Here is the Start of the automated cockpit!

[Rockie]

It is being developed by Airbus and an airline owner. Since Airbus's equipment was involved in these accidents I would bet that is why they are working on this world wide data grid which is mostlikly step one of the plan.

When the 320 first came out Airbus grossly underestimated the importance of the human pilot in their equations and the lesson has been painful to say the least. They've learned their lesson in my experience, but if they are still bent on removing the human that is disturbing and I'm not sure I believe it.

[trey kule]

I think you might be looking to far ahead.
The next step, I think will be the reduction of three crew to 2 in the cockpit and one on the ground.
When the cruise pilot is at the controls , there is another pilot there.
This has so many advantages, and I cant see a reduction in safety, but a savings in costs.
A ground based cruise pilot can go from aircraft to aircraft for their shift
Then a reduction from two in the cockpit to one and one on the ground.
These scenarios will not even tweak the public as safety issues.

Will the ground link be replaced entirely by automation? I believe so, but it will be first in cargo only flights for years because of the public perception that a human is necessary.

The time will come.

There is always the argument that pilots can make decisions better than computers. I would like to see that backed upmwith statistics. If we elimated the accidents and incidents from complete pilot error and substituted accidents incidents from automation would the accident rate be better or worse.
Well, we dont have those stats yet I dont think. So how anyone can catagorically claim humans will be safer is beyond me.

As to difficulty in who would train a single transport pilot. Really? You cant see a simole solution to that.

[Meatservo]

Any meaningful statistics would be difficult to collect. Who gathers any information on the times where human pilots make some kind of judgement call that seems relatively arbitrary, common-sense or overly conservative that results in nothing bad happening? Nobody? How can you ever know when a decision you make changes the course of history? Nobody can know.

You could, I suppose, look at the times a human pilot did make an error, and through simulation, figure out what a computer would have done. A lot of the time a computer would have been just as screwed. You'd be missing out on a lot of data, though. You'd be missing out on those times where the computer just goes off into la-la land and you have to press the red button and level the wings yourself. Everybody's done this. It's happened to everybody. These statistics don't get compiled. And, it doesn't even matter.

No, I think it's silly to debate this based on safety. Those who stand to make a profit from making humans redundant will manipulate the data to support their safety claims after it's already been decided that this is going to happen. Any feeling you have that you can make a reasonable case for human pilots is irrelevant because it's never going to be cheaper to pay a crew once they figure out how to get people into the automatic ships. I'm not confident anything will be safer, but it will be cheaper, and that will please everybody. Everybody, that is, except people whose jobs have been taken over by computers and now they can't afford plane tickets. I'll be OK because I'll be pretty damn senior by then. Maybe I can score a job as a ground-copilot.

I'm just having trouble figuring out who the hell would WANT a job as a single pilot on one of these things. What a lonely, dreary, soul-sucking job that would be.

[Rockie]

Actually I think statistics can be compiled on how many times a human prevents a catastrophe on a flight. All you need is an observer to count every time a pilot makes a decision. You can count the times a human prevents the automation from killing everyone by adding up every time they touch the FMS, manipulate the guidance panel or disconnect the autopilot. Won't be long before they get tired of counting.

[timel]

You can count the times a human prevents the automation from killing everyone by adding up every time they touch the FMS, manipulate the guidance panel or disconnect the autopilot. Won't be long before they get tired of counting.

Absolutely, how many occurrences are reported and not reported into airline safety departments and will never be known by the general public and by most aviation professionals? A lot. Most of the time pilots did an outstanding job and nobody heard about it.

You can't have backup for every system and possible scenarios, you will never replace a live person in the cockpit that smells, hear, senses, there will always be a delay between information relayed between the ground and cockpit and at critical phases it will put passengers at risks.

You can't replace experience, training and human intelligence.

Still, I am optimistic, there seems to be people even at airbus who understand the real issues.
http://www.wsj.com/articles/airbus-offi ... 1428853600

When I was a kid we were all suppose to have a flying car in year 2000. We are not even able to go back to the moon. The journalist should stop drinking the kool aid in their offices, they should come and sit in cockpits for a few months and see what is really happening up there. Ho, that's right, they don't have the budgets for that anymore...

