The Plane That Saved Its Pilots | Air France Flight 2510

[@MiniAirCrashInvestigation]

About the title, yes I know about the apostrophe s, This video is doing quite well and I dont want to spook the algorithm by changing it. Thanks

[@uzaiyaro]

Hey, awesome video! Although from someone else (who happens to be an A320 pilot), id love your thoughts on this one. Its about how an oil change wrote a brand new A320 off, and also caused the failure of all four elevator computers, all three hydraulics systems, and both engines, and generated an error so rare that no checklist existed for it: kzbin.info/www/bejne/fXqpfneZnK2Sbsk Thanks again for your hard work!

[@electricmaster23]

Ah, I see. Don't poke the bear!

[@MiniAirCrashInvestigation]

@@electricmaster23 You have no idea how tricky it is

[@MiniAirCrashInvestigation]

@@uzaiyaro Ive actually done a video on smartlynx 9001, you can find it on my channel :)

[@electricmaster23]

@@MiniAirCrashInvestigation Do you know if altering the title in any way messes up the algorithm, or is it just conjecture?

[@stt5v2002]

Also known as “engineers save the lives of everyone on board.” Thank you engineers!

[@mynameisgladiator1933]

*...after an incompetent pilot didn't know to use TOGA power, after Hamburg thinks it's OK to have a "special" relationship with their glide path Antennae.

[@SpearFisher85]

Ok. Ill give you this one. But don't get too cocky before I mention the old 747 cargo bay door latches. 😎

[@neilkurzman4907]

@@mynameisgladiator1933 I believe the narrator pointed out that a crash usually involves several things going wrong. And sometimes it only takes one thing going right to prevent the crash. The pilot, the copilot, both sets of controllers, all of the groups of people that decided to allow aircraft in an area that would affect the glideslope. And possibly the people who made the glide slope indicator. They all did something wrong. And all of those mistakes that could’ve easily led to an accident or counteracted buy a piece of software.

[@KutWrite]

@@SpearFisher85 ...and the 737MAX

[@deadbolt91765]

Engineering can be a somewhat thankless job. If you do everything perfectly, many people will never know that you exist... But your mistake can be front page news.

[@EvanBear]

We hear so much of automation causing pilots trouble so this is refreshing to hear!

[@MiniAirCrashInvestigation]

Exactly why I wanted to make this video

[@uzaiyaro]

But what you don't hear are the millions of flights happening every day where a potential swiss cheese model that would line up without automation, doesn't. Airliners are goddamn complex machines. The avionics bays are as large as some people's homes, and the computers contained within consume as much power as an office building. If you wanted to fly one by manpower alone, you would need a captain, a first offer, a flight engineer, a navigator, a radio operator, probably a second officer, where does it end? Automation cuts out a lot of the flight crew, and almost all of the errors they could possibly make. Statistically, fatalities due to automation are non-existent. In fact, more accidents have happened when pilots have failed to comply with the automation, than the automation itself crashing the plane. That is extremely rare, which is why the 737 MAX fiasco was such a big deal. Automation failures just do not happen.

[@turricanedtc3764]

Strictly speaking, what we're talking about here isn't really "automation" as most people usually consider it (they tend to refer to the Flight Management computers that control the aircraft's flight path automatically) - this scenario refers to the flight envelope protection logic which is built directly into the electronic flight controls.

[@turricanedtc3764]

@@uzaiyaro - Believe it or not, the electronic flight control computers in the A320 series are built around "hardened" versions of the Motorola 68000 and the Intel 80186 - CPUs that were effectively obsolete when the aircraft was being designed back in the mid-1980s. What the engineers needed was proven reliability over performance - the maths being performed in real time isn't particularly difficult, and much of the processing power is actually being used to cross-check against the other units. As I said to Marie, what we're talking about here isn't really "automation" as most people understand it - the Airbus electronic flight control system essentially takes pilot inputs along with the aircraft's current status and does its best to comply with what the pilot is requesting. From the A320 onwards, Airbus's selling point was that the EFCS enabled the short-haul A320 series to handle in a very similar way to the long-haul widebodies (A330/A340), which meant that airlines could save a significant amount of money on type conversion training for pilots. What happened with the 737 MAX was that Boeing tried to kludge a technical workaround (MCAS) into the design in order to make the aircraft handle like the earlier models (again, to minimise training costs and allow the aircraft to remain under a grandfathered type certificate). The problem was that the 737's fundamental design was laid out in the '60s and the airframe simply wasn't designed to do what they were asking of it - in effect, they were trying to kludge an electronic/software solution onto old-school mechanical/hydraulic flight controls, didn't tell the crews about it, and when MCAS behaved in ways the crews weren't expecting, they didn't know how to counteract it.

[@turricanedtc3764]

@Peter Wexler - If it's OK to ask, how do you mean "[tying] into the artificial horizon"?

[@johnt120]

I’m an airbus captain and I don’t know why this captain waited so long to add power. He performed what is called a “soft go around” which is used when you’r altitude is much higher above the ground. I will say it can get confusing if the glide slope is messing up, it may take a few seconds to figure things out, however if you’re that low, and something isn’t right, go around and figure it out with a lot of space below you!

[@elliott7268]

Agreed. Could be a case of "head buried in instruments".

[@Marlondurran]

Sounds good to me sir..

[@RaulGonzalez-pg1hr]

The captain did not performed the soft go around… he did not set the thrust levers at toga and then to climb detent, he neither performed a cancelled approach… push the approach push button and then select a vertical mode and lateral mode, just did not applied the correspondent SOP for that situation

[@thatguyalex2835]

@@elliott7268 Me personally, I would just look to see if the runway looks like a trapezoid out my windshield, keep my speed at 145-170 mph and land using visual approach. But hey, I don't fly planes. Lol...

[@noiserrr]

​@@RaulGonzalez-pg1hrsoft go around would be toga -> flx/mct and then climb once go-around thrust reduction altitute is reached. either way, was the soft go around even available on an a320 before the neo?

[@RmNrIHRoZSBDQ1AK]

While you do have multiple instances where automation caused a crash, the reason you don't have any of the opposite is because it works all the time. It's like intranet at a company; as long as there's a sysadmin, everything works. Then you fire the sysadmin because everything works and you don't need him. Guess what happens.

[@3zzzTyle]

Oofers, sounds like a personal experience.

[@stephanweinberger]

@@3zzzTyle It's called "survivor bias" and is actually a very common phenomenon. It's like the air force adding armor to the areas that were most often hit by Flak without any success, still losing plane after plane in combat - until someone had a moment of clarity and realized that the planes which where hit in *other* areas didn't even make it back to base to be examined.

[@Syclone0044]

@@3zzzTyle Happened to me too, personally. Well I wasn’t fired, rest is true. I realized the better I did my job administering our 20 servers for 650 employees, the more it caused employees to perceive that I did very little. I would routinely get random employees joking “What do you guys even do all day down there??” which was really insulting. I also realized the sad truth that if I were unethical, I could secretly manufacture a very visible crisis of my own making, and then promptly swoop down as the hero saving the company, and NOBODY would be the wiser, and I’d be hailed as indispensable. Of course I never did any such thing but I’m 100% certain that it occurs at least once in a while. There are a surprising amount of unethical or at least selfish and contempt-filled IT workers.

