II - yes sure I hear you and sure its not to suggest that the physical problem isn't isolated but some of these claims are pretty big and to say that their HUMS is going to capture everything to a 1 in a billion but then that same system seems so un-robust that it false alarms 14 times in 2 weeks....
Obviously that might be 14 times from 14 billion but on the basis it isn't.... If I have an oil pressure light in my car that is constantly "on" then its not really a valid system or do we disagree?
Especially since this system seems to have the buy in of the CAA, EASA and the operators.....
Edited to add:-
Going back to your point about 1/1bn flight hours - obviously this isn't back tested because the entire fleet hasn't done this many hours so OK its an extrapolation which might be fine if a certain level of robustness had been demonstrated around the process, which a high % of false alarms wouldn't suggest (if in fact 14 false alarms is a high % of the total run ups).
Thing is when you read the AAIB report on the two accidents and the element around HUMS you do find that this system and its thresholds are very specific to each airframe (hence I believe there is a fleetwide maximum threshold within the "learned" limits).
That might seem to be OK if it was business as usual but here it seems that the use of this data is being used to overcome limitations.
Last edited by Pittsextra; 5th Aug 2013 at 11:59.