Increasing tech and decreasing separation.

[KL-666]

Hello HJ1an,

Diverting wildly from a planned flight path is scaled as quite an incident and would be reported. To my knowledge there exists no report of any false positive. We can not go on with what-ifs that are far from realistic. Do you know of any false positive report?

Kind regards, Vincent

[jwocky]

Here is a case where two Quantas jets were involved over Adelaide. One of them got no warning from TCAS:
http://www.theguardian.com/business/2013/oct/10/qantas-near-miss-warning-system-failed-on-one-plane-investigation-finds

This is an ASRS report also referring in a general way with TCAS problems
http://asrs.arc.nasa.gov/publications/directline/dl4_tcas.htm

One other problem is, that TCAS since it is what they call nowadays "secondara radar" (an euphemism for transponder-based), will not show small aircraft without transponders, transponders accidentally set to wrong modes and inoperative transponders (think MH-370 here)

However, while I doubt the 100% reliability of any technology, by numbers, it appears TCAS is usually the safer bet and one has to construct a real unusual situation to cause really a crash, like wrong use of the TCAS system per se (as in user error).

[KL-666]

Hi Jwocky,

The first article is about TCAS not warning. That is not exactly the same as a false positive.

The second article contains just a few worrying real events, but mostly it contains worries of ATC about events that have not happened (yet). If you compare this meager yield over many years of TCAS use to the many successful evasions per month thanks to TCAS, then i agree with you that the balance goes to TCAS for now. Probably the concerns of ATC can even be dealt with in the near future. Good thing that they speak up.

Kind regards, Vincent

[HJ1an]

Hello HJ1an,

Diverting wildly from a planned flight path is scaled as quite an incident and would be reported. To my knowledge there exists no report of any false positive. We can not go on with what-ifs that are far from realistic. Do you know of any false positive report?

Was thinking of the Turkish B737incident. That one is not TCAS of course, it's another a ground radar instrument I think - I'm just wondering how -psychologically - a false positive from the machine can affect judgement later on when it really does matter.

*However* - on the other hand I read that humans trust robots more, even though the same robot lead them to the wrong directions. So given that if the machine is the one that is more accurate in this case we're talking about, then this should be a good thing. Unless of course, when it isn't accurate.. (also counters all the talk about relying on machine flying, what an odd balancing act this needs to be)

[KL-666]

Hello HJ1an,

Turkish is a bad example for not trusting the instruments. They never looked at their instruments, so there is nothing to not trust.

There are better examples:

China Airlines 747 near San Francisco 1985: During cruise they manage to let the plane get in a 90 degree roll (not scanning again). When they wake up, one of them sees the 90 degrees roll on his horizon and says: "Hey, my horizon is broken". On the cross check he exclaims: "Look, your horizon and the backup are broken too". They lost 30000 ft before recovering a heavily damaged plane due to exceeding of many structural limits.

British Midlands 737 at East Midlands Airport 1989: No 1 engine goes defect, but they shut down No 2. In their previous 737 model they claim that the engine indicators were notoriously unreliable. So they based their judgement on smelling smoke, which indicates engine No 2, because the air conditioning comes from there. Wrong again! Their current 737 model sucks air conditioning air from both engines.

Nothing good has ever come from trying to be smarter than your instruments. Therefore airlines that train their pilots lay extreme emphasis on using the instruments before anything else (btw. flight director is not an instrument in that sense).

Unlike you try to suggest, it is possible to train to force yourself to do the right thing. I know it from car driving, where on a quiet road i am tempted to fiddle with something inside the car for seconds. But i force myself to do it in stages of less than a second, so my eyes are never longer than a second off the road.

Kind regards, Vincent

[HJ1an]

The Turkish one, I'm actually saying, if the plane did not have a false positive prior to it, and during that approach on autoland the fault occurred at that point, the pilots would probably have caught it, because it wasn't shouting at the pilots for the entire 2 hour trip, for example.

British Midlands 737 at East Midlands Airport 1989: No 1 engine goes defect, but they shut down No 2. In their previous 737 model they claim that the engine indicators were notoriously unreliable. So they based their judgement on smelling smoke, which indicates engine No 2, because the air conditioning comes from there. Wrong again! Their current 737 model sucks air conditioning air from both engines.

Nothing good has ever come from trying to be smarter than your instruments. Therefore airlines that train their pilots lay extreme emphasis on using the instruments before anything else (btw. flight director is not an instrument in that sense).

Unlike you try to suggest, it is possible to train to force yourself to do the right thing. I know it from car driving, where on a quiet road i am tempted to fiddle with something inside the car for seconds. But i force myself to do it in stages of less than a second, so my eyes are never longer than a second off the road.

Kind regards, Vincent

Good point. Regarding the Midlands 737 - if the pilots had the engine thing drilled into their brain previously (from the -200), and Boeing had actually changed the design and made mention of it for the -300 but not retraining and drill it into the heads, I can understand them doing it out of previously experience.

[KL-666]

No pilot is ever drilled to smell. Their training drills them to use the instruments. There is no excuse for not using your training drill. If you can not adhere to a simple training drill, you have no business in any cockpit.

Kind regards, Vincent