[BoneCollector]

It seems that the Lion Air crash and Ethiopian crash have similarities between them.

The first concrete evidence of a possible link between two deadly Boeing 737 Max crashes came from space.

A new satellite network capable of tracking planes in high fidelity across the globe captured the flight path of the Boeing Co. 737 Max that crashed Sunday. The data was critical in persuading the U.S. to join the rest of the world in grounding the jet, according to industry and regulatory officials.

The erratic, six-minute flight of the Ethiopian Airlines plane convinced the Federal Aviation Administration that it was close enough to what preceded the Oct. 29 crash of another Max off the coast of Indonesia to warrant concern.

After reviewing the data “it became clear -- to all parties, actually -- that the track of the Ethiopian Airlines flight was very close and behaved very similarly to the Lion Air flight,” agency Acting Administrator Daniel Elwell said Wednesday.

Link- https://www.hindustantimes.com/world...pPtV5kq2J.html

[superbad]

All said & done, this just goes on to show that scope for any error in planes (human or technical) is ZERO.

[A350XWB]

Originally Posted by arijitkanrar

Completely agree with this assessment.
I'm not saying Boeing is not to blame for the Lion Air crash, but at the same time this begs the question "What were the pilots thinking?".
We were not in their place and we can and should not be judging their actions.

Not so simple. Like I told earlier, the main criticism against Boeing is that they kept the introduction of MCAS, kind of secret. You can react to a problem correctly, if you know about it. According to the news, there is no mention of MCAS in the FCOM (Flight Crew Operating Manual). Boeing thought that this might be too much of information to the average pilot.
From this article.

The airplane’s nose can tilt down suddenly during takeoff, pilots aren’t being adequately trained on the autopilot system, and the operations manual is “criminally insufficient.”

Another article.

A multi-page document issued by the airline’s flight operations department that highlights the differences between the MAX and 737 NG does not mention MCAS or any other changes to the auto-trim system.

Interesting read:
About MCAS System
Complaints filed by pilots about the MAX.

[HyperRetard]

https://www.nytimes.com/2019/03/14/w...-airlines.html

[srini1785]

I am of the opinion that technology is good only up to a certain point. The more complex you design, more problems. A case to the point is the F35 fighter which cost billions in development and when released also created problems. Correct me if I am wrong.

[longhorn]

When the Lion Air plane crashed, all pilots flying the 737 Max 8 would have analysed the information available in public domain. Chances that another pilot made the exact same mistake (if at all the first plane went down due to pilot error) is likely to be very remote, that too, something as simple as switching off two buttons. Somehow things don't seem to add up. I'm guessing there's far more to this than what is publicly known right now.

[Zappo]

Originally Posted by longhorn

When the Lion Air plane crashed, all pilots flying the 737 Max 8 would have analysed the information available in public domain. Chances that another pilot made the exact same mistake (if at all the first plane went down due to pilot error) is likely to be very remote, that too, something as simple as switching off two buttons. Somehow things don't seem to add up. I'm guessing there's far more to this than what is publicly known right now.

Wow! You almost read what was in my mind all this while but I was desisting from saying aloud. I too think there is far more to it than what Boeing has really spoken about. If it had been a simple matter of turning off a switch or two, which by the way many pilots say should be a natural reaction of any pilot in a similar situation even if they don't know about MCAS, by now after the Lion Air crash must have become so common that you will not expect it to repeat a second time. There is something more here than what meets the eye.

[pinin]

In multiple post, came across 'pilots did not switch off MCAS'. As a novice, my question is - why would you need to switch off some switches in case of an emergency? Why can the system not be designed for a manual override?

Take a scenario where an autonomous car suddenly starts to veer off the road. Can you imagine a situation where the manual says - Goto ICE --> Goto Steering control --> Switch off Auto mode --> Take control of your vehicle. All this while the driver is busy keeping the car on road!

Needless to say that I do not know anything about Aircraft controls. Experts, please let me know.

[HyperRetard]

Originally Posted by pinin

In multiple post, came across 'pilots did not switch off MCAS'. As a novice, my question is - why would you need to switch off some switches in case of an emergency? Why can the system not be designed for a manual override?