[@EShirako]

As a long-time I.T. Admin, whenever we have the "Do you people DO anything down there?" 'joke', I usually started looking for a new job. Idiocy usually arrives shortly after that starts to crop up. I move on, the company learns its lesson...again, in many cases...and re-hires admins until they lose enough managers to get 'Institutionally-retarded' and outsource IT support /again/-again. "It'll save us money!" Unlike the three previous efforts, which that person never even asked IT about, nor even looked at the situation that we'd all been hired to correct...namely (for one exact company's issue, naming no names) like one place needing to shout "Nobody use the CRM, the Domain Controller is down for a minute!" Why is the DC down? Because the print server died, so they need to reboot it to let the company print anything at all. Print-DC. Whatever. Why yell? Because it was already down/crashed/rebooting by then, and so 'network broadcast alerts' weren't possible even if they'd known how to do such a thing. Why did losing the DC matter at all? Because there was no BDC. There was one lonely DC, and a few valiant not-servers trying very hard to let the network survive off their cached Authentication data until the DC comes back again. Bonus Fun: The sole print-server-DC-mail-server-nightmare didn't always come back up because of a RAID backplane firmware bug that occasionally popped up (because that was never patched either, of course...) which made it need to be power-cycled to let it come up again. Five, ten, fifteen minutes of at-the-edge-of-crashing for this whole office. No isolated DC (like, the print server and Exchange and IIS and SQL shouldn't have been there...several of the services were unused, even! They used MySQL/Postgres, not MS SQL ever.), no BDC, bad drivers, unpatched servers...eventually I got the network *and* our IT group working SO well...that they joked about "What do you folks do over there?" and then they outsourced us. Again. And, ironically, called me a week later and begged for me to RE-make the 'In Case I Get Hit By A Bus' file that they had not just not accepted but had thought was a /legal violation/ on my part and almost wanted to sue me over. They made me delete it. MADE me. Sure, fine... Within the week, the BDC had been mauled (I never found out how), and my carefully-crafted SNMP communities and trap-domains had all been fractured and shut down because the dying BDC kept giving them "I'm dyin' here!" pre-fail alerts and they didn't know how to mute a repeat alert...so they just disabled and renamed SNMP stuff so the traps went nowhere-useful. CRM goes down, they can't figure out why it barfed. Augh. They made a hash of so much stuff in just two weeks. I hope they saved a LOT of money from outsourcing us to [OutsourceCountryOfTheMonth], because the company went out of business about a year later. But hey, gotta save them bucks!

[@kstricl]

@@EShirako I essentially got canned because I warned the company I was working for that a server failure was likely - 18 months before it happened. They moved so slowly that I was forced to do a week long rescue. When I was let go, it was the tip of the iceberg for all the people that were my biggest supporters - they were all gone within two years. Last tally I heard was that it cost close to half a million to remove me and the systems I built/maintained on a shoestring budget for nearly 10 years. Funny thing - it was a CRM with custom client interface that they were trying to move away from - into a CRM with custom client interface. I also did contract work for them for 4 years after they canned me, to keep existing system alive and in some cases to do further integration on the old system.

[@cdbtheclaw]

I was in a hotel really close to the runway 23 glideslope that day together with my father and brother. I guess an A320 landing in my bed wouldn't have been too nice of a surprise.

[@las10plagas]

reminds me of donnie darko (a weird movie were an airplane engine smashes through a house's roof, straight into someones bed... but there was no plane in the sky)

[@chillylizerd]

Was the hotel called "destination fucked"?

[@Ozymandias1]

This video shows that there are a lot of residential areas directly in the flightpath, it would have caused many ground deaths.

[@YanDaOne_QC]

Yo momma is an A320 landing in my bed....

[@madiekirkland3764]

I watch Air Disasters solely for the genius minds who have to solve what seems to be an impossible mission of putting together the facts that reveal what happened to cause the tragic crash.

[@fjlkagudpgo4884]

i admire the minimalism in your style, no flashy graphics and sound distraction, everything that's needed but nothing more huge thank u and keep it up

[@Syclone0044]

I agree. The one addition I could see benefiting, would be some medium/low-volume “sad, somber music” while mentioning fatal outcomes. Air Safety Institute’s fantastic crash postmortems have the perfect music for this, MACI, maybe you should check them out for inspiration.

[@MyOrangeString]

Yes! Exactly! Those TV reports are so flashy and void of substance, this tiny channel youtube is miles better!

[@HolySoliDeoGloria]

Agreed! There is way too much annoying background music in KZbin videos.

[@MIO9_sh]

I actually have quite a question here. To initiate a go around, isn't pushing the throttle to TOGA the first thing to do instead of pulling the plane up? The priority there just isn't going quite right cus as far as I know it, the TOGA detent is what telling the plane to go around, even you don't need TOGA power, you still need to push to TOGA, then back to MCT

[@HTown99]

I would think that if the aircraft is pitched too far downward then TOGO thrust could shoot the plane down into the ground. It should probably be a combined action of rotating the yoke backwards while the engines are spooling up their thrust

[@littlepastelkitten]

@@HTown99 airbus does not use a yolk system, they use sidesticks

[@MIO9_sh]

@@HTown99 in case you didn't know, planes don't pitch down when approaching the runway, their nose levels off and let gravity do the job for smooth descend. So, when TOGA activates, the engines will spool up, the aircraft levels off for a brief moment, and then a tiny input on the sidestick is enough to bring the aircraft climbing again

[@TimSheehan]

@@HTown99 they don't pitch down for landing, just lower thrust so that they're falling faster than the wings are lifting them up, so generally they could throttle to full and not change angle of attack to level off or climb slowly

[@markusloitfelder3469]

You do it simultaneously, advancing thottlesvto TOGA and as the aircraft accelerates instantly, pull back at the sidestick to pitch up. You never set thrust levers to MCT inflight as long as both engines are working. MCT is max cont thrust for one engine out, the same detent is used for Flex TO (on ground).

[@buskeiten3625]

Nice video. I have been flying A320 after a brief time of flying 737 and Embraer. I have to admit, I love Airbus and I wouldn't want to fly anything else. As for automation, I don't mind it. There is still a lot of things to do and take care of, even with it. Aside from that, the A320, even though a lot more automated, than Boeing 737 for example, does still give you some flexibility. In operated mode, you can control many aspects and without the Autopilot on, it is nice to handfly. The automation in my opinion does most of its job helping the pilots. For example, if you pitch the nose up 5 degrees after takeoff, the aircraft hold that pitch, even if you let go of the stick. The same is true with bank. Many of the automation programs are there to take workload of pilots and to prevent pilot error, which sadly happens a lot. For me, the airbus is just nice to fly and has the right amount of automatization, while still allowing you to fly it manually, if you want to. When I do fly a 737 in a simulator once in a while, it takes me quite a bit of time to get used to it and I get set up by several of its quirks. For example the process to enter step climbs is different and in the 737 you would "Arm" LNAV and VNAV as well as Auto Throttle before takeoff. The airbus doesn't require that at all. The Airbus also does not have a manual rudder with rudder pedals. I love the Airbus and I think it is just right as it is. This video shows that, because if those pilots would have been flying something else, they would have been in big trouble, likely not even alive. Sadly, the trend is going towards more automatization. I had a chance to peak at the A350 flightdeck and got to take a look at the flight manual. That's a complete different thing. On the other hand, it is also nice to be nostalgic. If I find time for flight simulation, I love to delve into X-Plane, flying things like the B727 with CIVA or the B732. Sometimes even the Embraer E-Jets, which while modern, yet are something completely different with close to no automatization. That said, I want to point out, that this is my personal opinion. I have been flying both and I love Airbus. It is just what I am used to, what makes sense to me and what I prefer. That does not mean, that Boeing is bad. The 737 have been around for several decades - even after the Airbus A320 was out - which wouldn't be the case, if it was bad. In addition, aircraft like the 777 or the 757 or the famous 747 are fabulous. In fact I like the 747 more than the A380, though I prefer the A340 over the 747. It is just a matter of preference. MAX has had a bad start, but Airbus had its flaws as well. At the end its like with a car brand or other things in life. And often it would also be what you are used too. Flying is complex and if you know how to fly one aircraft, eventually got used to it and all its quirks, you usually stick with it.

[@r.b.6592]

As an Airbus Captain with many thousand hours on the Bus, I can say that I never even got close to any of the protections that are installed. However, I am a firm believer, that these protections have surely catched the one or other pilot who was off guard and saved the day. Nevertheless, we will seldom find out the amount of times, when it was the pilot so saved the day by being smarter than the automation. Best bet: highly trained and proficient pilots in good working aircraft. Great video, thank you. Looking forward to more vids!

[@topethermohenes7658]

@ it's both, a very robust automation system and a poorly trained pilot for those automation doesn't work either queue the "what is it doing now?" conundrum. A robust automation system and a pilot trained to handle those systems is gold.

[@philismenko]

I would much rather a plane that always corrects itself over highly traimed pilots, the always is a major word there

[@topethermohenes7658]

@ as I said both go hand in hand. If there's no perfect humans then there's also no perfect system. Computer systems are also designed by imperfect humans no matter how complex they are. Computer systems are NOT the holy grail in aviation, it's both the human factor AND advances in these systems that will shape the future of aviation.

