This might not be the best place for this topic, but it has technical aspects to it.

I've noticed that a fair number of people on these forums seem to have a perception that technology/automation is the solver of all problems.

I've noticed this mainly in the automation of cockpits, but I've also seen it as a solution for just about any problem. Something on a checklist gets missed-add a warning system or have electronics make it an automatic function. No suggestion of adding in some training & put in a warning system.

My thought is that technology should be used to reduce workload freeing up the pilots to be more observant to catch the 1 in a million situation that the computers might not notice or be able to fix; you can't program for every single thing that could happen.

The aircrew & the computers work in conjunction to improve safety.

So I'm wondering where this perception of the computer/electronics will never fail vs the human brain will quite often fail. Where in fact, both are points of failure.

Is it lack of knowledge that causes this blind faith or other factors?

Not the greatest example, but there are now systems that can detect when a wheelchair air cushion is losing air, but what if the sensor fails & the person hasn't been keeping an eye on the inflation or doing body shifts to relieve pressure. If the person isn't doing body shifts, the lack of movement will result in skin breakdown, so automation is just another layer of safety.

Shouldn't there be & isn't there, a system of checks to improve safety? Mechanic's to fix, inspectors to check the mechanics work, pilots doing inspections & looking for hazards, ground control & linemen keeping an eye on things, radios for people to report if they see something.

It boggles the mind that people are expecting/trusting one single thing will catch everything & be able to deal with it.

I'm not sure where you're headed here.

There are plenty of checks and balances in aviation.

Trust but verify is very much alive in aviation. Certain mechanic's work is checked by an inspector. Pilots perform a walk around and pre-flight in an attempt to verify the aircraft is airworthy. Our non-certified folks are all trained to report anything that doesn't look right.

I've flown on enough flight decks to know that flight crews check each other and the automated systems.

The problem with freeing up pilots to catch the "one in a million" scenario is that you not only quickly run out of tasks that can be freed up, but you can't keep crews attentive long enough, over a long enough time, to catch that white whale they may or may not see during their career.

Another aspect is that the vast, vast, vast majority of aircraft crashes and mishaps in the last 70 years or so have been the result of pilot error, not equipment error. And even much of that equipment error is actually from pilots not understanding the consequences of what they told the computers to do. To train pilots to be more on the lookout for errant "computer things" will usually be at the expense of training them to not f-up by doing errant "human things", and if you're 1000 times more likely to die from pilot error than computer error, you're being penny wise and pound foolish to focus on those extremely rare instances where the automation is actually to blame.

One of my first flight instructors, when he showed me how to use the rather rudimentary autopilot in a Cessna 172, quipped, "Remember, the autopilot will kill you quickly". I still live by those words. Automation is a tool, but it is not perfect, can fail, and needs to be used correctly and with care.

Our perception is not that computer/electronics, or shall we say, automation, will never fail. Automation fails, and we train for that. We are also trained not to take things on faith. As fr8mech says, "trust but verify". We don't have "blind faith" at all. Every piece of equipment is under scrutiny, and modern aircraft have very robust master warning systems.

You talk about "one single thing" as if there is a single point of failure. Automation, like all the other equipment, is multiply redundant. One flight computer failing, for example, will have no effect on safety or handling. It doesn't even trigger a warning. Just a caution. When flight computers fail on the 'bus, we can reset (reboot) them in flight.

Computers are good at boring and repetitive tasks like checking parameters. Humans are not. We are singularly poor at observing more or less unchanging parameters for a long period of time.

As LyleLanley points out, equipment failure is a vanishingly rare direct cause of accidents in recent decades.

Looking back the C-5, advanced as it was for the day, was for pilots pretty simple—it’s was very mechanical and systems were easy to understand. I draw out how most them working and their failure points and what to do. The flight engineers had to know every switch, relay and valve, the power source and how the redundancy worked, down to the circuit breakers.

Speed forward to the new Global 7500, most of the details are electronically controlled and the actions displayed on synoptic displays. How the various line of code, relays, power sources are involved is a mystery except in a very generic way. You couldn’t fix the lines of code and boxes anyway. The fuel is both a marvel and pretty hard to figure out which I’m learning looking at a safety report.

