ACAS Xa

[LeadSled]

Folks,
I have just been catching up on the flight tests of the new system ACAS Xa, to replace TCAS II around 2020.

Sound like a very significant improvement, but still only pitching manoeuvers for resolution advisories. The system will now include 1090ES ADS-B/C as part of the calculations.

I did have to have a bit of a chuckles, given the resolute rejection by so many in aviation in Australia to any vestige of elements of probability theory being used in airspace management --- the basis of G through A designation.

Its internal workings in ACAS Xa are a complete departure from what we have had for years, finally TCAS II ( the "If-then?" logic stream). The new system logic is all based on probabilities, to the degree that it has at least some of the characteristics of AI -- Artificial Intelligence.

Compared to the "old" system, testing seems to have gone very well, and the problem of false positives is almost non-existent, apparently. Another great advance claimed is the ability to upgrade the basic operating system software without the testing and verification nightmares that have been a "feature" of the old system, which took years to upgrade for known problems --- that many of us experienced in early years of TCAS.

Tootle pip!!

T

[Capn Bloggs]

And the luddites objected to us demanding transponders in E airspace. Oh how times have changed. :rolleyes:

[LeadSled]

Bloggs,
Is illogic your only modus operandi?

It is probabilities that said transponders were not needed in E, and as for the AFAP maintaining that E was "less safer" than G --- words fail me.

Your comments are completely unrelated to ACAS Xa (and u) --- have another try.

The central nature of probabilities in the software structure of ACAS Xa reinforces the importance of taking into account real and assessed risk, and not the wishful thinking of perceived risk.

Tootle pip!!

[Capn Bloggs]

Your comments are completely unrelated to ACAS Xa (and u) --- have another try.
Oh, so TCAS Xa doesn't need a transponder? That's clever.

[LeadSled]

Bloggs,
Are you going senile in your old age, or is it what you are smoking?

I never suggested that TCAS didn't need a minimum of a transponder in one aircraft, I never mentioned transponders in the original post. I doubt there is an example, but you could, in theory, have TCAS/ACAS without a transponder.

The whole point of the post was about the use of probability theory in the new ACAS Xa, which will supersede TCAS II ---- and the fact that a group of you, for as long as I can remember, have rejected the concept of probability theory that underpins the whole of modern ICAO airspace management.

How often, over the years, have I had to listen to your representatives reject any possibility of probabilities as part of the answer, instead demanding "absolute safety" ( damned if I know what that is, and not just in aviation) and regulatory responses based on perceptions of risk --- even after admitting there is no demonstrable evidence of the risk perceived.

To more or less quote one of your not at all highly esteemed representatives: " We don't care how well it works in the US, we are not going to do it the way the septics do it".

Now you are going to have to wear full-on probabilities in the new ACAS Xa, whether you like it or not, because you are not going to have a choice.

Tootle pip!!

[cattletruck]

Interesting opening post regarding the new AI approach to this problem. Determinism replaced with probability (maybe not entirely), but both achieving the same desired outcome.

Many years ago I was jump-seating in a 744 when the TCAS threw an RA - climb FL310. Although we had visual contact, it came as a surprise to me that our fully loaded 744 had to spend a few crucial seconds accelerating from cruise to climb speed, after which the threat was no longer a threat, but we followed the RA anyway. Hopefully this new tech is a little smarter.

[Capn Bloggs]

I never suggested that TCAS didn't need a minimum of a transponder in one aircraft, I never mentioned transponders in the original post
But you did make yet another snide quip at those who have a different view of airspace and risk to you, hence my reference to you, who thinks that E doesn't need transponders, nor a radio. As for the codswallop that A is only as safe as G, I'll ask the same question as you asked me, what are you smoking?

Speaking of smoking, I'm as clear as mud on your description of the use of probabilities in the new TCAS. How is that different to the current system, where the gear calculates the CPA and generates a warning? Are you saying the new system will be saying "it's probable you'll hit so I'll pop up a warning" without doing any normal collision-calculations (although you did mention it using ADS)? or is it just the same principle but improved? You're either going to run into someone or you're not. Is the AI going to say "I saw this yesterday... looks similar... I won't issue a warning"?

Latest TCAS 2 6>7>7.1; a simple firmware change or a hardware change?

Lets all get away from "early years of TCAS" and "Many years ago". I'm sure there are plenty of warries around of the first iterations of TCAS but that was years ago and totally irrelevant to today's kit. 7.1, at least where I operate, lets us get pretty close without bleating.

[LeadSled]

Bloggs,
Do you realise how long ago, now, load 7.1 was?? Work it out.

Yes, properly managed, ICAO G through A is exactly the same separation assurance probability -- but your mob have never accepted that simple proposition --- much less accepted the fact that G through A was just an ICAO adoption and codification of FAA practice since around 1959 or so -- not long after FAA was first created.

You would not accept around forty years (40) of proven practice.

Worst thing was that it was called A through G, when the separation assurance standard remained constant, but you lot decided to treat it as an ascending or descending order of something called "safety", being apparently unable to functionally understand and accept the concept of risk management --- when it came to airspace management.

Go find the facts for yourself, the whole underlying design of the ACAS Xa software is new, a wholly different principle to TCAS/ACAS so far. It is NOT a further amendment to the existing TCAS II. It is NOT TCAS II 7.x.

From my point of view it is the delicious fact is that you are going to be stuck with a whole new probabilities based software program and system ( that will be undeniably less error prone than the TCAS logic strings) and you and your mates will not ( and have not) ever be consulted, much less have any say in the matter of its adoption.

Or put another way, you will not have the opportunity to degrade a proven system, and force increase costs, when adopted by Australia

It will be jointly mandated and adopted by FAA and EASA, presumably simultaneously by ICAO, some time soon after 2020.

Tootle pip!!

[Capn Bloggs]

Go find the facts for yourself, the whole underlying design of the ACAS Xa software is new, a wholly different principle
Leddie leddie leddie. You made the first post. You mentioned AI. You didn't explain in even the most basic terms how this is different to current, apart from raving on about my inability to change it because of it using probabilities?? Keh? I am not going to waste my time trying to work it out, either. As is standard with you, you mention a small bit of the whole story and then tell us to wrack off and find out the rest ourselves.

Besides, anything to improve my ability to keep away from you transponding bugsmashers, I'm all-for. Even better if you have ADS-B. Now there's an idea...

10 aeroplanes in the terminal area at an airport. What's safer, Class A or G? Thank you.

Could you answer my question. Was 6>7>7.1 a hardware change or a mere software/firmware update?