Automation vs Seat-of-the-pants-flying talking as devil's advocate - so no abuse plea
 

I posted this in the Asiana 214 thread but is was then closed. Was it a :suspect: Moderator who edited my post, changing *Land Rover* to Trabant? I left it as is because it made me laugh.

Automation vs Seat-of-the-pants-flying
talking as devil's advocate - so no abuse please.

Occasionally the pilot has to intervene because the automation is not doing what the pilot expects.

How many times does the automation intervene because the pilot is not doing what the automation philosophy demands? It is probably under reported.

From my perspective the automation has lead to increased safety in the last decade and the odd exceptions do not invalidate this perception.

Most car enthusiasts have had to acknowledge that they can't repair an engine problem on the side of the road any longer due to the sheer amount of electronics under the bonnet (hood, for the cousins) even on my diesel *Trabant*. But, you have to admit, modern cars are incredibly reliable, so it is a worthwhile trade.

That planes will fly without pilots within 20 years is indisputable. Driverless cars will come within ten years. And safety will continue improve.

changing *Land Rover* to Trabant

Just an aside .. the system did that automatically.

Don't worry too much about it .. the history is it was set up as a joke (a long time ago) between two of the the senior folks on the totem pole.

If I recall correctly, you just need to put some punctuation in, eg Land-Rover, to have the system miss the "offending phrase" and get around the replacement.

Same thing happens if one posts any of a number of naughty words of note and you can see evidence of this from time to time.

I am agree, the automation has improved the security, nevertheless, has converted the pilot like a simple manager of the plane, doing the training programmes more simple and short with the pretext that the automation will do that for you, so, you don't need to learn it.

Passenger or cargo planes without pilots will never exist.

I'm sure engineers at Boeing or Airbus could develop such a plane today. However, there will never be a company daft enough to take the risk of having such planes fly cargo or passengers.

There's just no way the lawyers would allow such a high liability risk to the company. The reason pilots are still in the cockpit is because Boeing and the airline need someone to pile the blame on legally when people die.

mross:

I'd disagree except you've made it indisputable. :D

Even drones needs a pilot...

Handflying raw data (visual) appraches does NOT use more fuel!
 

Posted in the Asiana crash topicWhere does this idea come from that handflying would use more fuel? That's bul%$hit! I'm sure the opposite is true! The pilots in my company save the company tons of fuel by flying visual approaches which are often 5 to 10 miles and 2 or 3 minutes shorter then many published full approaches or vectors to final. Almost every time I fly a visual approach raw data approach I land with MORE fuel in the tanks then my FMGC predicted I would have at touchdown! And even if you're hand flying a vectored approach you're not going to use more fuel then when using the autopilot!!!

Of course, I have to be careful. Very often when I hear a British carrier getting a visual approach, I can expect him to make MORE track miles then when getting vectored. :ugh:
Happened to me a few days ago: A Thomson B737 10 miles ahead of us requesting and getting a visual app when he was on downwind for Rwy 25 in Rhodos at 3000'. This guy maintained 3000' and slowed down to 160 kts on downwind, positioning himself to leave 3000' on final on the glideslope. We reported having him in sight and were cleared for a visual as nr 2. We descended to 1500', turned base when he passed us on final. Got our landing clearance at 800' on final when he left the runway. We exited the rwy one exit sooner then he did, taxied back to the apron and got on stand before he did, even if we never hurried or attempted to do so!

We saved fuel. He didn't! :cool:

Was it you in the cockpit perhaps, "Speed of Sound"? :} (just kidding)

The former I can understand arguing for, with the human element in the loop and the invaluability of human life, your statement may hold true for a long time. The question is, at what point will it become acceptable to stop referring to that human element as "pilot" in exchange for say, air vehicle operator, with suitably adjusted requirements for training and maintenance? Runway to runway automatic flight capability had been demonstrated a long time ago and many times over and all the initial speculation surrounding the Asiana prang makes a certain statement with respect to the respective "pilots" aptitude. The argument that "had the ILS been operational and the pilots' hands tied, the accident would have been avoided" will be heard in the upcoming lawsuits. From there it is again one step closer to de-labeling the human operator as "pilot" in favour of a yet more automation oriented role, since it was once again demonstrated that the respective human element has failed in its ultimate role to take over for failing automation.