Keep in mind with Automation is that Garbage in and Garbage out. Once you input something in, it is important to verify that the computer will do the task that you have instructed it to do. For example setting the MCP altitude and verifying that the Altitude on the primary flight display is set prior to selecting the climb/descent modes on the MCP to program the A/P the way you it to reach your new altitude. Then after setting the A/P climb/descent modes on the MCP, you want to verify on the flight mode annunciator that the A/P is doing what you have told it do it. If it it isn’t, then intervention needed by selecting the appropriate modes to do what you want it to do, or de-automate and fly the plane.

Having a Automation is a great tool to reduces one’s workload in the flight deck, however it can increase one’s workload as well in some cases. This why it is important for us crew members understand the system when Automate and de-automate.

From my experience, reading on these forums. When folks talk a preach about automation and want to rely on it. It usually comes from folks who don’t have experience in the flight in the decks of any operating aircraft.

Where I'm going with this is that it seems more or more people want technology to be the solver of all problems with nothing to countercheck the technology.

Posters have already elaborated on how more automation is monitored and verified by pilots. No one believes everything is solved by blind-faith in technology, so I’m not sure what you’re getting at. Same as your “technology countercheck”. Do you mean pilots? If the technology is designed to help keep everyone alive is your “countercheck” designed to try to kill everyone?

Can you be more specific with what you’re getting at, because at this point it seems more like a generic rant than a cogent question.

I like the idea of this thread but I do not see the point you are trying to make. Yes, both the automation and the human are prone to error. However, who has the control of the automation, most of the time? We do. In the airbus planes, there is a control where the captain can invalidate the F/O's sidestick inputs and control the plane himself. The crew is the one who programs the FMS/FMC every time before the flight to direct the automation where to go. We engage/disengage/set the autopilot. That's just some examples. Also, the overwhelming majority of crashes are caused by pilot error, rather than mechanical failure. Even if there is no direct indication of pure error on the pilots' part, skeptics will always say "well this could have been done by the pilots to prevent this from happening." Air crashes are not caused by a singular event, rather a chain event, which at some point in the chain, human error took place in some form or another.

So in reality, most incident/accidents are caused by human error. And even that is astronomically low risk.

(someone can correct anything I said, I just felt like posting. I am only a high schooler and its just my 2 cents)

I suspect question is not about if automation belongs in cockpit (anyone wants to hand pilot transpacific flight all the way through?), but where two aspects should meet.
My impression is that every piece of equipment - not only aviation, but in general - is designed for existing level of user experience and understanding, but necessarily affects that baseline level for future users.
So any target is a moving target.

Sorry, I didn't do a very good job defining the question. Right now, in the cockpit & the ground, we have a system where everything is cross-checked with redundancy in the system. However it seem like more & more people think or see computers/automation as being the way to solve all problems, so a single system instead of multi-tiered like it is now, I'm wondering was has led people to believe that computers are infallible & the answer to all problems.

As the example I have above, say something is missed on a checklist or not entered right. Pretty well every non aviation persons response to that sort of thing is add a alarm or automate. The only ones that suggest more training as part of the fix are pilots & others directly involved. The perception I see is that that the computer will always solve the problem & the human is where the problem is. For some reason they don't understand that a computer can't be programmed for every single scenario imaginable.

Canuck600 wrote:

Sorry, I didn't do a very good job defining the question. Right now, in the cockpit & the ground, we have a system where everything is cross-checked with redundancy in the system. However it seem like more & more people think or see computers/automation as being the way to solve all problems, so a single system instead of multi-tiered like it is now, I'm wondering was has led people to believe that computers are infallible & the answer to all problems.

As the example I have above, say something is missed on a checklist or not entered right. Pretty well every non aviation persons response to that sort of thing is add a alarm or automate. The only ones that suggest more training as part of the fix are pilots & others directly involved. The perception I see is that that the computer will always solve the problem & the human is where the problem is. For some reason they don't understand that a computer can't be programmed for every single scenario imaginable.