@@topethermohenes7658 , humans evolved to survive in particular environments. They did not evolve to have maximal abilities in an airplane's cockpit. On the other hand, technology is rapidly evolving. There's no need to invoke the "But it's not perfect!" straw man. No one expects perfection. But look at airplane accident data before automation versus the past decade or so. Add up the numbers. Aviation automation has been a thing for what -- 60 years from experimental deployment? It's in its infancy. Get back to me in another 40 years and lets see what's what.

[@topethermohenes7658]

@ but why do man have the need to evolve in an airline's cockpit when an airline cockpit itself is evolved for human use? Well I did not, you were the one who said that humans aren't perfect? Then an invention of a human then follows to be well not perfect. It's not only automation responsible for safe flights as I'm repeating again and again 😅 it's both! Human factor include new crm procedures, better atc, better training etc etc I'm NOT saying that computer systems did not help aviation to grow, it did, tremendously. But foregoing the strides that human pilots have sacrificed and overcome doesn't bode well with me. I don't know whatll happen 40 years from now, I'm not a computer engineer nor nostradamus to predict the future, I'm just a commercial line pilot who knows that the current key to safety is both me and the computer systems working in harmony

[@traemo]

I would love to hear more stories like this of bad situations that end well, whether by automation, great airmanship, or luck. Great video as always! As a student pilot these are very interesting to me. Thanks for doing what you do!

[@OscarScheepstra_Artemis_]

I love those videos! You are a great storyteller. You don't scream, you speak clearly, and your Flight Simulator videos are amazing to illustrate the video! I would love to see a video on Gol 1907.

[@Syclone0044]

Is “Gol 1907” the one in Netherlands where the guy had to force land in a small opening in a forest, striking treetops, with the fuselage breaking into 3 pieces, and yet somehow I think everyone survived?

[@OscarScheepstra_Artemis_]

@@Syclone0044 no, the one in Brazil. A Boeing 737, from Gol (a Brazilian airline) was flying over the Amazon rainforest, on a "dead space" for radar. On the other direction, an Embraer Legacy Jet was flying without its TCAS system activated, and at the wrong altitude. This caused a mid-air collision. The private jet actually survived with small damage. It seems that it was a series of factors that caused this accident - from the lack of radar coverage to the poor English skills of the Air Traffic Control team.

[@ellenorbjornsdottir1166]

@@Syclone0044 Sas 751(?) was in Sweden.

[@Gyrocage]

“If you ever find your self low and slow on an approach get your power up NOW! Don’t mess around, get it right up there immediately.” - Tony LeVier, one of many pieces of advice given in a talk at Wright Patterson in the 1980s.

[@ursodermatt8809]

that was 40 years ago

[@gunnermurphy6632]

@@ursodermatt8809 im glad you can do math

[@ursodermatt8809]

@@gunnermurphy6632 i am amazed you seem to be able to do math too!

[@Logarithm906]

@@ursodermatt8809 I just like the irony that the 80s are between 41 and 32 years ago. So neither of you can do maths :P

[@ursodermatt8809]

@@Logarithm906 i am glad you can do math. congratulations

[@tieck4408]

Reminds me of how the F-16's g-lock ground avoidance system recently saved a trainee pilot who had blacked out during a difficult maneuver.

[@BottleOJamie]

whilst the boeing MCAS actively tries to kill you

[@MultiJoe11111]

@@BottleOJamie LOL true 🤣

[@MultiJoe11111]

Woow is it? That's really something 👍

[@grahamstevenson1740]

g-LOC (loss of consciousness) !

[@EShirako]

@@BottleOJamie Your comment made me actually LoL aloud IRL. I don't do that a whole lot. :D

[@Barabyk]

It was airmanship/CRM issue as well. Reliance on automation is one thing, but they still had some tools at hand - approach charts come with recommended approach profile and altitudes/heights for particular distances. For 3 deg approach path it's around 320Ft/NM - they could cross-check their DME/Alt readings with glideslope to confirm they're at the right place.

[@EstorilEm]

Yeah this wasn’t an issue of reliance on automation- no Airbus driver ever gets into a situation where a fight protection kicks in on purpose. They definitely had an unstable approach. The incident report showed that of the aircraft hadn’t intervened, they were within a second or so of stalling. It’s also interesting that this is VERY close in flight profile to Asiana 214 that hit the rocks short of the threshold at SFO. An Airbus would NOT have allowed that to happen.

[@Bartonovich52]

@ Alex Thomas Different philosophies lead to different risks. Airbus Industrie 129 would have never happened to a 777 that preserves airspeed rather than angle of attack in its flight envelope protection. Air France 447 would have also never happened to a 777 because the pilot monitoring would have had full situational awareness that the pilot flying had the controls full back the entire time.

[@millomweb]

Maybe they should have gone for runway 33 as originally planned but overflown it then do a go-around to runway 23 - on the presumption this would give them time to reprogramme the plane for the new runway - and then reverted to their normal flying/monitoring jobs.

[@agps4418]

@@satunnainenkatselija4478 forgetting his responsibility due to some distraction, probably. he just wanted to land quickly, instead of land properly

[@madiekirkland3764]

The Plane That Saved It's Pilots was spellbounding in a good way. Thank you for a well spoken documentary!

[@coca-colayes1958]

What a fantastic explained video, I really like that you said the “ plane rudely interrupted” I’m surprised after learning from your video that it only takes this plane on taxi to interfere, with so much traffic it’s good I don’t hear much of issues like this , love your content even the adds

[@chris-hayes]

I imagine a team of engineers must've spent an absurd amount of time designing, building, and testing this system, a system they hoped was never needed.

[@KuK137]

And then there is Crasheing where all they do it cut corners and say 'oops' when a plane with 300 people rams the ground just because boss wanted to steal some more $$$...

[@KaitlynnUK]

I like how you do 'near accidents' instead of focusing on the actual accidents. As a sim pilot I find some of these fascinating.

[@EstorilEm]

It’s interesting to note that the system doesn’t have a set speed (speed protection is different than alpha prot btw.) Based on the speed decay rate and the engine settings, it can actually trigger the warning even sooner (or activate toga) if it knows that the engine spool-up time is now at a critical margin. This is critical because many pilots would have likely adjusted the throttles or even hit TOGA, but it would have been too late based on speed decay rate and spool up time. I think they plotted this out on a graph and the aircraft was within a knot or to (aka a second maybe) of potentially stalling. So yeah the speed protections are more for energy state, the alpha triggered on pull-up. It’s fun digging for situations where the flight protections kicked in - though I’m sure they are hardly ever documented. You only hear about automation getting people in trouble.

[@Syclone0044]

Wow that’s impressive and brilliant engineering. If I understand you correctly, when the situation gets dire, the computer code detects when they get too close to the limit, and preemptively applies throttle to get the engines “pre-spooled” so they’ll be much more responsive when the pilot applies throttle, compared to the typical up-to-5-seconds it can take an airliner engine to spool up from idle to full throttle?

[@Syclone0044]