Take a scenario where an autonomous car suddenly starts to veer off the road. Can you imagine a situation where the manual says - Goto ICE --> Goto Steering control --> Switch off Auto mode --> Take control of your vehicle. All this while the driver is busy keeping the car on road!

Needless to say that I do not know anything about Aircraft controls. Experts, please let me know.

Because MCAS works on sensors, and the sensors malfunctioned.
The sensors believed that the plane was stalling, so they automatically pushed the nose down and increased the engine speed, to prevent the engines from stalling.

To put into perspective, lets say, a self driving Tesla thinks that you are going too slow, so it increases speeds, while you are going at just the right speed, and you end up crashing.

If you can diagnose the problem fast enough, you just take control of it, but if you dont, it will just keep accelerating.

The problem here is not the MCAS, but the sensor that feeds the wrong data.

[rpads1]

Aircraft takeoff and landing in my opinion are very critical . I'm surprised that the expectation is to turn off a software in case things go wrong.
I'm making a simple case of a very complicated problem at hand. The approach should instead be, turn on the MCAS software once pilot explicitly gives over control to software. A manual control should always be given preference to the automated one.
Again sensors are bound to fail and there should be an alternate check mechanism to warn or disable any software dependent on the same.

[Jeroen]

Originally Posted by rpads1

I'm surprised that the expectation is to turn off a software in case things go wrong.

In many cases that would you do. If you drive on autocruise and it is going to fast for you liking, getting to close in front of the guy in front of you, you switch it off.

It the autopilot, for whatever reason is not doing what you think it should be doing, you will switch it off and take manual control

Originally Posted by rpads1

A manual control should always be given preference to the automated one.
.

Why, what evidence do you have for that?

You need to realise that automation / computers are infinitely better at doing a lot of stuff than humans. During the take off commercial jets use something called auto throttle. Why, because it is able to control the exact thrust you need much more precisely than a human. It is safer, more reliable and more efficient.

Automation has played a huge part in the current incredible aviation safety records. But sometimes the automation goes wrong, moreover, pilots seem to be getting confused over what automation is doing.

Something as flying a modern airliner at mach 0.8 at 35.000 feet is not something any pilot would undertake manually for any length of time. It is just to exhausting.

Current so called separation criteria (how close aircraft can get horizontally and vertically) are so tight, that it is mandatory to have advanced automated flight system. Humans can not be relied on flying this precise for any length of time.

Be also aware on what you define as “manual”. In many planes no matter what, so called manual flying still involves a lot of computers controlling stuff. E.g. how do you think fly by wire works?

Etc. etc.

Jeroen

[rpads1]

Originally Posted by Jeroen

In many cases that would you do. If you drive on autocruise and it is going to fast for you liking, getting to close in front of the guy in front of you, you switch it off.

It the autopilot, for whatever reason is not doing what you think it should be doing, you will switch it off and take manual control

My thought was more like, the car does not go on autocruise on its own, default is disabled and one has to explictly enable.

Originally Posted by Jeroen

Why, what evidence do you have for that?

You need to realise that automation / computers are infinitely better at doing a lot of stuff than humans. During the take off commercial jets use something called auto throttle. Why, because it is able to control the exact thrust you need much more precisely than a human. It is safer, more reliable and more efficient.

Automation has played a huge part in the current incredible aviation safety records. But sometimes the automation goes wrong, moreover, pilots seem to be getting confused over what automation is doing.

Something as flying a modern airliner at mach 0.8 at 35.000 feet is not something any pilot would undertake manually for any length of time. It is just to exhausting.

Current so called separation criteria (how close aircraft can get horizontally and vertically) are so tight, that it is mandatory to have advanced automated flight system. Humans can not be relied on flying this precise for any length of time.

Be also aware on what you define as “manual”. In many planes no matter what, so called manual flying still involves a lot of computers controlling stuff. E.g. how do you think fly by wire works?

Etc. etc.