Which "people" are you talking about that "think or see computers/automation as being the way to solve all problems". I have not seen any evidence of these people in the industry.

The solution of a procedure is not "working" is holistic. Why does this issue exist? Can it be solved with redundancy, automation, procedural change, training change, or a combination? Throwing more automation at the problem is not the default solution in the industry.

If you come up with concrete examples instead of vague concepts it would help the discussion.

Starlionblue wrote:

Canuck600 wrote:

Sorry, I didn't do a very good job defining the question. Right now, in the cockpit & the ground, we have a system where everything is cross-checked with redundancy in the system. However it seem like more & more people think or see computers/automation as being the way to solve all problems, so a single system instead of multi-tiered like it is now, I'm wondering was has led people to believe that computers are infallible & the answer to all problems.

As the example I have above, say something is missed on a checklist or not entered right. Pretty well every non aviation persons response to that sort of thing is add a alarm or automate. The only ones that suggest more training as part of the fix are pilots & others directly involved. The perception I see is that that the computer will always solve the problem & the human is where the problem is. For some reason they don't understand that a computer can't be programmed for every single scenario imaginable.

Which "people" are you talking about that "think or see computers/automation as being the way to solve all problems". I have not seen any evidence of these people in the industry.

The solution of a procedure is not "working" is holistic. Why does this issue exist? Can it be solved with redundancy, automation, procedural change, training change, or a combination? Throwing more automation at the problem is not the default solution in the industry.

If you come up with concrete examples instead of vague concepts it would help the discussion.

He didn't say people in this industry, he said people on this forum. This is the latest one I've seen so far.

viewtopic.php?f=3&t=1468293&start=100#p23107607

AirKevin wrote:

Starlionblue wrote:

Canuck600 wrote:

Sorry, I didn't do a very good job defining the question. Right now, in the cockpit & the ground, we have a system where everything is cross-checked with redundancy in the system. However it seem like more & more people think or see computers/automation as being the way to solve all problems, so a single system instead of multi-tiered like it is now, I'm wondering was has led people to believe that computers are infallible & the answer to all problems.

As the example I have above, say something is missed on a checklist or not entered right. Pretty well every non aviation persons response to that sort of thing is add a alarm or automate. The only ones that suggest more training as part of the fix are pilots & others directly involved. The perception I see is that that the computer will always solve the problem & the human is where the problem is. For some reason they don't understand that a computer can't be programmed for every single scenario imaginable.

Which "people" are you talking about that "think or see computers/automation as being the way to solve all problems". I have not seen any evidence of these people in the industry.

The solution of a procedure is not "working" is holistic. Why does this issue exist? Can it be solved with redundancy, automation, procedural change, training change, or a combination? Throwing more automation at the problem is not the default solution in the industry.

If you come up with concrete examples instead of vague concepts it would help the discussion.

He didn't say people in this industry, he said people on this forum. This is the latest one I've seen so far.

viewtopic.php?f=3&t=1468293&start=100#p23107607

Ooooh. Thanks for clarifying.

Well, I'm not really worried about it, then.

Starlionblue wrote:

Canuck600 wrote:

Sorry, I didn't do a very good job defining the question. Right now, in the cockpit & the ground, we have a system where everything is cross-checked with redundancy in the system. However it seem like more & more people think or see computers/automation as being the way to solve all problems, so a single system instead of multi-tiered like it is now, I'm wondering was has led people to believe that computers are infallible & the answer to all problems.

As the example I have above, say something is missed on a checklist or not entered right. Pretty well every non aviation persons response to that sort of thing is add a alarm or automate. The only ones that suggest more training as part of the fix are pilots & others directly involved. The perception I see is that that the computer will always solve the problem & the human is where the problem is. For some reason they don't understand that a computer can't be programmed for every single scenario imaginable.

Which "people" are you talking about that "think or see computers/automation as being the way to solve all problems". I have not seen any evidence of these people in the industry.