The frustrating thing to me is, as even an amateur programmer and remote drone pilot, watching so many videos here and elsewhere that could absolutely have been avoided by almost trivially easy code. Such as: IF terrain elevation ahead is dangerously close to current altitude according to GPS topographical maps’ listed elevation; AND closing speed IS GREATER THAN (time until impact / safety margin); THEN sound alarm(“🚨 High Terrain ahead! Pull up!”); Another one that’s killed a ton of pilots especially in GA, here’s how I’d save tons of them: IF fuel_calculated_hours_remaining(currentAltitude, currentSpeed, currentWind, aircraft_glideSlopeRatio, distance_toDestination) < safety_margin; AND (currentAltitude / calculated_poweroff_glide_distance) < landing_distance_safety_margin; THEN soundAlarm(“🚨Critically low fuel! Fuel in danger of falling below minimum safe to reach destination!); alertPilot((calculate_destination_runway_length / highest safe landing speed)); AND engage_maximum_fuel_conservation_flight_parameters; // NOTE: these would be things like, auto engage fuel tank transfers, auto feather propellers on any engines stopped or at idle, flaps retracted until necessary, landing gear up until necessary) AND alertDestinationATC(“Emergency: Incoming aircraft $(aircraft_tailnumber) at or below minimum safe fuel, Estimated minutes remaining: $(estimated_fuel_surplus_duration_remaining), Current Altitude: $(current_altitude), Current Descent Rate: $(current_descent_rate), Estimated likelihood of reaching destination runway: (estimated_flight_time_remaining_byFuelQuantity / estimated_time_to_destinationRunway_touchdown)); Here’s one for GA pilots to avoid the death spiral: IF airspeed_Vmc > (current_airspeed * safety_margin); THEN alertPilot(“🚨 Stall Imminent! Speed descreasing too slow. Add throttle! Pull Up! Maintain speed above Vmc”); AND engage_stallSpinAvoidanceProcedure; // this would do things like automatically add throttle to avoid falling below Vmc (minimum controllable speed, below which you’ll likely lose control and the plane will drop into a steep banking dive straight into the ground), and using artificial horizon vs bank angle to auto correct out of a steep descending death spiral. These are just off the top of my head but you get the idea. In this day and age, it’s unconscionable to me that so many thousands of avoidable GA (and sometimes commercial) accidents occur due to circumstances that could easily be translated into computer code that prevents the pilot from continuing down a path of certain doom. I fly the old, original DJI Mavic Pro professional drone (about $1400 when it was new) and the amount of incredible refinement to their software to do everything they possibly can do to avoid allowing the user to crash the drone at high impact or other scenarios likely to result in total loss is amazing. They even utilize the 2 auxiliary downward facing digital cameras (mainly used to photograph the launch site, so it can engage a Precision Landing upon return to home, by matching the scene to the photos recorded upon takeoff to increase accuracy substantially beyond what GPS alone can offer, which works SO well that I tested a “helipad” improvised with colorful tape marks indicating precise takeoff point - and it routinely can land ITSELF totally hands-off WITHIN 3-6” of where it launched!!!). But they also use these downward cameras AND the downward sonar-based ground-detection/AGL sensor (“VPS”) to detect when the final landing surface is unsuitable for landing, whether it seems like tall weeds, uneven, or if it detects WATER, any of which cause the landing to interrupt and notify the user “Unsuitable Landing area. Please move to another location. Or hold down left stick to force landing.” There’s so much more than that, but point is in the ~3-4 years this drone has been out, they’ve basically PERFECTED the flight control software to where you have to be pretty reckless, ignorant, and/or foolish to crash your drone. And that’s an industry just growing out of its recent infancy the past 5-10 years! Why do SO many airplane crashes continue to happen under the most common and obvious scenarios where the recovery procedures are SO well established??? Under GA I can understand all those carbureted 1960’s and 70’s Cessnas with wood-paneled dashes just lack any automation, but what about all the rest??? I mean this IS CRITICAL, LIFE-DEPENDENT STUFF!

[@Stoney3K]

What is interesting to me is that the "speed, speed" aural warning in this case did not cause either of the pilots to instinctively slam the throttles forward into TOGA. No matter what other things you're doing, if you hear that warning, which is basically saying "You're going way too slow, punch it Chewwie!" you can always get to the thrust levers because they are within arm's reach. I'm also kind of surprised that they were flying the approach at flap 3 so slow to begin with, if you're flying the yellow zone in the speed tape, it means you have to keep a hawk-eye on your speed and one hand on the thrust levers... with the pilot monitoring ready to pop another notch of flaps when things get nasty.

[@-DC-]

Trust but Verify, Airbus have a good balance of assistance protection and pilot override ability.

[@Argosh]

To be honest, with Boeings shitty record on automation I consider Airbus the true pilots plane by now. All they do is enforce the flight envelope unless the sensors are fucked. Which is extremely safe.

[@albertabound5124]

The AoA protection works wonders. There’s a reason why every dual engine failure incident involving Airbus on this list (5 total) resulted in 100% survival en.wikipedia.org/wiki/List_of_airline_flights_that_required_gliding (Not saying non FBW MD/Boeing’s can’t, as we saw with the Air Canada and SAS “gliders”, just more challenging without the AoA protection)

[@BlackFoxFalcon]

@@Argosh A true Pilots Aircraft is a Piper Cub with the bare minimum of instruments that you need to have onboard due to regulations. Add to the mix that it`s a taildragger. Good Pilots know how to handle themselves and the aircraft without the aid of automation. Automation should be viewed as a seperate system that needs to be monitored aswell. While there is no doubt that automation saved a lot of lives, there is a reason why Student Pilots get trained on aircraft which most of the time, don`t even have a basic autopilot. There is nothing against automation. But, I don`t praise it as much as you do. That`s just the way I think. A modern day Pilots Aircraft would probably be something like a King Air 350 with the old EFIS 84 Avionics stack. While it already has more modern systems in it, it is part of the Complex Aircraft family. I personally like flying aircraft which depend upon the pilots making the right decisions and being ahead of the plane.

[@jamessheppard4372]

@@BlackFoxFalcon unless the pilot suddenly dies

[@Argosh]

@@BlackFoxFalcon a true true pilots aircraft is a glider. Whether it be the ASK21 or something comparable. Gliding is pure flying. Though you might also look at hang gliding. It's the truest form of gliding. So a hang glider is the true true true pilots aircraft.

[@bobbieratclif9519]

Your stories shared are great! As a heavy flyer all of my life, and after writing serious college level thesis about Aviation, your channel has shared unknown stories I’ve never heard about. Thank you so very much for sharing stories around the equator and south of the equator that appear to be left out of American news. Thank you!

[@mandywalkden-brown7250]

Thank you, another good one. The.notification popped up as I was fully immersed in reading an historical fiction series set in Ancient Rome, so i found your TOGA/toga reference a rather amusing synchronicity. I shall now return to the Forum ...

[@JW-vi2nh]

What series are you reading? It was a historical fiction novel set in the Republican Roman era that made me change my major in college and gave me a lifelong love of ancient history. It was called "The Memoirs of Cleopatra" by Margaret George. From there, the Masters of Rome series by Colleen McCullough blew me away and then, if you want something a little more "fun," I highly recommend the Roma Sub Rosa series by Steven Saylor, beginning with "Roman Blood." It tells, among other things, the story of Cicero's first murder trial. Absolutely fantastic! If you are interested in going a bit further back in history, I highly recommend basically every book Pauline Gedge has ever written, specifically "Child of the Morning," a book about Eqyptian Pharaoh Hatshepsut. Gedge even wrote one historical fiction novel set in Ancient Rome called "The Eagle and the Raven" and it is spectacular. It is about the struggle between the Romans and the people of Britannia, focusing a lot of the story of Boudicca. I LOVE historical fiction and would be thrilled to give more recommendations if you are interested. Enjoy your book!

[@dschroeder4380]

@@JW-vi2nh I love historical fiction too but I usually read more recent history. I'm really interested to check out the ones you suggested from ancient history.

[@joeb5316]

An audible warning of John Belushi chanting, "Toga! Toga! Toga!" would be amusing.

[@IdliAmin_TheLastKingofSambar]

And if one of the pilots accidentally hits the TOGA switch, you could have Hoover saying we can’t have a toga party. 😆

[@Milesco]

@@craigjensen6853 : What exactly is a "90's" voice?

[@peterkoch1676]

@@craigjensen6853 When the pilot gets a "Pull up!"/"Terrain!"/"Speed!" warning, the sound of the voice will be his or her least matter of concern in that moment.

[@jetaddicted]

Be cool with them pilots; even their aircraft is calling them retards upon landing.

[@te0nani]

The 737, blocking the ILS: "Ha ha, crash Airbus peasant!" A320: "Wanna see a real safety system?"

[@chibani-]

Air Berlin: "ze little frenchie iz gunna crash" Air France: "let me show you le pro move"

[@adamp.3739]

Virgin 737 MAX: "I'm greener... And more deadly than before!" (crashes with all dead) Chad A320(neo): "I'm gonna do what's called a pro gamer move. " (saves itself, pilots and passengers) "Much better than your MCAS bullshit."