[CID]

When the 320 first came out Airbus grossly underestimated the importance of the human pilot in their equations and the lesson has been painful to say the least. They've learned their lesson in my experience, but if they are still bent on removing the human that is disturbing and I'm not sure I believe it.

Wow. Just "wow".

[garfield]

Well, if it's ever happen I just gonna have to move to Congo to really FLY an airliner. In an ideal world you would fly the ac from T/O to ldg, AP only for the cruise portion.

I'm a pilot, not a system manager sitting behind a desk!

[goldeneagle]

...and the Shuttle was perfectly capable of performing all of the above without the crew!

Actually, it wasn't. There was one component of the shuttle that could not be activated by folks on the ground, the landing gear. The rationale at the time, inadvertent deployment of the landing gear meant re-entry was no longer possible, so, folks in the decision train that were scheduled to fly it, veto'd that one item.

The shuttle was capable of a fully autonomous flight, with one detail excepted, it still required a live human in the cockpit to extend the landing gear.

[Rockie]

The shuttle was capable of a fully autonomous flight, with one detail excepted, it still required a live human in the cockpit to extend the landing gear.

Aren't you forgetting the room full of engineers providing 24/7 monitoring and control of systems plus the hundreds of other engineers and scientists behind the scenes providing additional support and the worldwide tracking and communications network? Also don't forget the years of mind-blowingly comprehensive planning that goes into every second of a space flight. There are literally thousands of people constantly involved in a space flight beginning years before launch until well after recovery.

It's impossible to see how anyone could consider the space shuttle in any way autonomous.

[Meatservo]

This is what I pointed out earlier. Everything that plane did automatically was calculated beforehand by very clever human beings, laboriously loaded into the computers and when the program was executed, the thing was like a model train on rails; it executed all the pre-considered manoeuvres and any variables were dealt with by the crew. Considering this to be a good example of autonomous flight would be like taking an old SLR camera, putting it on the tripod, composing the shot, calculating depth-of-field, figuring out the exposure, focussing, setting the self-timer and pressing the button- and then telling people the camera "automatically" took the picture because it waited ten seconds before the shutter fired. Worst possible example in support of automatic airliners. People don't seem to realize- steering a plane is the easiest and least important thing they need to figure out how to automate. Autopilots have been around since the 1920s. Just because a plane can "land itself" doesn't mean it doesn't need a crew- non-pilots always get this mixed up because they think the hard part is already automated.

[Mach1]

5. Radio Operator
4. Navigator
3. Flight Engineer
2. First Officer <- You're next buddy.
1. Captain.

The number of flight crew have been in decline for decades. Just think, how big of a bonus check can a guy cut himself by eliminating 1000, 2000 or more First Officers? Oh my! If I can get that much money by getting rid of the FO's, how much of a bonus check can I give myself by eliminating all those Captain salaries? I can quit or get fired, take my golden parachute and move on to the next company. Might even be able to retire.

If you think any other thought process is at play in any business, not just airlines, you are being naive.

[trey kule]

You missed cruise pilot...they can quite easily be replaced, and will be the next to go, and be replaced by a ground based pilot

[CID]

It's impossible to see how anyone could consider the space shuttle in any way autonomous.

The term "autonomous" is troublesome. Some people consider some unmanned spacecraft "autonomous" but using the strict definition of the word, there is really no such thing. But here's the thing. There are no autonomous pilots by definition either. So why measure the ability of an unmanned aircraft by such a term? An unmanned aircraft that would perform flights repeatedly with an acceptable margin of safety simply needs to automate (not make autonomous) several individual tasks that feed in to a main "program". The word "program" in this context doesn't mean software program. It means the overall purpose of the combined tasks.

So can Rockie work autonomously to drive his 777 from Vancouver to Hong Kong? Not a chance. There are many many people and a great deal of automation performing the many tasks that are necessary to complete such a "program".