Jeroen

If use is over-riding the auto mode should it not let him over-ride and have preference over the controls. I do understand manual still means a lot of electronics is involved but at-least the input to the electronics is human and not an automated input driven by a sensor that could malfunction. Also whatever the level of automation involved, I think unless you know the exact logic it follows it would be very difficult to quickly grasp (in a panic situation) what is going on and how to circumvent it .
Do note, all my observations are theoretical and I have absolutely no exposure to the aeronautical electronics technology.

[Jeroen]

Originally Posted by rpads1

My thought was more like, the car does not go on autocruise on its own, default is disabled and one has to explictly enable.

Sure, but bear in mind these days cars have a lot of automated system that are always activated. Some you can switch off, some you can not or only partially.

ABS as a very simple one, and under certain (ice) condition it might well be preferred to have it switched off. You also never know if it will work when needed. In most cases it does of course, but occasionally these systems will fail and you might never notice until you brake on a slippery road.

Think about all these electronic limited slip differentials, electronic stability systems, all are always switched on. The car and handling were designed with everything switched on permanently. Some you might be able to disconnect, many of these system you can not or only partly disable.

Somebody, the design team, has decided for you what the performance of a car is going to be and how that will be achieved. And that includes some permanent electronic trickery these days.

Plenty of functions on an modern aircraft are automated. They hum away doing their thing quietly with nobody ever knowing. Some of it you switch on deliberately (e.g. an auto throttle, auto pilot, navigations systems). Some are on by default, e.g. by design.

In case of the 737 MAX the discussion is about whether the pilots should have known there was this MCAS device that was permanently switched on. (but only does it’s trick under very specific conditions). And there was a routine, non normal checklist that would disable it without the pilots knowing it. Because if this thing runs amok, it manifests itself as a runaway auto-trim (which it is). Disabling the auto trim correctly disables the auto trim system and the MCAS.

Whether Boeing should have told/instructed/trained pilots on the actual presence of MCAS is something currently hotly debated in aviation circles. We will just have to wait till the regulator (FAA) passes judgement.

Originally Posted by rpads1

If use is over-riding the auto mode should it not let him over-ride and have preference over the controls. I do understand manual still means a lot of electronics is involved but at-least the input to the electronics is human and not an automated input driven by a sensor that could malfunction.

You have heard of glass cockpits, no doubt. All main instruments you use to fly and all main indicators and annunciators are all on digital displays. Guess what all that information on what the plane is doing (altitude, vertical speed, airspeed, ground speed, drift, slip, attitude etc) are all derived from countless sensors and computers. So even flying manual these days, all your instrumentation is driven by banks of computers, systems, sensor etc.

If the input is wrong (in this case an incorrect AoA) it does not matter whether the autopilot is flying or the pilot as such. Because initially they will start to respond (to the faulty signal) in an identical way.

Originally Posted by rpads1

Also whatever the level of automation involved, I think unless you know the exact logic it follows it would be very difficult to quickly grasp (in a panic situation) what is going on and how to circumvent it .
Do note, all my observations are theoretical and I have absolutely no exposure to the aeronautical electronics technology.

This is a very relevant point when it comes to automation in general. Aviation, process automation, but for instance also driverless cars.

Slightly off topic, rarely spoken of, but the biggest challenge in driverless cars is actually this automation-human interface in case things go wrong. At the best of times it takes an average person some time to respond to an emergency. If you were head down reading your iPad and suddenly the computer yells at you to look up and take control it will take considerable time.You have no situational awareness. You are startled, you need to look up, look around, figure what is wrong, decide what action to take and execute.

People are very poor at this sort of thing. Say you are driving at 72 km/h, which is 20 m/s, this could easily take 2-4 seconds, that means the car has already driven 40 to 80 meters before any action starts happening.