[@Jet-Pack]

What I don't understand is why the MCT = *maximum* continuous thrust was not enough. Moving the thrust levers from idle or from CLB up to the MCT detent should disengage the autothrust and after around 5 seconds of spool up delay it should add thrust that is higher than normal climb thrust. MCT is about 90% of the available thrust, so TOGA adds just 10% on top of that. Keep in mind that with flap 3 a go around can be performed even on one engine at 100% thjrust. So with two working engines, each at 90% of the available thrust, that should be plenty to climb, that's still 80% more thrust than they would theoretically need in that situation! So something else must have gone wrong. I can imagine that due to the descent at 2900 ft/min that the engines were at approach idle setting, not at the stabilized 40-50%ish setting from which a go around is usually performed during training. The engine acceleration time from approach idle to 95% go around thrust is about 5 seconds as per engine certification. Approach idle is increased by the engine FADEC to meet that time constraint. That means either they added thrust too late or didn't actually move the thrust levers into MCT and kept it in CLB or they pulled up too soon or too aggressive. The activation of A.FLOOR probably just coincided with the time that the engines finally spooled up enough and the last like 10% thrust were added. Its not like the A.FLOOR saved them in my opinion. Yes it activated and added additional thrust were needed. But it's not like MCT would not have been enough there if it had been established in time. The thing is that the CFM engines have quite a slow acceleration from low power settings up to mid range. So the first 2.5 to 3 seconds only very little thrust is added. Then when the compressor finally gets up to a more safe stall margin thrust can be increase very quickly. The acceleration speed from mid power to high power is very quick, which is probably what they were used to from practice. But yeah, still not sure why 5 seconds delay caused the speed to drop that much. -

[@stephanweinberger]

I guess it's a combination of the high descent rate (because the pilot thought the plane was above the glideslope and had just pushed it down) which needed to be canceled out before the plane would start to climb, low flap setting (which increases the stall speed, resulting in less margin for a climb) and extended landig gear (which adds drag). Floor protection takes enginge spool-up time into account and counts against the rate (and presumably even the first derivative) of energy change - so when it activates this is actually the last moment before the plane's speed will drop below a safe level otherwise.

[@gasdive]

Seems to me the decay in speed mostly came from the pilot thinking "pull back makes plane go up". Throttle makes the plane go up, pitch controls speed. If he'd applied MCT and then managed speed with the pitch to hold best L/D speed he'd have got maximum rate of climb with the available thrust at a safe speed. Edit: obviously following company SOP would have been the best.

[@ampersand.]

You explained the situation so well even I could understand what had happened. Very interesting, thanks.

[@coca-colayes1958]

Nice , you are early with this video ,it’s 9:42 pm Sydney time and your video came out 11 mins ago , I’ll watch every second of. It , my favourite channel , Can’t wait to meet you

[@spauldo82ftw]

When learning instrument flying, you're taught to trust your instruments but they could always fail. Cross checking all the indicators is how to ensure everything is accurate. The problem here is that there was nothing wrong with the instruments. I think the only way the crew could have figured out an ILS problem would be to reference their distance with their altitude on the approach plate. If they were below glide path but the altitude shows they could be dead on would mean something isn't right. That's a lot to do in a critical phase of flight though.

[@koborkutya7338]

There seems to be way to many systems onboard and on ground, and each of them seem to have way too many modes of operation and possible failure modes. Then you have a computer running such cross-checks but it's still a computer so it can have its own failure modes you need to monitor. That day they would have uneventfully landed a DC-3 by visual.

[@EdOeuna]

Whilst established on the glide slope they should be seeing a vertical descent of about 800-900fpm. If they had to increase to >2000fpm to, for some reason, recapture the glide slope! Then something is seriously wrong.

[@CertifiedDynamite]

Bernard Ziegler deserved a drink on that day for preventing AF 2510 from meeting the same fate as TK1951.

[@turricanedtc3764]

Bernard Ziegler's title was Senior Vice President of Engineering, however his role was more-or-less sales-focused and an evangelist for the technology. The man responsible for leading the actual work on the A320/A340 EFCS pilot engineering and testing effort was the late Captain Gordon Corps.

[@toblexson5020]

This is definitely a case where failsafes are great. Having a plane ensure that it doesn't stall sounds perfect. There shouldn't be a case where the pilot relies on this feature, which is probably one of the best cases you can have. And I think generally from the incidents seen, the problems are primarily human based anyway. I wonder how many flights onboard computers have saved without the crews even noticed or considering. Enviromental sensors, gyroscope stabilised, assisted control movements, computer regulated engines. There's probably countless little programs running on even a simple aircraft that prevent situations that would have been regular occurences 50 years ago. I'm no pilot, but from these incidents, as well as others in other areas, I'm fairly confident saying that the problem with automation is not it existing, but when the human side of it. Some isn't tested enough, some is used with glaring errors, some is relied upon instead of being used for it's purpose as backup, and some is just used wrong. In this case, ILS was wrong, but the crew did their job of flying the plane, and aborted the landing, and the plane's computer did its job of catching the crews' mistake.

[@JamesF0790]

Agreed entirely. The two 737 Max issues were as much pilot training as anything and that was on the airlines. MCAS should never have relied on a single sensor but it would have been avoided if they'd known to disengage the system.

[@baguskusumaloka]

@@JamesF0790 if the plane not use single sensor, if boeing give proper information about this and give a guide for pilot. Maybe we have different story

[@JamesF0790]

@@baguskusumaloka Oh to be clear Boeing is not blameless, I'm not saying that. I'm saying that there's a mix of blame here.

[@thorstenschoop205]

Just take a look at flight QF 72. It shows a different picture...

[@LunaticTheCat]

@@JamesF0790 It's still crazy to me that Boeing allowed the 737 Max to go into production with the MCAS system relying on a single sensor that has a history of occasionally failing.

[@boahneelassmal]

just a quick note: you don't necessarily need TO/GA for a go around. MCT should in most cases suffice. It's important though, that you don't pull up immediately but rather let airspeed build and then pull up. Now, the thrust applied depends on GW and altitude above ground. A new feature introduced with the neo series is something called soft go around. Here you won't demand TO/GA thrust but rather MCT allowing for a more comfortable, slower and engine saving go around. But, even soft GA is initiated by moving the thrust levers to Toga and then back to MCT. You can't just intiate a GA by setting thrust to MCT itself. Long story short: you don't necessarily need TOGA thrust but you need TOGA thrust setting to initiate GA. Lastly: If the PM was distracted from monitoring because they still hadn't set up the MCDU correctly the question should be asked whether they were good to accept the approach in the first place or whether it would've made more sense to request delay vectors.

[@skipsoft]

I don’t fly big airplanes yet, but one thing I feel like wasn’t mentioned in this video was it is the pilots job to brief the approach and actively pay attention to your distances and altitudes. Even if the glideslope was giving you an anomalous reading you should be able to look at fixes along the approach and say wait a second this is not right.

[@HolySoliDeoGloria]

That's astonishing that the flying pilot didn't set go-around thrust. I haven't flown any Airbus aircraft, but from the first airplane that a student pilot flies, one is taught to set go-around power or thrust as an absolutely necessary part of the go-around or missed approach procedure. It should be second nature from the early days of flight training.

[@Taletad]

Even higly trained professionals make dumb mistakes some of the time. It just came at the worst moment You will probably miss a critical item on your pre flight checklist a few time in your pilot life if you fly often, but it will turn out fine because the plane was in good condition anyway. And you will never know about that oversight. That is why there are systems in place and redundancies to make sure one failure or oversight doesn’t crash a plane

[@jamesturner2126]

The automated TO/GA saved the damn plane. Great video on a seriously near miss. Everything from the glide slope issue to the 373 taxiing too slow.

[@nikiandre6998]

Like always, you did amazing job! Beautiful picture and good storytelling! I like your channel were much!!!

[@00muinamir]

Honestly, the question of automation vs. none isn't the first thing that comes to mind for me here. ATC dropped the ball and that's what got this scenario rolling.

[@Timoteo53575]

@Christina T. I believe at this time, the possible effect of a nearby plane by the ILS was unknown. ATC may have assumed (ignorantly but not carelessly) that #2510 was aware of the glide scope problem. Which they were NOT, due to interference....

[@ShimrraJamaane]

A lot of balls were dropped. The poor design of the ILS that it can be interrupted by ground traffic, ATC training or error, and failure by flight crew to properly initiate a TOGA procedure. Had any single one of these not been a factor, then there wouldn't have been a serious issue.

[@KelseyDunlevy]

Placing the blame only on ATC isn't helpful in this situation. This was a Swiss cheese situation where many things were going wrong. I'm glad that the automation didn't allow then to pass through the final hole of the cheese and crash.

