Flydubai Flight 981 Crashes on Landing in Rostov-on-Don, Russia

[elaw]

Well you can "do an ILS" with no autopilot at all... just tune your nav radio properly and keep the right needles in the right position and you'll... slam into the approach end of the runway if you don't slow down and flare but you get the idea.

Or, you can "do an ILS" with anywhere from 1 to 100 autopilots, in visual conditions where the pilot is carefully monitoring what Otto is doing and is ready to take control of the aircraft in case Otto gets distracted by a shiny object or whatever.

If I understand correctly, the term "autoland" refers to doing about the same as described above when the pilot does not have visual contact with the ground and therefore does not have the same ability to identify and deal with improper actions taken by the automation. I think in those cases, 2 or more autopilots are required to be working to legally attempt landing.

Two other random points:
Sure a go-around involves keeping the nose "generally pointed upward" but if you point it too far upward the plane can lose speed, stall, and end up emulating a lawn dart. Having multiple APs online reduces the chance that an AP fault could cause an "oopsie" like that.

The focus of most go-arounds is "aviating" as you describe... ensuring the plane remains airborne in a safe fashion. But in some circumstances, a tall mountain near the departure end of the runway for example, navigating becomes important too. A fully-automated go-around in such a circumstance requires more "smarts" on the part of who- or whatever is flying the aircraft.

[3WE]

...aircraft...all have GPS...my iphone gps is accurate...how difficult is it to have approach plates [.] programmed into the flight computers?...

...the very same plane can autoland...yet we don't trust it to do an automated go around...

Time out.

The plane CAN do exactly what you describe.

We have indications here that the pilots chose not to do it that way.

If things are truly as they are seeming, we are stuck with the same old arguments of why should the pilot even touch the controls, ever, versus exercising an ordinary skill which should work fine and which is a good idea to practice from time to time...

...and how did they mess up something that is a very basic skill (whether Otto was involved or not).

Finally- there is the twist of the very gusty, turbulent conditions- there might be an argument that that's a time for a real human to do the flying...and that's a very important caveat- the human isn't there to do a zero zero landing, the human is there to do a sloppy ILS to make visual contact with the ground and then land manually (using well-honed landing-in-gusty-wind skills).

Like I keep arguing with Evan, after the go around- it should be look at the instruments and point the nose properly skyward, and I think that even us parlour MSFS talkers might even be able to do that with fair competence (assuming I'm not a zombie from lack of sleep).

[3WE]

...Sure a go-around involves keeping the nose "generally pointed upward" but if you point it too far upward the plane can lose speed, stall, and end up emulating a lawn dart...

Really? I was not aware of that. And so, the plane stalls because of the slow speed?

(and I did use the adjective "properly" upward...you know, as opposed to... "relentlessly" upward )

[Evan]

PS 1- Gabriel is saying you can do an ILS with one autopilot, correct...I hear his point- why in the hell is 1 ok for an ILS but you need 2 for a go-around?

I just answered that. If you don't like my answer, take it up with Boeing...

PS 2- Is some of this irrelevant since you said auto land was not an option due to the crosswind (ok, you meant the gusty winds)

Autoland is for poor visibility, not gusting crosswinds. I don't know why a pilot would go to autoland when the winds are this turbulent and the runway is clearly in sight. I can only understand a dual-channel AP approach down to the runway, with a hand on the thrust levers ready to hit the go button.

[3WE]

I just answered that.

No you did not answer.

Gabriel, POST#159 (Bold is mine) "Sorry but I don't get it. If one single autopilot was good enough to fly the approach down to X ft AGL, I don't see why it would not be good enough to fly the go around from the same X ft AGL."

You see the word "why" in that sentence?

You keep arguing about the importance of a backup for all sorts of critical FLAREING things (that's a landing operation, NOT an autopilot flown ILS to 'typical' minimums...

Your comment, "ask Boeing" is where you finally get close to the answer- but you have yet to answer "Why" but instead argue with Gabe (It's Ok to argue with me- the truth lies somewhere between us (even though you'll argue that)).

It defies logic that you have to have two autoplilots to climb into a big open sky but only need one to go down a very narrow corridor...

Hell, yeah I get it for autoland / Cat II / whatever ...give me two Ottos if they are actually going to LAND the plane...but to execute the transition from an 800 FPM descent to a climb (without stalling and without undue altitude loss) that I can execute fairly well with a pretty fair degree of reality on MSFS down to 20 ft AGL???

[Evan]

No you did not answer.

I think Gabriel understands why you want a soft-fail AP provision when transitioning from a nice ILS glideslope into a critical pitch phase at low altitude. Hard fails tend to bring on those human factors. And then there are transitional autoflight-to-manual issues like stab trim positioning. Remember, these are passenger-carrying aircraft built around redundancy by a manufacturer with a pretty strong safety culture for a lifeform that doesn't always react well to the unexpected. Go-around is a critical phase that has humbled more than few cowboy pilots.

Get Boeing on the phone. I'm just betting they have a list of good reasons.

[Gabriel]

Really?

Gabriel, what reason, aside from a failed or MEL'd AP, would a pilot have for not flying an ILS approach in approach mode with both autopilots?

You completely missed my point.

- Gabe: Will the plane let me fly an ILS approach on a single AP down to 200 ft AGL?
- Evan: Yes (this is my guess for your answer, you didn't actually say that... yet)
- Gabe: If at that point (200 ft AGL on an ILS approach on a single AP) I hit the GA button, will the AP fly the GA?
- Evan: Of course not! A single AP flying a critical phase so low? What's wrong with you? It will disengage itself and let you fly the initial GA by yourself.