The solution of a procedure is not "working" is holistic. Why does this issue exist? Can it be solved with redundancy, automation, procedural change, training change, or a combination? Throwing more automation at the problem is not the default solution in the industry.

If you come up with concrete examples instead of vague concepts it would help the discussion.

A lot of folks here who believe Airbus side stick/FBW system and automation is the best thing since slice bread….

I'm talking about people on this forum that the key to aviation safety is getting rid of the pilot & let computers do everything.

Canuck600 wrote:

I'm talking about people on this forum that the key to aviation safety is getting rid of the pilot & let computers do everything.

I think that’s a bit of a generalization. Whenever the topic of 1 pilot or no pilot cockpit comes up, users are very against it.

It is a bit difficult to argue against though. Most of the “automation” incidents or crashes still boil down to pilot error. Airline pilots these days aren’t Chuck Yeager, getting a feel for their airplane. They need to be qualified to fly it under virtually any circumstances, but they also need to understand the computer (within reason, not saying these folks need to be IT pros)

I was trying to look up some examples of automation incidents, and outside the MAX8 issues, I was actually having difficulty. Most accidents that place blame on automation include the phrase “the crew failed to….” Which in my eyes takes automation out of fault. The Crew put in the wrong waypoint, the wrong speed, altitude, etc etc.

If you burn your dinner because you set the oven to 500 degrees and the timer to 6 hours, do you blame the oven?

FGITD wrote:

Canuck600 wrote:

I'm talking about people on this forum that the key to aviation safety is getting rid of the pilot & let computers do everything.

...I was trying to look up some examples of automation incidents, and outside the MAX8 issues, I was actually having difficulty. Most accidents that place blame on automation include the phrase “the crew failed to….” Which in my eyes takes automation out of fault. The Crew put in the wrong waypoint, the wrong speed, altitude, etc etc.

If you burn your dinner because you set the oven to 500 degrees and the timer to 6 hours, do you blame the oven?

Is FGITD referring to the recent Emirates 231 incident? If so, FGITD, please state as much.

FGITD wrote:

If you burn your dinner because you set the oven to 500 degrees and the timer to 6 hours, do you blame the oven?

It's obviously the oven's fault! Back in my day, we used to cook over open fires and you'd have to hold the meat over the fire with your bare hand. You knew exactly how long you'd had the meat in the flame, so we didn't need any fancy "timers". You youngin's today don't know what real cooking is. Probably shouldn't even be called cooks, but button pushers.

People don't need computers to kill themselves. They've been doing a splendid job of that for a long time! Airplanes crashed long before automation was ever introduced and they'll still happen because humans continually relearn lessons learned in the past. The first person to stall a plane and die was probably, what 1903? 1904? Still happens today, because people will always find a way to kill themselves. At the end of the day, if something can be f*cked up then someone, somewhere will f*ck it up. From the looks of the EK incident it looks like the crew would've figured out another way of being stupid.

I just appreciate the "I spy with my little eye" contortions needed to figure out what this topic even asked.

FGITD wrote:

Canuck600 wrote:

I'm talking about people on this forum that the key to aviation safety is getting rid of the pilot & let computers do everything.

I think that’s a bit of a generalization. Whenever the topic of 1 pilot or no pilot cockpit comes up, users are very against it.

It is a bit difficult to argue against though. Most of the “automation” incidents or crashes still boil down to pilot error. Airline pilots these days aren’t Chuck Yeager, getting a feel for their airplane. They need to be qualified to fly it under virtually any circumstances, but they also need to understand the computer (within reason, not saying these folks need to be IT pros)

I was trying to look up some examples of automation incidents, and outside the MAX8 issues, I was actually having difficulty. Most accidents that place blame on automation include the phrase “the crew failed to….” Which in my eyes takes automation out of fault. The Crew put in the wrong waypoint, the wrong speed, altitude, etc etc.

I can think of a couple incidents where the automation wouldn't help. I would agree that automation didn't cause the incidents, but just pointing out that automation isn't going to help in every situation.