[@freedomthroughspirit]

Nice to hear about a plane actually helping the pilots instead of working against them like more recent Boeing software incidents.

[@kylebroflovski5333]

Really liking the new footage with flight sim footage, really helps with understanding

[@zadman49]

I’ve been watching and listening for a while and while you have always been a good narrator I feel like you’re even better know. Your narration is very natural and easy to understand. Keep it up! 👍

[@EvanBear]

I still think a "trust but verify" attitude would be best. Automation is a great help but it needs to be supervised closely by pilots. I think pilots being vigilant is still the most important thing in air traffic safety.

[@matthewwilson5019]

Yup and this proved that just fine

[@EstorilEm]

@@matthewwilson5019 Exactly - they were out of options here, the aircraft only kicked in at the last possible second. The report on this is actually pretty scary. They really should have hit the toga button when they heard the “speed speed speed” warning.

[@matthewwilson5019]

@@EstorilEm not only that it was traffic control tower and the signals were blocked by an airplane so, if the airplane had crashed it would have been air traffic controls fault

[@Bartonovich52]

That’s literally how pilots operate automated aircraft. It’s actually harder than hand-flying in most cases.

[@tomaszenko2080]

Well yes, but human is a lot worse in monitoring that acting. When making a GA with no autothrottle it is easier to not forgot about power.

[@Fir3Chi3f]

That's really interesting and super great analysis! We usually don't hear about these sorts of things when there is no loss of life. I'm super thankful for the engineers at Airbus!

[@jacekatalakis8316]

I'll copy my comment from a DotC video: There's always something interesting to watch on this channel

[@MiniAirCrashInvestigation]

Thank you :)

[@Syclone0044]

@@MiniAirCrashInvestigation Your channel’s notoriety is growing. I have seen at least 2 recommendations for your channel in comments on random aviation videos. One was Air Safety Institute’s top-notch crash postmortems.

[@ericcss7]

Another very insightful video also you are a great storyteller making your videos very enjoyable. Keep it up

[@caleb0606]

Automation is so great. It worked exactly as designed! That is great.

[@lightningstrikestwice6302]

While I would never be smart enough to ever fly an airplane I've always loved them. Listening to these stories is enlightening and scary at the same time. The narration is perfect! He has such a pleasant voice and is so knowledgeable! Such a refreshing change from a lot of the stuff on KZbin and the news. Thanks for taking the time to do these. Makes me smarter every day. And that's quite a heavy lift!

[@justmoritz]

I think next gen automation needs to have a feature where it brags about saving the day on twitter or something after an incident :p

[@minikipp8549]

😂 exactly what we need

[@thomasmcdermott132]

This is very well researched and explained. Long time watcher, just decided to subscribe. This video converted me into an Airbus fan.

[@essbe7158]

It's a tough question... This is a story of automation helping a plane...you have multiple instances where automation has caused the crash, especially in those 737 max situations with MCAS. I guess the answer is automation and failsafes like alpha floor are important, but pilot training must be kept up, and manual flying skills must be stressed as well.

[@EstorilEm]

Airbus has designed systems like this into their family since creation- its predictable and doesn’t intrude on flight ops. In fact something like MCAS would never be required on an Airbus as their core control and protection laws already prevent it from happening. Even if they wanted to program in a tweak for engine location mods (like the max / MCAS was for) they could do it with software. The Airbus systems are already triple-redundant.

[@danilon3121]

Automation is why we now often go years between air accidents in modern airlines that operate new planes. Your reference to MCAS is misplaced as this wasn't really a failure of automation, but a failure by Boeing in trying to rush their product on the market, while selling it as identical to the prior generation 737. If they had installed a fail safe, as is the norm in the industry, and if they had have trained pilots properly with regard to MCAS, there would never have been any MAX crashes. In any event, the number of cases where automation has truly failed in recent history are virtually non-existent, and usually the real issue was human error in maintaining planes, or in responding to problems in flight. The most important thing in all of it is that for every 1 crash caused by automation, 99 crashes are avoided because of it.

[@Bartonovich52]

@ Alex Thomas Lol “predicable” Like hopping out of Normal Law without telling the pilots who then try and use the now disabled Alpha Floor and Alpha Protect to fly out of the stall for 38,000 feet before crashing into the South Atlantic. Nice try.

[@bennylofgren3208]

Bartonovich52 Any airliner is an extremely complex piece of machinery. The software running these systems is equally complex, if not more. I’ve been a software engineer for close to 40 years, I know exactly the kind of complexity we are talking about. Add to that the man-machine interface that needs to be logical and intuitive and easy to use in general. Sometimes, we can’t predict every possible scenario, and then things go wrong. A bug in the system isn’t an inherent problem with the technology. Even with a number of accidents caused directly or indirectly by those flaws, FBW-automation is still safer than normal manual flight controls by a large margin, simply because it gives the pilot additional help exactly in situations such as this one, where they were saturated with other tasks and forgot to fly the plane. There is a reason Boeing too now implements this technology as well, starting on the 777 and 787.

[@millomweb]

I don't think pilots now do enough hands-on flying. Real experience gets embedded and you get used to what's a normal response to input and what sort of response you get from flying slow. Then you can learn to use flight control input to assess real airspeed irrespective of what the instruments are saying.

[@mellanyvanderboon2046]

Very well explained! I throughly enjoy your videos. Thanks so much and look forward to watching all of them!

[@tomstravels520]

I’m assuming that the pilot was operating the thrust levers manually. Otherwise if they had been left in SPEED mode when he dived the autothrust would have idled and the reapplied power when pitching back up again

[@adb012]

In the Airbus planes autothrust doesn't command more than climb thrust. There is one thing the video doesn't mention: Boeing planes have a TOGA button that you press to initiate a go around and set the plane in go around mode. Airbus doesn't have such button, the way to do it is byt moving the levers fully forward to the TOGA detent. So not only that the pilots didn't apply thrust as you would do even in a Cessna 172 for a go around, they didn't activate the go around mode either.

[@tomstravels520]

@@adb012 no it wouldn’t (unless on OEI mode at MCT) but with autothrust engaged it would have automatically maintained (as best it could) the speed selected or at VAPP. Even I’m with autothrust engaged the aircraft should not reduce below VLS which is way above Alpha Floor

[@adb012]

@@tomstravels520 ... Yes... but... The way it works in Airbus is like this: If the thrust levers at above CLB, then the selected thrust prevails, no matter if the AT is on or off. If the thrust levers are at or below CLB, then with the AT off the selected thrust prevails (the automation can still do some things with the thrust even if the AT is off, like adding thrust if you reach alpha floor), and if the levers are at or below CLB with the AT on then the AT will select the thrust according to what is required for the selected flight mode (like hold an airspeed or a vertical speed) *up to* what is selected with the levers. SOP is that after take-off, when you reach the thrust reduction / clean up altitude, you put the levers in CLB with AT on and it remains like that for the reminder of the flight up to the point of the flare when the plane calls you RETARD and you idle them for touchdown (note, if the logic of the AT was not to limit the max thrust to whatever the thrust lever is set at, idling the levers in the flare would have no effect, the AT would still keep the thrust to maintain approach speed and even increase speed to keep APP speed through flare). At 2:00 the video said that, because the were way above the glide slope, the pilot started a steep descent and pulled the power back. So whatever the new setting it was after he pilled back (idle or somewhere between idle and climb), that was going to be the max thrust that the AT would apply. So yes, the AT would try to keep the selected speed (not less than VLS) by applying the necessary thrust *not exceeding* whatever setting below CLB the thrust levers were set by the pilot. Now, why would the pilot pull back on the thrust levers to initiate a descent? Good question, there should be no need, the AT would reduce the thrust by itself to keep the selected speed. The only reason I can think is that it will take a little bit for the AT to react since it reacts to trends in airspeed (so the speed would need to achieve an increasing trend before the AT starts reducing the power). The pilot probably wanted an immediate reduction of thrust while he was pitching down so he pulled back manually. Now pilots normally understand that there is only one "person" handling the thrust. Either the captain, or the first officer, of the autothrust. If you are in autothrust and you start fiddling with the thrust levers, you are doing something wrong. Want to take control of the thrust? Ok, disconnect the AT. Want the AT to start controlling it again? Ok, connect the AT and put the levers in CLB. I have the subjective idea that the Airbus logic for the AT somehow incentivizes this blended behavior or being more or less in AT but more or less in manual thrust too. In a Boeing, where the AT physically moves the thrust levers, you have to physically wrestle the AT to put the levers where you want and then you will need to wrestle to keep them there too because the moment you let go, the AT will put them back to wherever the AT wants. So with the AT on the AT is always in charge unless you are physically overriding it or turn it off.