Do you understand the contradiction there?

[Gabriel]

You completely missed my point.

- Gabe: Will the plane let me fly an ILS approach on a single AP down to 200 ft AGL?
- Evan: Yes (this is my guess for your answer, you didn't actually say that... yet)
- Gabe: If at that point (200 ft AGL on an ILS approach on a single AP) I hit the GA button, will the AP fly the GA?
- Evan: Of course not! A single AP flying a critical phase so low? What's wrong with you? It will disengage itself and let you fly the initial GA by yourself.

Do you understand the contradiction there?

EDIT: I wrote this just after reading your initial not-response and before reading your discussion with 3we where you didn't convince me either.

[Evan]

EDIT: I wrote this just after reading your initial not-response and before reading your discussion with 3we where you didn't convince me either.

Well, maybe this will convince you:

NOTE: This is an FCOM from 2000. Avionics have evolved and this is before GLS I think, but AFAIK, this is still vailid, as the core principal of having a fail-passive autopilot for critical pitch manuevers at low altitude is sound. A single channel ILS approach should be terminated before 158ft AGL (JAA) or 50ft AGL (FAA).

[Gabriel]

Well, maybe this will convince you:

No, I still don't understand it.

Approaching on a single AP down to 50 ft then clicking Go Around and having to take manual control of the plane in a situation where trim and underslung engines induced pitch can be challenging is acceptable.

Going around at 1800 ft on a single AP is not acceptable,

Sorry, I don't get it. And it is not that I don't believe you that this is how it is. I already bought that. It is that I don't understand why it is like this,

[3WE]

No, I still don't understand it. Me neither, it seems illogical.

Approaching on a single AP down to 50 ft then clicking Go Around and having to take manual control of the plane in a situation where trim and underslung engines induced pitch can be challenging is acceptable. Yes and no...and I think you know what I mean

Slight edits: BUT Going around at 1800 ft on a single AP is not acceptable???, or at 500 feet? or 200 feet?...200 feet with good monitoring?

Sorry, I don't get it. And it is not that I don't believe you that this is how it is. I already bought that. It is that I don't understand why it is like this, You do read my posts right?...the emphasis on that word "why".

Evan gave the correct answer: "I don't know, ask Boieng". Unfortunately his absolute blind love and trust of all things engineering and procedural (and procedural with little complex caveats) prevents him from saying, "Yeah, it does seem to be counter intuitive".

[Gabriel]

You do read my posts right?...the emphasis on that word "why".

Yes I do and did.

[Evan]

Evan gave the correct answer: "I don't know, ask Boieng". Unfortunately his absolute blind love and trust of all things engineering and procedural (and procedural with little complex caveats) prevents him from saying, "Yeah, it does seem to be counter intuitive".

1) Why does my simply pointing out the behaviors of the 737NG autopilot result in attacks on my intelligence (emphasis on 'why')??

2) If it seems counter-intuitive, it is because procedure is in place to overcome flawed intuition. There's a lot of flawed intuition out there.

3) It seems counter-intuitive to me to not fly a dual-channel AP ILS approach if it is available. Why would you choose to fly it on a single channel (emphasis on why)? The STANDARD procedure (NG or A320) is to engage APPROACH and then BOTH autopilots. Why? Because aviation safety is built on redundancies. Yes, in a failproof world, one channel can do everything just fine. Do we live in a fail-proof world? No, we live in a world where systems fail and pilots fail to react correctly, become disoriented, lose situational awareness and slam into the ground far too often. That world runs downward at 700fpm but can quickly change to 4000fpm...

Unfortunately his absolute blind love and trust of all things engineering and procedural

Do you know why autoland is prohibited in crosswind component exceeding 20kts? (hint: engineering)

[3WE]

Do you know why autoland is prohibited in crosswind component exceeding 20kts? (hint: engineering)

Really, mister black and white? The challenge of teaching a computer to land in a gusty 20 knot wind?...engineering is the answer?

I'll bet you that that it's easy to get a computer to land a plane in a 20 knot wind and royally kick the butt of any human who tries to challenge it.

Because I'm not closed minded like you, my answers include things beyond engineering:

Meteorology: It's very hard to have zero zero conditions with 20 kt wind so there isn't that much point in having autoland.

Legalities & Human nature: No, we just don't want to trust HAL, do we.

Yeah, I feel better dying from a simple brain fart than I do a rudder reversal from a well-designed servo, or some goofy confusion that the automatic go-around won't work in situation Y- yeah, sure, that is poor logic and as to the why, that's another post.

While you can indeed ignore me, It would appear that autoland engineers have caved in (not JUST to me, but to a lot of other folks including some folks in the legal dept. at Autoland Industries, a division of Otto Inc.). This may not be that great of logic, but neither is allowing an ILS to 200 and 1/2 on one autopilot, but then refusing the go around button- where pilots are supposed to monitor because only one Otto is on.

Economics: (Given the afore-mentioned reasons including some engineering costs).

[3WE]

1) Why does my simply pointing out the behaviors of the 737NG autopilot result in attacks on my intelligence (emphasis on 'why')??

Your intelligence is high. Your closed-mindedness is off the chart.

Gabe asks why, you argue vehemently (what's the post count now?), but never answer why.

Your arguments are laden with a strong love of engineering and complex processes-with-caveats and a blindness to common sense.