Now, pilotless cargo planes? Ten years tops. A consortium centered around BAE Systems is vying for unmanned aircraft to be granted access to common civil (in their speak 'non-segregated') airspace and have recently executed a successful flight trial, see here. Once that access is granted, it is only a matter of finding the right balance between the value of your cargo in transit, the cost of transportation against the risk-hazard scenario at hand (potential liability and the probability of an accident). Insurance to cover such will be available from day one, for adequate premiums of course. Notwithstanding that there is a myriad of unsolved questions surrounding the prospective commercial operation of unmanned aircraft and some will quite possibly have to be resolved in courts.

So in order to sum it up, my half interested and somewhat educated yet by no means expert guess says that in foreseeable future we will see pilotless freighters - or more precisely, piloted from the ground - while passenger traffic will retain manned pilots. Their role will however continue to evolve towards yet more focus on automation and become somewhat more that of system operators rather than seat of your pants aviators.

As a flight engineer, I have always flown older generation aircraft. Whilst I am conversant with modern avionics/FMS, I sometimes wonder if they generate a lower arousal state on the flight deck? I would never denigrate the SA they can give, nor deny the flexibility and time savings they present, but when I read some of the tech threads with people contradicting which law the system is operating in, it does make me ponder over the validity of some of the training packages that people are given. Do any of you 'more mature' operators miss the higher workload of yesteryear ????

Talk to any sim instructor and they just shake their heads at 'so how did he get that flying job?'

Simply put, boxes break, in the real world and simulated in the simulator. And when they do, the pilots that can't fly with out the boxes kill people.

Guys that talk safety with regard to automation take the stance that pilots are idiots, that a box needs to fly the plane, and the pilot watches the box. I can't really disagree except somewhere around the 80s, we departed from hiring pilots that didn't need a box, to pilots that couldn't fly with out one. So in the end, if you hire idiots, yes, automation makes flying safer. If you don't hire idiots, then automation makes it easier.

Simply put, if you can't hand fly, navigate and land the aircraft, by yourself under IMC conditions, under partial panel, you aren't a pilot, you are something else, most likely someone who 'networked' 'got lucky' 'bought your way in' or whatever and shouldn't be getting paid to carry passengers. Argue the point if you want, either you can fly or you can't.

So it's more of a case if being able to step up to the mark when required, ie having the background knowledge and ability. Can newer pilots ever get that training or experience these days?

I was lucky and learnt to fly my first big jet B732 in days of visual flying around the islands and into big UK airports. Thus we learnt all aspects of operations, with & without ILS's, and learnt NPA's for real. The a/c was basic, it was just faster than the Navajo I'd spent years on, and bigger than HS125 I'd cut my jet teeth on. The foundations were solidly laid by some good captains and by scaring myself enough times when single crew. (there were lot's of "I learnt about flying..." moments).
I loved the B757/767 technology. Great a/c to hand fly and the info presented made it so much more accurate and easy to be precise. You could rip them around the circuit in some tiny places with full confidence where you were relative to the target. Never did forget or lose the basics. When it got murky the automatics were a joy and abbreviated arrivals still possible with confidence. I had great trouble trying to convince the newbies to hand fly more often and turn off the LNAV/VNAV. Look out of the window and use DME. That was 25 years ago. Now I teach B737NG. The newbies are straight off a Cherokee, or the lucky ones an EFIS trainer. The SOP's demand they fly LNAV/VNAV, so they do. They've read the books backwards. I try to encourage to think of the a/c as faster Seneca; the basics remain the same, things happen quicker. Don't forget the foundations you've already laid, but I do also teach the automatics, above and beyond.....

The SOP police and training gurus hate it. 1 year later you meet the budding pilots on the line and they are bored. No visuals encouraged, and LNAV/VNAV required if you do. The company policy is to dilute piloting skills, but they'll never admit it. It might not be the primary intention, but it is certainly the result.

If anyone knows how to post a link to the classic scene in Space Cowboys where Tommy Lee disconnects all the computers on the Space Shuttle for a landing. The young experience astronaut crew, on the jump seats, are horrified. Post it on here next to the 'Children of the magenta line' video. Next to that post a link to Asiana/SFO & THY/AMS. Then we'll have a view from both sides of the argument.

Actually try a few years ago, can't/ dont have time to find the reference at the moment but I remember reading of some enterprising drug smugglers using pilotless planes to get "cargo" over the border.

Now as to sanctioned cargo operations ten years sounds possible, will likely start in isolated areas then slowly spread as expereince is gained.

From a previous thread, this chap summed it up well;

"Automation was designed to reduce your work load, not fly your airplane because you can't."

More and more, I see the term "pilot" being replaced with "operator".