Can a computer replace Rockie? Very likely. But only in the context of a "pilot" performing tasks required in flight. Un-required tasks still need a human touch. Tasks like eating, shitting, gossiping, forgetting to put the gear down, not noticing that the autopower has gone offline...etc. The "Rockie" computer need only provide the interface and perform the tasks assigned directly to Rockie. Of course other aircraft systems need to be modified to properly interface with the "Rockie" computer but FANS and NEXTGEN are slowly closing that gap of automation.

It's a romantic notion that an airplane or car or a can opener can't possible operate effectively without a human component in the "driver's seat" but that's all it is. Technically it is possible to replace a pilot with automation and maintain a reasonable margin of safety. Possibly even enhance it as so many other automation measures have.

[goldeneagle]

The term "autonomous" is troublesome.

Yes, I used the wrong word. Unmanned is more appropriate.

It was suggested earlier in the thread, this type of unmanned automation has not yet even reached trains, the simplest form (1 degree of freedom to automate, speed up, slow down, on tracks). I would suggest that person has never ridden the skytrain in Vancouver.

And for those who insist it cant happen, here is some 'light reading' to demonstrate.

http://www.cbc.ca/radio/day6/episode-23 ... -1.3119963

[timel]

There is a bunch of people in here that way too excited, you should change careers and apply to the world without mankind program, go make some extra pennies and save the world from the freaking humans.

This is just another old bean counter program whose purpose is to fill some greedy pockets with $ they won't have time to spend.

Safety? ... Yeah right BS.

[fish4life]

The biggest difference between planes and any other vehicle is when something goes wrong on a driverless truck or train you can program a simple apply brakes and throw error code until someone can come fix it. That is the biggest hurdle for aviation autonomy, when the systems fail it still needs to know what is going on and continue to work because you can't simply apply the brakes and throw an error code.

[Mach1]

I don't think anyone is saying this is going to happen tomorrow but, it is coming and a 20 year timeline is more than reasonable, possibly less. No sooner than 10 years would be my guess... not because of the technology but because airplanes are expensive and it takes time to certify them.

[CID]

The biggest difference between planes and any other vehicle is when something goes wrong on a driverless truck or train you can program a simple apply brakes and throw error code until someone can come fix it. That is the biggest hurdle for aviation autonomy, when the systems fail it still needs to know what is going on and continue to work because you can't simply apply the brakes and throw an error code.

Here's an even more interesting difference. The accident rate for trucks are horrendous. Train accident rates are much better and have only improved with automation. The main reason for truck and train accidents remains human error. Travel in airlines is extremely safe. In fact the safety margin has been improving over the decades even though the capabilities and performance of the aircraft have improved at an even faster rate. And like other modes of human controlled transportation, the primary causes of accidents in aircraft are lumped in the "human error" category.

Taking a trip across the Atlantic on a Pan-Am "Clipper" in 1925 was a technological triumph but it was slow (27 hours from New York to Lisbon with a stop in the Azores), uncomfortable and in today's terms of reference, quite dangerous. Those big flying boats needed a crew of 11 people (including 2 flight attendants) for 22 passengers. Besides the flight crew, they needed a navigator and a radio operator as well as a flight engineer.

Fast forward to present day where it's routine to fly high mach numbers half way around the world at the edge of the stratosphere with hundreds of people on board in full comfort complete with in-flight entertainment. And we're not limited to the low latitudes either. It's also routine to fly over the poles or other remote areas of the world. And....it can be done by two pilots. Except of course for those long routes where the operate needs to keep a couple of spares in the bunk. Although the performance has increased significantly, safety has improved, and workload has been reduced drastically.

All quite impossible without automation.

[Meatservo]

I wouldn't say "impossible". At most I would say "inconvenient" and "expensive". Unless you take a broader view of what "automation" means, of course. Automation needn't be some mysterious computer lurking in the avionics bay. It could be better-designed fuel system that employs ejectors, feed tanks and level sensors to maintain good trim without an engineer having to fiddle with valves. It could be an advancement in communications technology that allows a pilot to talk on the radio and switch frequencies with a knob rather than a radio operator having to key things out in morse and fiddle with antenna lengths. It could be a needle that points at a ground transmitter with a known location, rather than a navigator measuring the relative angle between stars and the horizon. And so on. People have been replaced by machines for as long as there have been machines.

