Following on from Capt Stable's comments,
(a) the ILS is a very old technology system and has its share of problems. One Australasian airline very nearly put a 767 into the water a year or so back when the automatics "captured" a non-existent glidepath and the crew were slow to pick up the error. Fortunately the safety culture in the particular airline was sensible and proactive (rather than punitive) and the resulting very extensive investigation produced a number of surprises ... indeed there was a suggestion that some seemingly inexplicable crew action accidents in the past may well have been caught out by a similar event ...
(b) consider also the complementary problem of false course captures .. some of the older automatics will quite cheerfully capture whichever signal occurs first .... GP or LLZ ... false or correct .. pick whichever you like. The same can happen to the unthinking pilot who is handflying the approach with brain in neutral. I had an instance with a 732 sim session not so long ago ... command upgrade chap (working more than quite hard under my tender and gentle care, I might add .. and doing pretty well) .... he selected the automatics for capture with some distance to go and then allowed himself to become distracted by other distractions (would I put in distractions ?) ... whereupon the machine captured a false LLZ path quite some considerable angle away from that which was desired. The subsequent "approach" proved to be quite puzzling to the crew as the picture wasn't right ... but, never having seen it before, and apparently not familiar with the possibility, they were unable to work it out under the remaining workload pressures ... I had allowed it to progress to see if the prioritisation penny dropped .... after we eventually froze the sim ... it ended up being a useful flight management discussion point.
The main point of my commentary is that both of the above problems ... and a good many others, can be avoided if we use the ILS in the manner intended.... it should be considered as an electronic funnel .... it is up to THE PILOT to find the entrance to the funnel and then it works just fine.
Far too much these days, we see the automation passive pilot flicking switches and pressing buttons with gay abandon but without any regard to the fact that these things are computers .. and computers screw up with regular monotony. If you trust it with your life without a healthy degree of suspicion and scepticism .. then there is a not insignificant risk that eventually you will pay for your casual attitude with your life and those of your passengers.
It is most important that the pilot take active responsibility for feeding the aircraft into the ILS at the appropriate point ... and actively check that both the location (by whatever means the approach plate specifies) and height are correct .. and likewise again at the marker ... it all works well most of the time .. but when it screws up .. it can do so big time ....
Last edited by john_tullamarine; 22nd June 2002 at 02:55.