A pilot can pretty much get into any airplane and fly it with a little training.
An "operator" has to memorize the manuals so he knows what buttons to push when and hope the machine does what he wants it to do. And, if the machine doesn't do what he thinks it should be doing, accidents/incidents happen.

It seems more and more operators are being trained than pilots.

Your argument isn't helped by having posted a link to training in your company.

That would suggest that your pilots are trained to fly, as well as manage and as such will have above average flying skills. That means that the majority of pilots aren't as good as your lot, especially if savings are also gained by alternative route/approach planning rather than simply efficiently used thrust, optimum configuration etc. etc.

As flying skills are diminished in favour of training 'systems managers', the number of good flyers will also diminish and more fuel-efficient flying will inevitably tend towards the automatics. Then the bean counters will say 'look, the FD/MCP is consistently outperforming the pilots so let's put even more restrictions on hand-flying'.

At present, with more and more hand flying being discouraged, it is becoming harder to demonstrate statistically that a good pilot using their brains can indeed give you a fuel cost advantage over automation. Catch 22 as Yossarian said!

Of course, even if this could be demonstrated, they may then argue that a training regime such as that at your company is prohibitively expensive, but that argument should be countered by pointing out that good training is a 'one-off' cost vs a whole flying career of fuel savings which should increase exponentially as fuel costs rise. And if more hand-flying/planning is the way to go, there is an added advantage of keeping those skills sharp and current because you are using them on the line rather than in the sim.

As I said, this will probably require regulation* (under the auspices of greater safety)

*or enough particularly enlightened airlines.

I was going to say, that I guess in the good old days, but then I really don't know just how 'good' those days were, before all of the modernization that we have today.

Never the less, back when I was flying, I'll use the 727 as an example, the only time we every flew a coupled (autopilot) approach was in recurrency. For some reason the FAA felt that we had to show that we were able to let the autopilot shoot an approach to minimums, while we sat on our hands. I usually took advantage of letting the autopilot shoot the approach in the simulator, by taking a short nap. Admittedly this could cause a slight problem, if the guy in the right seat also took the same advantage to take a nap as well. :p

Back then most of us used 10,000 feet as our 'transition' altitude. That meant after takeoff, that the autopilot was turned on, the shoulder harnessed came off and the seat back was slightly reclined. On the descent it was just the opposite. Seat and seat back to normal, shoulder harness on and autopilot off.

Then I'd hand fly all the way to landing, no matter the weather. I don't think I can remember anybody ever using the autopilot on a visual approach. Back then we landed at quite a few airports whose runways did not always have any type of glide slope or a visual glide slope. Just used the old 3-1 method for judging a glide path. Worked very well.

It looks like for a lot of people, that has changed in today's world. Is it better? To be honest I don't know, as it seems that there are still way too many landing accidents, even with all this automation.

However, an accident such as the one at SFO, completely baffles me that with four pilots in the cockpit, two very experienced, that an accident like this could happen, modernization or no modernization.

@isanengOne study which has looked into this is Masalonis, A.J., Duley, J.A., & Parasuraman, R. (1999). Effects of manual and autopilot control on mental workload and vigilance during simulated general aviataion flight, Transportation Human Factors, 1(2), 187-200. (I can access a full-text copy, just PM me if interested.)

Another article that may be of interest to several of you is available online. It also gives some examples of how pilots avoid being reduced to mere button-pushers without reverting to (probably not SOP-compliant) pure raw data handflying: Trust but Verify | Flight Safety Foundation


@Teldorserious Actually I don't think it's that simple. I fully acknowledge the notion of a pilot having to be sufficiently proficient in some tasks even if they are not part of the daily business in highly automated flying (and not of particular interest to bean-counters who are but the executive power of customers who want cheap fares and above all investors who want profit). On the other hand, I posit that even a proficient pilot can be under par or find him/herself in a situation where the aid provided by automation is more than welcome. But of course that's a different story than being completely automation-dependent.

reply to John_Smith
 

I know the loss rate for military drones is very high. But these are not pilotless planes, they are remotely-piloted planes. The losses might be related to loss of comms. Does anyone know the reason (for the losses?) Don't the Russians have autonomous cargo spacecraft supplying the MIR space station? The MARS Rover mission was based on autonomous vehicles because of the too-long delays in radio signals for remote piloting. But today I am still very pleased to see two skilled pilots on board!

At present, with more and more hand flying being discouraged, it is becoming harder to demonstrate statistically that a good pilot using their brains can indeed give you a fuel cost advantage over automation. Catch 22 as Yossarian said!