Qantas 32: Uncontained engine failure causing enough damage to the point where the automation was useless.
Qantas 72: Faulty ADIRU caused an autopilot disconnect.
Cathay 780: Contaminated fuel resulted in loss of engine control. Automation wasn't going to help here.
Northwest 85: Rudder hardover meant the automation wasn't going to help here.
Turkish 1951: Faulty radio altimeter caused the plane to think it was about to land, resulting in the auto-throttle pulling the power back far earlier than it should have.
Aeroperu 603: Pitot static ports taped during maintenance and wasn't removed, resulting in faulty data. Automation wasn't going to help here.

AirKevin wrote:

I can think of a couple incidents where the automation wouldn't help. I would agree that automation didn't cause the incidents, but just pointing out that automation isn't going to help in every situation.

Qantas 32: Uncontained engine failure causing enough damage to the point where the automation was useless.
Qantas 72: Faulty ADIRU caused an autopilot disconnect.
Cathay 780: Contaminated fuel resulted in loss of engine control. Automation wasn't going to help here.
Northwest 85: Rudder hardover meant the automation wasn't going to help here.
Turkish 1951: Faulty radio altimeter caused the plane to think it was about to land, resulting in the auto-throttle pulling the power back far earlier than it should have.
Aeroperu 603: Pitot static ports taped during maintenance and wasn't removed, resulting in faulty data. Automation wasn't going to help here.

I think you and I are agreeing. The pilot still needs to know how to fly the airplane, that goes without saying.

LyleLanley wrote:

]

It's obviously the oven's fault! Back in my day, we used to cook over open fires and you'd have to hold the meat over the fire with your bare hand. You knew exactly how long you'd had the meat in the flame, so we didn't need any fancy "timers". You youngin's today don't know what real cooking is. Probably shouldn't even be called cooks, but button pushers.


I just appreciate the "I spy with my little eye" contortions needed to figure out what this topic even asked.

Advanced! We used to throw rocks at the meat until the impact friction cooked it



I wasn’t thinking of the EK incident exactly, but it’s a great example. The plane only did what it was told to do.

Or an extreme example, Germanwings in the alps. The plane knows it’s flying low and towards terrain that poses some incompatibilities with remaining in flight. But the human imputed directions to do this, so this is what it’ll do.

The automation is only as smart as the person behind it

LyleLanley wrote:

FGITD wrote:

If you burn your dinner because you set the oven to 500 degrees and the timer to 6 hours, do you blame the oven?

It's obviously the oven's fault! Back in my day, we used to cook over open fires and you'd have to hold the meat over the fire with your bare hand. You knew exactly how long you'd had the meat in the flame, so we didn't need any fancy "timers". You youngin's today don't know what real cooking is. Probably shouldn't even be called cooks, but button pushers.

People don't need computers to kill themselves. They've been doing a splendid job of that for a long time! Airplanes crashed long before automation was ever introduced and they'll still happen because humans continually relearn lessons learned in the past. The first person to stall a plane and die was probably, what 1903? 1904? Still happens today, because people will always find a way to kill themselves. At the end of the day, if something can be f*cked up then someone, somewhere will f*ck it up. From the looks of the EK incident it looks like the crew would've figured out another way of being stupid.

I just appreciate the "I spy with my little eye" contortions needed to figure out what this topic even asked.

As an ART, I used tell the “utes”; “I’m an old dog teaching new dogs old tricks”.

bigb wrote:

Starlionblue wrote:

Canuck600 wrote:

Sorry, I didn't do a very good job defining the question. Right now, in the cockpit & the ground, we have a system where everything is cross-checked with redundancy in the system. However it seem like more & more people think or see computers/automation as being the way to solve all problems, so a single system instead of multi-tiered like it is now, I'm wondering was has led people to believe that computers are infallible & the answer to all problems.

As the example I have above, say something is missed on a checklist or not entered right. Pretty well every non aviation persons response to that sort of thing is add a alarm or automate. The only ones that suggest more training as part of the fix are pilots & others directly involved. The perception I see is that that the computer will always solve the problem & the human is where the problem is. For some reason they don't understand that a computer can't be programmed for every single scenario imaginable.

