The use and value of 'AI' can be debated elsewhere, but it is now opportune to engage with a thinking tool.
For subjects which can be fact checked - systems, books; a ChatGPT (Large Language Model) summary based on a wide source of information is more likely correct - law of large numbers, preferable to limited and variable human opinion. Consider the number of factually contributory posts vs the irrelevant ones in this thread.
Referencing ChatGPT is considerate, opposed to copy and paste without acknowledgement.

The glib conclusion of 'trash' lacks supporting evidence or reference.

Similarly 'the clear safety view', lacks reference. Airbus documents explain the purpose of BTV, not safety re landing overruns.

Additionally, proposing design improvement 'without a clue and based on a guess', is a typical misunderstanding of complex systems - man, machine, environment.
Ackoff or Chapman have described the human aspect as 'those who propose solutions, are part of the issue'.
----

ChatGPT stated that BTV 'automatically controls deceleration using braking and reversers', 'considers reverser use dynamically', which implies thrust modulation; thus for clarity, further research:-
'Does the BTV system directly control the thrust reversers':
Response, 'No, the Brake to Vacate (BTV) system does not directly control the thrust reversers. However, it takes into account the reversers’ effect on deceleration and adapts braking accordingly.'

I do not agree with the ChatGPT conclusion inference ( my inference ) that the BTV (alone) reduces the risk of overrun, but if considered together with ROPS, then the latter is a valuable system. Possibly a weakness in language and context, but no more than in some posts.

This seems like one of the more straightforward and reported questions that can be asked:

https://cimg9.ibsrv.net/gimg/pprune....2066359422.png


Some of the statements are correct. In all but one case (MD-80/MD-90), the reference to specific aircraft types is incorrect.

It does not 'conclude' - it is not a thinking machine. It is just a text generator that fills in the most statistically likely word. That's why, when you ask it 'does [transport category aircraft] have a RAT' it shamelessly returns 'yes' because statistically the most likely answer to appear there is 'yes' because more aircraft have them than not. It's that simple. It seems very sophisticated but it takes only the slightest experimentation to realise that it's ultimately useless for anything connected to reality and facts because it just predicts words without having any understanding of the physical reality.

Do not be fooled by the marketing!

Please get this thread cleaned up...

An interesting discussion. I fly the 737 so was not aware that Airbus has a clever computer to calculate the best autobrake setting for a particular runway exit. In the age of criticism of pilots for being children of the magenta line, it seems counter-intuitive to keep adding automation that is either insufficiently thorough in terms of the factors it considers or takes so much thinking away from the pilot that, in the absence of an operating BTV system, they would struggle to make a sound brake to vacate decision.

We adhere to the brake to vacate mantra at my (very large) airline but it relies on a pilot considering flap, brake and brake cooling factors to achieve a safe, controlled exit at the speed appropriate for the RET or 90 degree exit in the prevailing conditions. A good decision should also consider passenger comfort in my opinion. The more we add automation the less pilots will understand how all these factors interact.

Just my two cents worth.

Given the number of aircraft including many many 737s that have had overruns, I'm not sure that the pre-BTV situation is that awesome.

Seems to me that the situation is ripe for automation: under normal conditions, it works fine. It's well-positioned to run effective self checks. If the self checks fail, it can either command maximum autobrake or hand over directly to the crew. If you get enough systems failures that it fails, there's nothing wrong with doing maximum braking to a stop if necessary.

ROWS/ROPS is intended to catch situations where pilots are landing long or not applying enough braking and thus predicted to be running out of runway, and tell them to go around or brake harder as the case may be. BTV is pretty minimal extension of that.

It's been in service more than a decade now. Anyone know of any relevant incidents on ROWS/ROPS/BTV aircraft?

Thanks for all the useful responses. As for the others; if it floats your boat then keep on floating…

I am capable of the following:

A) stopping any vehicle I operate comfortably and safely.

B) Learning more about technical things that interest me.

And… you may not believe this because you may not be capable of it… I can actually do BOTH at the same time! :ok:

The world wonders :ugh:

BD

Belief in safety
 

And … believe it or not, … experts in human behaviour and limitations would say that you are seriously mistaken.

Are you a pilot? Your responses seem more engineering focussed - which obviously has it’s place, however you seem quick to want to take the human out of the equation - which is unusual for a professional pilot.

I've used BTV on the A350-1000 and it works brilliantly. Slightly unnervingly it doesn't always start braking immediately but does it in a fantastically controlled way that brings you to 10kts at the point you have selected as your turn off. You can of course override it at any point. I fell in love with it very quickly, great for managing brake temps.