Believe it or not there are airlines that are real LNVA/VNAV sticklers for conformity, but they still have monthly fuel burn league tables. How does that work? You plug in at 1000' and plug out at 1000'. You have no influence, but just fly the route with the computer. Ripping visuals are discouraged so the fuel burn is not under your control. What is the point of league tables. Beats me.

Maybe someone is shooting at them?

The other reason beyond comm link issues is likely that they are used in ways that would not be considered if a pilot was at risk.

In case you missed it, just search for 'drone license' on cnn.
The guy proposing the ordinace admits it is illegal and also highly unlikely that anyone would be able to bag a drone using the allowed 12 gauge shotgun.

For most "drones", that is correct. A better example is Global Hawk which is capable of taxiing out, taking off, flying its entire mission and landing -- all autonomously.

Commercial airliners will never go to pilotless operation in a single jump. It does seem unthinkable -- today.

But in 1958 if you told the proud five-man cockpit crew (pilot, co-pilot, flight engineer, navigator, radio operator) on a Lockheed Constellation that one day two-engine planes with only two cockpit crew would fly 8,000-mile non-stop trans-Pacific routes carrying nearly 400 passengers, they'd think you were crazy. If you told them each of the two engines would have 115,000 pounds thrust and the plane could essentially fly the entire route and land by itself, including automatic flare, landing, braking and rollout guidance in zero-zero conditions, they'd say you were certifiable.

Just as we didn't get from four-engine propliners with a five cockpit crew to a two-engine 777-300ER with a two cockpit crew in a single jump, we won't get to "pilotless" airliners in a single jump.

It also won't be done on today's technology but in evolutionary refinements of that.

The next step might be replacing the co-pilot with an on-call flight attendant trained for contingencies. That may seem humorous today -- if you're thinking only of today's technology and procedures. Any such step would be carefully studied, then tried on a very limited basis in strictly controlled conditions while gathering data. This is similar to how ETOPS was rolled out. You didn't go from four-engine trans-Pacific routes to two engines in a single jump.

Over an extended period, much attention will be given to contingencies and redundancy. They might incorporate satellite-linked remote piloting as a contingency. Whatever the path, it will be gradual, with each incremental step studied and tested.

"TRUST BUT VERIFY" FSF as quoted by Armchairflyer #18 above.

The statistics for pilots in the survey of 273 Pilots suggests that 82.5% were between the ages of 41 and 60.

The other 17.5% do not come in this age range, (and may be younger ?)

6.2% of all those surveyed stated that they used A/P as soon as possible after T/O.

12.1% kept the A/P on for as long as possible.

I wonder which age groups they came from.

mross,
In the early days of "glass", two new standard phrases were added to general flight deck use:
1) What's it doing to us now? and;
2) I've never seen that one before!
They are still in general day to day use, even after twenty five years plus.
Every time there is a new software load, there is an opportunity for new idiosyncrasies.
Remember, the software is only written by humans --- think of the flight deck crew as the final quality control gate for the software.
Pilots are going to be around for a very long times.

Obviously not a pilot. Of course cars, where electronics have improved engine performance and suspension has improved safety, can merely be pulled over and left to be fixed instead of tweaking the distributor of wiggling the plug leads.

Automation, in the context of handflying, doesn't intervene to save the flight. Far and away the biggest safety improvements have been with safety systems eg GPWS, TCAS and database Constant Descent approaches. These are not automation.

As for "not invalidating your perception", ask the victims of the Turkish 737 or Asiana 214 accidents about what they perceive automation and it's deleterious effect on the ability of pilots to actually fly.

in twenty years, not today
 

My first post was meant to be provocative :). But I did not say pilotless planes would be around soon, I said in less than twenty years. From the Wilbur brothers' flight to the Apollo moon landings was only 66 years and the pace of progress seems to be ever increasing.

@MurphyWasRright

Yes I had seen that story about Deer Trail. I will do some research but I thought the majority of piloted drones were lost in the landing phase and recall a loss rate of about 20%.

@Capn Bloggs

Are you saying that automation contributed to the Asiana 214 crash??? It sure looks like an automated landing would have prevented the crash. You make my argument for me! :D (with great respect sir, since I am only a PPL)

I wasn't comparing cars to planes directly, only making the point that the advanced technology has more advantages than disadvantages in terms of safety and reliability. ABS can diminish braking in soft snow but sooner or later the cars will 'learn' to deal with that.

Yeah but didn't he stack it the first time? ;)

True, but it is a tongue in cheek tilt at the button pushing philosophy and encourages practice to achieve more success.