The real question is whether a pilot's brain can be replaced with an acceptable alternative. And the other question is: to what extent does "less expensive" constitute "acceptable"?

..and, of course, there is no question of what "expensive" means. To a bunch of fat cats, it means guys like us having food on the table and a house. To us, it means the opposite. But we have given the decision-making power to people who can vote not to automate their jobs, and profit personally by automating ours. And that's where Timel and I agree about something.

[Boreas]

Hot MILF Tatiana works some knobs

[Rockie]

Many things a pilot does with their hands and feet can be done by automation within certain limits. Anybody who actually flies automated airplanes should be familiar with those limits. The reality is that aircraft are flown all the time in conditions that are inappropriate for, and even exceed the automation's capability. But that just deals with the physical control of the airplane which many people think is all there is to it. They're wrong.

What a pilot does with his brain is far more important and cannot be replaced by a computer. I hate to use cliches but this one fits "A superior pilot uses his superior judgement to avoid situations requiring his superior skill". Computers possess neither judgement or skill.

[Rockie]

The accident rate for trucks are horrendous. Train accident rates are much better and have only improved with automation. The main reason for truck and train accidents remains human error.

Trucks must be steered on a road filled with lots of other traffic of varying degrees of performance and speed controlled by other drivers, and are affected by weather conditions which require greater or lesser degrees of skill and caution. In other words a crowded, completely uncontrolled environment under sometimes wildly different conditions. Trains are big heavy vehicles riding on rails that are much less affected by weather conditions - although they still are to some extent. They also have deadman switches, and if the engineer's foot comes off the whole thing comes to an emergency stop where it sits on the rails with the brakes set until someone resets it. There are also thousands of times more trucks than there are trains. Those are the main reasons for the difference in accident rates, not human error and not automation.

It's not like trains are immune to accidents either and automation isn't helping there, just eliminating people. There is also a big difference between small people movers operating on short, tightly controlled and isolated tracks like the Skytrain in Vancouver or between terminals at YYZ, and a 120 car freight train for instance on the mainline between Calgary and Vancouver. The latter requires a high degree of knowledge and experience in train dynamics and an intimate familiarity with the track.

[xsbank]

http://www.cbc.ca/news/technology/world ... -1.3393437

[CID]

"A superior pilot uses his superior judgement to avoid situations requiring his superior skill".

And an autopilot. And and FMC. And an autothrottle. That "romantic" quote was coined in an era where it was quite impossible to automate the operation of an airplane. And of course it still applies but to a much smaller degree.

Computers possess neither judgement or skill.

That is just so wrong. A computer possesses the judgement and skill of the programmers and it's interfaces. And it can reproduce the results of those skills and judgement repeatedly ad nauseam without variability. And it never gets tired or mad. It doesn't fret about the mortgage or it's cheating motherboard or the meaning of life.

A computer that perfects execution of critical tasks at near 100% of the time, will maintain the level of safety even if it fails to save the day following a rare event. That's because it will save the day all those other times that flightcrew falls to complacency or fatigue or other human factors that lead to human error.

[Rockie]

And an autopilot. And and FMC. And an autothrottle.

These are just tools a pilot uses that still require skill and judgement in addition to manually flying the airplane. In many ways they make a pilot's job more complex and difficult - not easier.

A computer possesses the judgement and skill of the programmers and it's interfaces.

No they don't. Computers react to a set of pre-programmed parameters with another set of pre-programmed parameters - and that's it. Many times they do it inappropriately for the circumstances because there are rarely a single set of circumstances, they do it incorrectly, and they never do it with any degree of skill. Engineers sitting on the ground cannot possibly program in judgement and skill even if they had any themselves when it comes to flying aircraft, which they don't.