[@tomstravels520]

@@adb012 I know exactly how the autothrust works. What I don’t get is what point you are trying to make regarding my OP. If the aircraft had autothrust left on it would be maintaining the speed as best it could (pilot could have dived so steep even idle wasn’t enough). But seeing as the speed managed to bleed off all the way past VLS to ALPHA floor then the autothrust would most likely have been off or it would have tried to increase thrust to maintain the speed required. Unfortunately I can’t find a final report in English which details what happened

[@adb012]

@@tomstravels520 ... Well, first of all, don't take all that I say as a direct response to you. You may very well know how the AT works, but many (most) reading the comments won't, so I always try to explain and support what I am saying with background information. Second, I (maybe wrongly) took your original post as implying that the AT must have been off. But on a second read what you literally said is that he was operating the thrust levers manually, which does not necessarily mean that the AT was off. If you ask me, that is still a bit ambiguous, The pilot may have operated the thrust levers manually once, a split second, bring them to idle or a low setting when they started to pitch down, and never touch them again. Then, sometime later, when they pitched up again, the AT would "try" to keep the selected speed but would be unable to go beyond the setting left on the thrust levers. At that point, and for some time, the pilots WAS not operating the levers manually. He DID operate them manually once in the past. But if what you meant is that the pilot must have reduced thrust at some point, then we are in perfect agreement, but that adds no new information since that was stated at 2:00. Yes, I realize now that "autothrust would have automatically maintained (as best it could) the speed selected" includes the possibility that it didn't do anything because it couldn't because the pilot had left the thrust levers at idle a while ago ("I did nothing at all" and "I did the best I could" are not contradictory when the best you could was nothing at all). But understand that this was confusing to me, and I am sure it is even more for the readers not familiar with the operation of the autothrust in the Airbus family of planes. Now, to add more to the confusion, read to the (almost) last thing you said: "But seeing as the speed managed to bleed off all the way past VLS to ALPHA floor then the autothrust would most likely have been off or it would have tried to increase thrust to maintain the speed required." Where is there the (in my opinion) at-least-as-likely option that the AT was ON but the pilot had retarded the thrust levers partially or totally earlier in the dive and left them there, so while the AT would try to maintain the selected speed it couldn't because it would not command thrust past the now retarded thrust levers setting? I am sorry if I am getting you wrong. I cannot say that what you say is wrong but it is at least very confusing, ambiguous and not clear to me.

[@BillyAlabama]

Your explanations are so clear to this non technical guy. Thank you

[@talk4dews]

There's a video of a fighter jet saved by Auto GCAS on KZbin. Automation, when it works correctly can definitely save lives.

[@Ananth8193]

Every time this channel never disappoint it's subscribers ❤️❤️❤️.... Loving all the contents

[@MiniAirCrashInvestigation]

Thank you so much!

[@carlschneider689]

Looks like clear weather. I'd always trust what I see.

[@damaliamarsi2006]

I think the correct answer is "it depends". There are cases where ignoring automation has caused crashes and cases where following it has caused the same. I think the best answer is to always review and always treat any system with the thought that no system is perfect and pilots and engineers alike should always be challenged to ask "what if" questions and even try to find a way to break the system. MCAS is a good example. Boeing threw a technical automated system in to fix a design limitation and didn't notify or give pilots the ability to override. Training and excellent and clear documentation with an always "never trust" mentality towards automation is a good way to go. Automation lessons pilots work loads and in this case provided a reaction or faster reaction time. It all really depends on the flight crews situational awareness as to how they are going to react to these situations when they occur. Great video, thanks for taking the time to inform educate and entertain.

[@youtube-kanal2606]

I love how the airplane has the registration MINI-AIC.

[@pop5678eye]

Thanks for this upload. Far too many people get the incorrect impression that automation is dangerous because of a number of incidents where it was a factor in a crash. However the countless times it saves lives are rarely mentioned because that doesn't make for flashy headlines.

[@quasarsavage]

Airbus: Here Imma save u Boeing: Naw fuck em let's kill em PITCH DOWN YEET MODE

[@vizender]

That’s a golden rule I’ve learn while in training in small planes : when you go around, first engine at full power, and then the rest.

[@kylebroflovski5333]

These are interesting, even a successful landing can have a lot of interesting content

[@dschroeder4380]

I love your videos! The narration makes it so much more pleasant to watch cause of eye strain. I'd love it if you did one on the Andes tragedy (I forget which flight that was)

[@Robbanswe96]

First of all, Great job on the video! Perfect! So with the automation things. It's there to remove stress on the pilots and increase safety for all aircraft, here we see that the flight computer is to advanced and take to much time but at the same time the alpha floor protection work beautifully and save the aircraft from a crash induced by a stall. So the FMS should be easier to controll and the ATC should have been better at their jobs. Airbus saved the day there Well done video! 👏

[@DoolittleSailing]

ILS or not, a pilot should always monitor his position. Every ILS approach chart has published altitudes to the respective DME distance. If there is any doubt about the correct position, Go Around. The GA procedure is very simple and standard at the 320, so consider me surprised, that TOGA thrust wasn’t the pilots first impulse. The Flightplan in the onboard navigation system is split into flight phases. During the approach the active phase is the Approach Phase. Normally you can’t get out of that phase with only two exceptions. Either land or select TOGA thrust to get into the GA Phase, which usually includes the published missed approach procedure. Even if TOGA thrust is not needed, just select it for a split second to get the correct sequencing. Well, that is an odd incident, but as someone who worked onboard this aircraft, I’m happy that Airbus takes care of our mistakes.

[@mistiffiecation]

Regarding automation: last year I had a class that touched on av safety topics, workloads, etc. Also automation. I remember a chart, that basically said that you have best outcomes, when operators of (partially) automated systems are aware of pros and cons. Thus they don't undertrust or overtrust those systems. As you said - there's no right answer, really.

[@jblyon2]

Agreed. Automation and protection systems are great, but the pilots need to be aware of exactly how they work, what they're designed to do/not do, and how to override the system if necessary.

[@robbiealixsantos]

It's so nice watching this one after binging a bunch back to back. The relief that I felt when they decided to proceed with the go around, wow lol

[@karllung2649]

Isn't that common sense that the first thing you do for go around is to add power? With the increased power, the pitch should go up even without pulling back.

[@turricanedtc3764]

Correct. The Alpha Floor protection is designed to kick in should a pilot (for whatever reason) forget.

[@gasdive]

I'm constantly amazed by the "pull back makes plane go up" idea that so many pilots have. Pull back makes plane slow down. Throttle makes plane go up. When you're doing a go round you add throttle... Or so one would expect, but time and again, from microlight to jetliner, they seem to fail to understand what that control does.

[@karllung2649]

@@gasdive Actually should take a look at the crash of China Airline 140 (CI140) at Nagoya in 1994. The co-pilot accidentally hit TOGA. Without knowing this, the pilot decided to go around and pull up, causing the plane to nose up steeply and stalled.

[@seeingeyegod]

assuming he thought he had enough power in to climb, which he quickly realized he didn't from the "speed, speed, speed" warnings, it begs the question why he didn't immediately go to TOGA instead of max continuous thrust. Does going to TOGA look bad on a pilots record or something?

[@gasdive]

@@karllung2649 wow. Do everything wrong.

[@BlackwaterBud]

Excellent job of explaining concepts non-aviators might find confusing.

[@rilmar2137]

So much human errors and bad CRM, thank god for the computer's great airmanship!

[@quentinzerr3929]

Great video! Many parallels with Emirates Flight 521 and I wonder if that crash would have actually happened if it was an airbus instead of a Boeing.

[@cpt_nordbart]

used a bit of google fu to try and find some news in german about this incident. But there hasn't seem anything written about it. At least nothing that i could find easily.