Pilots, well, at least commercial pilots are selected to be much better at this sort of stuff than the average person. Also, they get trained in all sorts of scenario. The have all kinds of routines and procedures drilled into them, memory items, checklist etc. As mentioned before take off and landing are the most crucial phases of a flight (most accidents happen during these two flight regimes). But it is a relatively short space of time, they will be very focussed and concentrated. So if anything goes wrong whether they fly manually or have engaged the autopilot the response time is minimal and they will have / should have excellent situational awareness. For instance, on a Boeing doing an autoland the pilot flying will have one hand on the yoke and one hand on the throttles. The autopilot (lots of computers) has actual control of the plane and it moves the yoke and throttle accordingly. The pilot’s hand follow though to what the computer is telling the controls to do. It enhances his situational awareness. If he needs or wants to take manual control, he pushes a little button on the yoke under his thumb and he disconnects the autopilot and takes manual control. Very little lag/response time.

(not sure how this works on an Airbus, the side stick doesn’t move, not sure about the the throttles.)

Once you are in cruise things tend to relax a little more. Also, if anything does go wrong, usually you have a little more time to figure out what it is.

Say for arguments sake that in the case of the Lion air crash, the crew really messed up. If they had just followed the procedure and handled as if it was a runaway auto trim, that would have been the end of it.

If you are in charge, what would your recommendations be narrowed down to three simple options:

- Train/instruct pilots better and more?
- Automate less, because obviously one crew could not deal with it
- Automate more, in order to exclude human error in this manner

It is probably fair to say that all three will improve the chance of this never happening again. Neither of them will be able to provide a 100% safeguard. Most likely the more automation option gives a better improvement over the first two.

These things are hugely complex, but less automation is not necessarily the best solution.

Jeroen

[ph03n!x]

Not sure if anyone here follows AVHerald - the posts are pretty technical, and the comments by readers - mostly pilots - are interesting.

Link for the ET-302 crash

[Jeroen]

Originally Posted by ph03n!x

Not sure if anyone here follows AVHerald - the posts are pretty technical, and the comments by readers - mostly pilots - are interesting.

The AVHerald is a very good source. Very good write ups. I doubt very much that most of the members are pilots.

Even on pprune, ProfessionalPilotsRumorsNetwork, a lot of the members these days are non-pilots, or pretend (armchair) pilots.

https://www.pprune.org/rumours-news/...-texas-30.html

On both anybody can join, you do not need to be a pilot. PPRuNe does have private sections where there is a more elaborate way of verifying who you are.

I am a member of both, but again, anybody can join and comment.

Of course, you do not need to a pilot to be able to contribute on an aviation/pilot forum. Lots of very knowledgeable people about. Some of them working in aviation, some not, some just having a keen interest. There are people who, purely out of interest, are able to build up a phenomenal understanding of a certain topics. We have some contributors in various threads on our forum too, on various topics, where that is very apparent.

But you need to separate the wheat from the chaff as they say. Just somebody uses a lot of technical lingo does not mean he/she is a pilot or real knowledgeable. Internet is a source for anything. Anybody can google, down load an aircraft operating manual, pick up some technical lingo.

You see it on any forum, also our forum. Posts that use elaborate and very technical language. Lots of own opinion, no real insights, but just beefed up with technical jargon picked up somewhere on the net. Basically show-offs. Not really contributing. Just a water fall of verbal garbage.

For some reason there are a lot of people that like to “pretend” on internet forums. Aviation forums, or aviation threads on a car forum like this, will always attract people who pretend to be a pilot.

Even so, both AVHerald and PRuNE are probably some of the best, if not the best aviation forums around. Apart wannabe/pretend pilots there are quite a number of genuine aviation experts (professional or by interest) about.

I mean, the guy Simon, who started AVH is not a pilot either. Even so, nobody would doubt his aviation knowledge. He is considered quite the authority and has access to many sources. News media will call him and you might find he is often used in main stream media as a source/ He will write to aviation regulators to seek clarification on what happened. And he usually gets response. So leaving the members/comment section aside the AVH is a very good source in how it reports on aviation incidents and accidents. Simon keeps adding to the story as more information comes to light. So it is a pretty comprehensive database on aviation incidents and accidents.

PRuNe is more about the discussion around a particular topic. So you need to built your own picture/image on the topic based on the comments and how you rate those comments.


Jeroen