It's sole reason for existing is to reduce runway occupancy time. In fact on the 350 the ND will show you approx runway occupancy time when selecting the exit. As such the aircraft will brake as late as possible given it's programmed g load limits. Pilots new on fleet usually will be a little worried, disconnect early and brake manually, but once comfortable it's best to let it do it's job until reaching a speed that is suitable for the given exit. If you watch the brake pressure 'gauge' what you will notice in most cases is that after nose wheel touchdown one will see slight brake pressure (unless the exit is far in which case braking starts later), after which the pressure keeps increasing as one gets closer to the exit untill the aircraft reaches the BTV disconnect speed.

Hm, that sounds like the opposite of what I would do for pax comfort... In my agricultural Boeing I try to keep a constant deceleration until reaching 40-50 at the beginning of the high speed exit - when deadheading I hate when people clearly set an autobrake setting that's too low to make it, and then step on it hard to try and vacate using the first exit.

Braking as late as possible gets you:

• Highest average speed and therefore shortest time on runway, clearing the runway for the next arrival/departure ASAP (as noted above)
• Maximum energy lost to reverse thrust, drag, spoilers etc.
  • Therefore, minimum energy absorbed by the brakes and hopefully minimum brake wear - carbon makes that a bit complex.
• Minimum stopping margin, because you're assuming you have full expected braking performance right when you need it, and any shortage will cause an overrun (past desired exit, not necessarily off the runway)

My understanding is this is one of the main ways that automatic train operation and moving-block signalling lets you fit more trains in: computers are better at doing the braking maths than humans, so can operate with less safety margin but still be safer overall.

I find that surprising. Airlines don't care much about runway occupancy. They do, however, care a lot about brake wear and money. As for brake cooling, how often does that matter in the slightest with longhaul flying with their lengthy turnarounds? Even with shorthaul airlines with 25 minute turnarounds brake cooling is rarely of practical consequence.

Incidentally, a competent pilot can achieve the same thing (brake later and then at a moderate, comfortable rate, to achieve the exit at the chosen point, thereby remaining in the loop throughout.

BTV is perfectly compatible with minimising brake wear, as previous posts have explained.

Until the day that they can't.
Like every competent pilot can stop within the expected landing distance, but some don't.

ATC and the airport operator are interested in runway occupancy, traffic flow, density, revenue; thence back to the operator.

Yes and no. Reducing ROT means getting able to fit more arrivals and/or departures in, leading to increased runway utilization and possibly the availability of more slots into crowded airports. Airlines and the industry on a whole prefer this.

It's not necessarily the case the BTV increases brake wear. Carbon brakes do not wear based on pressure applied but on the number of applications, so a single constant application of increasing force shouldn't be any different to any other application.

Usually with a regular autobrake pilots will disengage during the landing role and then reapply brakes a second time manually, so that's two applications per landing vs one with BTV.

The benefit with BTV is that there is less braking during the high speed phase of the landing role so the reversers and drag do the majority of the work as the deceleration rate increases towards the end of the landing. This actually leads to lower brake temperatures in most cases especially if using a high speed exit as BTV can be disconnected early and the highest planned brake pressure is not even reached.

What puzzles me is the punters who select a deceleration rate about 45 minutes before and almost immediately on touch down stamp the foot things resulting in a jerk, immediately release, wait for a bit and then smash both clompers in some sort of uncoordinated way to achieve a “safe” but ##### exit from the runway…

Still learning what’s new; still stopping where I should…

BD

I didn't say they don't care. I said they don't care much.

I rarely use autobrakes at all, make the desired exit, manage the brake temp and company says we do not have any brake issues. It is possible! Aircraft have become an engineering playground. Auto TCAS, Auto Emergency Descent, everything Auto...

SLF and retired Software Engineer.

... and also consider the spate of "daft - but sounds authoritative" responses to some questions.

"You Can’t Lick a Badger Twice": Google's AI Is Making Up Explanations for Nonexistent Folksy Sayings
https://futurism.com/google-ai-overviews-fake-idioms

... and a host of follow-up examples involving non-existent idioms. I think that the AI's interface has been changed to prevent answering similar questions, but IMHO it exposes fundamental flaws in large-language model applications.

PS I don't like the term Artificial Intelligence , mainly because it covers too wide a range of algorithms. But some "AI" software -- e.g. screening medical scans for cancer, etc -- seems significantly better than a human (let alone a human reviewing scan after scan).

I still fondly remember the occasional use of "Artificial Intelligentsia" in some technical journals back in the 60's and 70s. Referring to those producing the overblown marketing hype around the term AI.