[@swissTom124]

Something like this normally won't appear in newspapers because it's too complex for normal everyday people and nothing exiting happened. This was probably more an investigation and since all pilots should understand it, English is the best language for it. Most universal knowledge is written in English.

[@pontmercy8]

Excellent presentation, I’ve been following your channel for a while. Interesting and engaging analysis

[@havoc23]

Another proof for my unconditional love for this airplane. Funny, that I as a mediocre Flight Sim Pilot would never initiate a go-around without going to Toga or MCT power. One question remains: Why does the Plane go missed on RW23, goes around and then lands on RW33, where they initially wanted to land?

[@tomstravels520]

Probably because of the faulty ILS but they didn’t realise at the time it was due to be blocked

[@killingfields1424]

In Airbus Industrie, its their engineers and technicians built their airplanes, not their marketing personnel as in Boeing or any American Aircraft Manufacturer

[@dmor6696]

i like how you explained why they fly against the wind "because it makes things a lot easier" ahahahahahahah

[@Robidu1973]

The point is, since the wind is coming against you, although you have a sufficiently high speed relative to the wind, your speed relative to the ground is reduced proportionally to the wind speed. That essentially allows you to stop much sooner.

[@BloodSteyn]

@ 747 comments, really didn't want to mess that up, but I just want to say, great video, lovely use of MSFS, great narration. Well Done.

[@williamgeorgefraser]

I have so much more trust in Airbus than Boeing. I'd never set foot on a Boeing, especially the 737 Max.

[@daniellamb7828]

Lol

[@ryanfrisby7389]

Awesome video as always man, this further confirmed my belief that we do need automation in aviation, though I still want my pilots as well!😸😸

[@DoDo-dq7yf]

I just started working for Airbus this week... so thanks, I guess But no seriously I don't work on the avionics

[@INSOMANIAC911]

Awesome video my friend looking forward to another ☺️

[@Argosh]

Imagine brain farting so hard you call Go Around but forget to put in thrust 🤪

[@Argosh]

@thetimekillerx that's called "being irrational". Because outside of mistakes like these, which are addressed by CRM changes and training, it is still much safer. And no, pilots in commercial aviation do know how to fly.

[@Argosh]

@thetimekillerx and it happens far more often in general aviation than in commercial aviation. Hence the massive difference in accident rates...

[@MrPig40]

This was a good one! Well presented, as always.

[@wolfgangwust5883]

Clear pilot error. That has been a close call. In remembrance of AF447.

[@bennylofgren3208]

Not entirely. This incident was induced by the ILS glideslope problems, unknown to both pilots and controller at the time of the incident. But yes, in an ideal world they should have caught up to the uneven glide path earlier. But in the end, thanks to both man and machine, history describes this as an incident and not an accident...

[@patriciaramsey5294]

Really like this. Please do more vids of near crashes.

[@snowfoxsean]

Its*

[@RaivoltG]

I'm not a pilot but I've heard a phrase something like "trust but verify." Proper procedures must be followed too, adding power to climb in a go around is a necessity I believe. Like you said, every situation is unique, has to be handled in a unique way. I'm glad everything worked out for this flight! 4:20 Are you sure TO/GA isn't an ancient Roman power? I think that's what they want you to believe! 🤣 Great video, I love your channel!!

[@killingfields1424]

Airbus engineers were really doing their job the best they can. Exactly the opposite of the competition

[@turricanedtc3764]

That's not really fair - as much as I'm an advocate for Airbus, I have to admit that the B777 EFCS design was really bloody nifty.

[@Bartonovich52]

Yeah... Airbus Industrie Flight 129. Had Airbus made flight envelope protection based on airspeed rather than angle of attack, everyone on board would have survived. Air France 447. Had Airbus put a control yoke rather than a side stick in, everyone would have survived.

[@turricanedtc3764]

@@Bartonovich52 - Re: your claim about connected yokes. If that's true, how do you explain Birgenair 301?

[@tomstravels520]

@@Bartonovich52 Atlas air 3591, if that aircraft had a pitch protection like Airbus or didn’t have a button that could accidentally engage TOGA then everyone would have survived

[@WingKLok]

Thank you for the video. I recall someone mentioned elsewhere that it is a design philosophy between AirBus vs Boeing that perhaps causes or prevents mishaps - in this case, AirBus’ systems was designed to save itself, vs Boeing’s products relies/values the flight crew’s decisions (and inputs) over automation, MCAS aside?

[@pizzablender]

I suppose predictability is important here. Automation should be known and predictable. Think ABS or ESP in a car, those are always there, but only in specific conditions that will not occur normally. Now imagine a lane following system that sometimes gives up, for example when the road color is wrong. That will surely lead to accidents.

[@karthika....9884]

First again

[@MiniAirCrashInvestigation]

Confirmed first!

[@karthika....9884]

@@JockMacHH WELL THANK YOU 😊

[@AlessandroGenTLe]

What's wrong with you guys (meaning people commenting 1st! Or "It is so early then...")? Who the f*ck cares about who put the 1st or 2nd comment??? Sometimes is better to shut than to give a demonstration of the proper stupidity.

[@KyrilPG]

@@AlessandroGenTLe Thank you, that's exactly what I think every time I see a moron commenting "first". They have to be so desperate for attention that they want to be recognized as the first to leave an absolutely useless comment. It's so childish, it reminds me of the worst of kindergarten.

[@coca-colayes1958]

@@MiniAirCrashInvestigation first to comment ? Or at watch video? Who was first to watch video and the adds and then comment ?

[@christopherwaltercheetham5905]

I like this mini video, was well informed. Going to be honest, I’m going to miss “it’s brighter here” though after watching a horrific crash

[@neilpickup237]

I have been in an A320 which had to do a go around at the last moment (wind shear according to the announcement later) and we were certainly aware that everything wasn't good. There is something quite unique about appearing to just hang there with screaming engines and then (unlike a normal take off and initial climb) make a very slow ascent to the go-around approach.

[@EdgyNumber1]

Sully would not have landed in the Hudson as cleanly in a 737. His idea of starting the APU was genius. It meant that the A320 could assist his landing by keeping him just away from the edge of stalling by varying the pitch angle on approach to his dead stick landing. Right pilot, right aircraft.

[@KYEMattFiBeR]

This recently happened to a South African Airways flight out of Brussels, where they had their weights completely wrong, and had their departure settings completely incorrect. Toga was set on departure due to Alpha Floor protection. They later said the plane "lied"....

[@josephcameron530]

Great video. Thank you for providing this video.

[@fr89k]

Why not automatically disable localizer if an airplane is in the critical area? Having no data should generally be preferred compared to wrong data...

[@nebukadnezar4431]

funny to hear that the FO who was "busy monitoring the speed" did not notice the speed drop below a critical threshold ...

[@jpmthecvrrecorder4462]

this is the great channel for Air Crash Simulator and Simulation, for sure :-)

[@jeremywaller4077]

The title is correct, it's is only a contraction of it is. Great video

[@tirsocanales8366]

If you go around on A320 the procedure is Toga flaps one step up positive climb gear up, check FMA toga GA TRACK , if you are high you can do a soft go around which is toga to arm go around mode and then climb to arm AT, you can push NAV to fly the published GA or HDG for radar vectors , wx avoidance , etc. Nice video!

[@tomstravels520]

Actually it’s TOGA then FLEX/MCT for soft go around. You can also do it at a low altitude if lightweight as long as terrain isn’t an issue

[@bullwinklejmoos]

@@tomstravels520 It’s TOGA then CLB from where I’m from.

[@tomstravels520]

@@bullwinklejmoos it doesn’t change depending where you’re from. To get soft GA mode it’s always TOGA then immediately FLEX. TOGA to CLB is the normal go around or for a TOGA takeoff

[@bullwinklejmoos]

@@tomstravels520 At my airline, FLEX/MCT is used only for takeoff or single engine operation.

[@tomstravels520]

@@bullwinklejmoos yes…..like nearly all airlines. But you can also use it for a go around if the aircraft has soft GA installed. Thrust levers to TOGA to active GA mode then immediately back to FLEX, SOFT GA enabled and aircraft will climb at 2000ft per minute until reaching acceleration altitude

[@Marten_Zeug]

0:20 I actually live in Hamburg, Germany. Is a nice City!