Which "people" are you talking about that "think or see computers/automation as being the way to solve all problems". I have not seen any evidence of these people in the industry.

The solution of a procedure is not "working" is holistic. Why does this issue exist? Can it be solved with redundancy, automation, procedural change, training change, or a combination? Throwing more automation at the problem is not the default solution in the industry.

If you come up with concrete examples instead of vague concepts it would help the discussion.

A lot of folks here who believe Airbus side stick/FBW system and automation is the best thing since slice bread….

I am one of those people. I find the system delightfully elegant and robust.

However, I also am acutely aware that this wonderful system can get me into trouble if I mismanage it. Automation is a tool, and as such can be used correctly or incorrectly.

Starlionblue wrote:

bigb wrote:

Starlionblue wrote:

Which "people" are you talking about that "think or see computers/automation as being the way to solve all problems". I have not seen any evidence of these people in the industry.

The solution of a procedure is not "working" is holistic. Why does this issue exist? Can it be solved with redundancy, automation, procedural change, training change, or a combination? Throwing more automation at the problem is not the default solution in the industry.

If you come up with concrete examples instead of vague concepts it would help the discussion.

A lot of folks here who believe Airbus side stick/FBW system and automation is the best thing since slice bread….

I am one of those people. I find the system delightfully elegant and robust.

However, I also am acutely aware that this wonderful system can get me into trouble if I mismanage it. Automation is a tool, and as such can be used correctly or incorrectly.

Be honest, it’s the tray table…. Lol

bigb wrote:

Starlionblue wrote:

bigb wrote:

A lot of folks here who believe Airbus side stick/FBW system and automation is the best thing since slice bread….

I am one of those people. I find the system delightfully elegant and robust.

However, I also am acutely aware that this wonderful system can get me into trouble if I mismanage it. Automation is a tool, and as such can be used correctly or incorrectly.

Be honest, it’s the tray table…. Lol

Touché. Yes, yes it is.

And the footrests. Perfectly placed so you can bump the autopilot disconnect button with your knee if you're not careful, but in all other ways marvelous.

Got on the plane once. Checked the tech log. "F/O Sliding table and Retractable Foot rest unserviceable. Removed. Up to 60 days." What the f----? That was a long flight.

LyleLanley wrote:

The first person to stall a plane and die was probably, what 1903? 1904?

If we include gliders, the first person to stall, crash and die was probably Otto Lilienthal in 1896.

Nobody believes that computers are immune to failure. That's why they're usually in triple-redundancy, as well as many other systems. Humans, on the other hand, only have one brain. Automation can reduce the pilot's workload and can catch the most common mistakes.

bigb wrote:

Starlionblue wrote:

Canuck600 wrote:

Sorry, I didn't do a very good job defining the question. Right now, in the cockpit & the ground, we have a system where everything is cross-checked with redundancy in the system. However it seem like more & more people think or see computers/automation as being the way to solve all problems, so a single system instead of multi-tiered like it is now, I'm wondering was has led people to believe that computers are infallible & the answer to all problems.

As the example I have above, say something is missed on a checklist or not entered right. Pretty well every non aviation persons response to that sort of thing is add a alarm or automate. The only ones that suggest more training as part of the fix are pilots & others directly involved. The perception I see is that that the computer will always solve the problem & the human is where the problem is. For some reason they don't understand that a computer can't be programmed for every single scenario imaginable.

Which "people" are you talking about that "think or see computers/automation as being the way to solve all problems". I have not seen any evidence of these people in the industry.

The solution of a procedure is not "working" is holistic. Why does this issue exist? Can it be solved with redundancy, automation, procedural change, training change, or a combination? Throwing more automation at the problem is not the default solution in the industry.

If you come up with concrete examples instead of vague concepts it would help the discussion.

A lot of folks here who believe Airbus side stick/FBW system and automation is the best thing since slice bread….

Which has very little to do with the alleged increasing over reliance on automation.