old Tom, Dick or Harry could write code with all the skill taken out of it.
Your analogy does not hold for aviation.

As a software engineer you will know that there are always going to be difficult areas - the 'otherwise cases' that drop through the IF-THEN-ELSE logic or the places where the analysts and designers cannot come up with a simple fix as the number of potential permutations in the real world make a simple solution difficult. In the FMC and Autopilot software the way out of those nP problem areas has been for those systems to failover to the human pilot handing them the bag-of-bolts and expecting them to recover the situation that the automation could not.What you are saying is true that many areas of work have had automation creep in and deskill the operator. But aviation is unlike those as people will die if the automation goes wrong and the software analysts, designers and implementers have ducked the final responsibility when it gets difficult and handed it to the pilots. You know the 'sky gods' you keep running down. Because management in their ignorance treats automation as fault free they are paying off the 'sky gods' and replacing them with systems managers. So they are in the process of removing the final safety stop that the software analysts, designers and implementers always assumed would be there to save the day. In other words management and the software analysts, designers and implementers have worked together to create a system that fails dangerous rather than fails safe. Indeed short sighted attempts at graceful degradation have actually made the failure cases far more difficult for the human to take over as the actual state of the automatics can be indeterminate - but they cannot be switched off.

So the systems builders expect a 'skilled sky god' to save the day when they find it too difficult to build automation to cope and their system fails. Their systems do not fail in a simple way but sometimes in unexpected and complex ways compounding the original problems and management, trusting the system providers who say how very safe their systems are start deskilling the pilots so there are no 'sky gods' available in the cockpit to pick up the bag of bolts.

This potential problem was foreseen long long ago. One of two things will have to happen, the systems builders clean up their act and build 100% fault tolerant systems that can cope with multiple sensor failures and never lose control and automate the pilot completely. This may happen sometime but on current showings the systems builders are not yet clever enough and couldn't afford the insurance. Or the industry starts ensuring the pilots can actually pick up the pieces WHEN the systems fail. This means very limited training in the routine and very intensive and extensive training in systems faults and LOC including live manual flying and LOC training.

I get the impression that at the moment heads are firmly in the sand they will need to be extracted.
Ian W is offline