I can't think of too many professions where the basic hands on skills of the chappie in 'supposed' control has diminished so much so fast. Comparisons would be interesting to note. How alone. or together, are we in this dilution of skill?

Watching a UAV land is one of the most comical sights aviation has to offer.

The USMC already operates such a thing. K-Max. Pilotless cargo hauling aircraft (helicopter)

I seem to recall that F-4 drones have been flying pilotless for some years at various weapons ranges for the USAF and the USN.

The tech to do this already exists. Has for some years.

Err, no, it is an obvious example of the deleterious effect of automation (using too much of it) on handflying skills.

As was R2D2:

mross:Of course automation would have prevented the accident. It does every day. But that wasn't an option since GS was AWOL.

Since you are relatively uncontaminated by auto-this/n/that, I contend that given a few manual landings in a 777 sim, you'd have a better-than-even chance of a completely successful manual landing at SFO 28L on a sunny day.

Well, mross, in a way "automation" did certainly contribute to the Asiana crash. It's obvious that the pilots where relying on the auto-thrust system to keep their approach speed for them, while failing to monitor the airspeed. Much like what happened to the Turkish 737 in Amsterdam. The Turkish WAS on the ils with the automatics switched on, but still that didn't save them. On the contrary! The Turkish crew still had the excuse that the A/thr did not perform as designed, because of a failure of the radio altimeter system. Most probably, the final report in the Asiana crash will say that the crew mismanaged the a/thr system. No hardware failure involved.

I'm sure that if the Asiana crew had tried to fly the approach with a/thr off like in a Cessna, they would not have crashed! (Even if they were very rusty in flying without A/thr)

So, yes, in a way one could say that automation (or the overreliance on it) was a big contribution to the crash. We'll talk again when the final report is out.

You might be interested in what my former Sabena colleague, now Airbus' Jacques Drappier has to say about the lack of basic piloting skills.

to Sabenaboy
 

It's a mistake to blame the automation! The auto throttle was effectively off (wrong mode). This was the pilots' fault. The automation did not fail. It only failed to do what was expected. (Of course the investigation may tell us a lot that we don't know yet) but we are on this website to idly speculate - well I am anyway :8 and also to learn from pilots who are well informed.

I do accept that too much automation will dilute pilots' flying skills but I don't agree that automation is to blame for this accident. It is obviously risky to fly in mixed mode when it is not well understood (again this is speculation).

Kefuddle

Laugh away at this carrier landing then:
First Arrested Landing of a Tailless Unmanned Aircraft Aboard an Aircraft Carrier | UAS VISION

UAS have come a long way. There are now several pilot optional UAS as well such as 'Little Bird'

UAV
 

Just an update for those of you who have not caught the news the X47b a drone with a 60+ foot wingspan has taken off and landed on a carrier. Manned military flight is about to end, soon. F4 Phantoms are being turned into drones to aid in training with weapons systems (we are automating them just to shoot them down). The reliability of UAV craft is getting much better, public perception of the ability to land on a carrier will certainly increase confidence as they are marketed to the civil sector..............Ian beat me to it.

The original post highlights a (quite common) lack of understanding when it comes to the differentiation between automation (i.e. autopilot/FMS) and modern flight control architecture with things like flight envelope protection.

Automation is for the most part either on or off, and will not "correct" a pilot when engaged, because the pilot has delegated control to the automation. Envelope protection and the like is a different matter entirely, and only enters into the equation extremely rarely.

Automation vs Seat-of-the-pants-flying

I'd like to point out to our OP that he might want to learn a few things about flying before making such a post, or such a title.

The thread is headed by a false dichotomy.

Seat-of-the-pants flying can get you killed whether you have automation or not. Death spiral in IMC flying due to failure to have an instrument scan working, or the correct instruments available for IMC flying, has nothing to do with automation, or its lack. RIP JFK's son.

Automation can kill you if you don't know how it works and make an incorrect choice, regardless of what flight regime you are in.

If the tension you wish to explore is between hand flying and automation, fine.

Seat of the pants is THE WRONG TERM to use here. Use of an integrated scan is what professional pilots apply when hand flying.

That is all.

Exactly. I kept my input regarding the OP to the technical side in the hope that a pilot such as yourself would point this out. Nice one. :ok:

Settle troops. The intent of the OP is pretty obvious. ;)

Actually, I find those who are good at seat-of-the-pants flying produce the result best down final when the machine is pitching about and the autothrottle is working hard (or it's out). Immediately reacting when you feel the @rse falling out of the aircraft instead waiting for the sight picture to change always produces a better outcome.