IEEE Spectrum Article: "When Will We Have Unmanned Commercial Airliners?"

[ross_M]

Interesting article in the IEEE Spectrum magazine (Dec 2011 issue).

When Will We Have Unmanned Commercial Airliners? (http://spectrum.ieee.org/aerospace/aviation/when-will-we-have-unmanned-commercial-airliners/0)

Curious to see what Ppruners think.

[Denti]

I found a good quote by Richard Aboulafia, an aviation analyst from the Teal Group about that topic on a blog post about UAVs:

Then there’s the oft-mooted civil UAV market. There’s something there too – homeland security and border patrol, to name just two roles, but nothing that will subtract from manned aircraft markets. Think robots will replace pilots on commercial and business transports? Pilots are a tiny part of the operating costs equation. Any savings from replacing them would be more than erased by higher insurance premiums, to say nothing of unmanned system costs.

The full post can be found here (http://www.richardaboulafia.com/shownote.asp?id=344).

[Clandestino]

Thanks for the link :ok: The author of the article is so ignorant of the way things work out in the air that he managed to produce stupendously hilarious text, which might or not have been his intention.

[Checkboard]

I've always thought that we will never have unmanned airliners, however if we make the next step to sub-orbital hypersonic regular public transport - then that might also include full automation.

Just as in today, you get onto a roller-coaster, staff help you strap in and check you're safely secured - and then they get off and you go for the ride, trusting to the design.

It wasn't always that way - roller coasters used to have "brakemen (http://themeparkreview.com/forum/files/018_21.jpg)" riding on the coaster to control it - no one thinks twice that they aren't there now.

I could see a hypersonic flight, taking an hour or less, working in the same way.

[grounded27]

We may not see it in our lifetimes bun unmanned flight is the future. Automation has improved safety (with a learning curve) and will decrease insurance premiums. The largest problem today in the cockpit are pilots who do not hand fly as much as yesteryear due to automation and do not respond properly to adverse flight conditions when they disconnect the A/P. The radio operator, navigator, F/E. All flight crew members replaced by automation, I am sure the aviators of their time would have said no way in hell in their time. The F/O is next with UAV control from the ground capability in the event the PM (that will really be your job) is lost.. Got to ask yourself what FANS/ADS-B is really for???

[MathFox]

I am with Denti, the operating cost of two pilots is fairly small, compared to general operating costs. On the other hand, the cost of development and verification of an autonomous control system is very high. UAVs have an operator, who directs the vehicle during its mission.
I think Cruise Missiles can truly be called autonomic vehicles... but their control systems only needed to be validated against "has a decent chance of completing a suicide mission", which is not exactly the standard that an airline passenger expects.

I expect that there will be progress in the area of autonomous vehicles, both land-based and airborne; but for planes I expect to see them first in military and GA (patrol, crop dusting) applications; it will take a long time before we will see autonomous airliners.

[911slf]

Objections to overcome in ascending order of difficulty.


Technically feasible
Economically viable
Politically acceptable

[Golf-Sierra]

I am with Denti, the operating cost of two pilots is fairly small, compared to general operating costs. On the other hand, the cost of development and verification of an autonomous control system is very high.

But this was in fact addressed by the author of the article quoted in the first post.

On an A380 or B744 indeed the cost of the two crew may seem a very small proportion of the overall cost. But - consider the need to carry relief crew, have the crew sleepover at a remote destination, have a contingency for a crew member falling ill - it all adds up.

It was also mentioned that such flight automation could be applied to small aircraft rather then large ones. In the case of a 10 pax plane the cost of having two crew is quite significant. This could open up possibilities which today are simply economically not viable.

Think also of today's ATC. Despite an enormous amount of technology both in the planes and on the ground - the bottleneck of the system is the controller needing to communicate with the pilots via AM radio which dates back about a 100 years.

[keitaidenwa]

Unmanned drones make much sense in military gear. Cost and weight saving as you can eliminate life support systems. Better performance, as UAV can ignore any g-force limitations puny meatsacks enforce on manned craft. And finally, nobody cares if an UAV is shot down, while an enemy force killing or capturing a pilot is major political disaster (at least for western countries that depend on goodwill from media and public).

Civilian transports gain none of those advantages. You still need life support for slf. Civilian planes don't need to outmanoeuvre other planes. And if an unmanned plane full of slf crashes, the media is REALLY gonna have a field day. Unlike in military, in the civilian world pilots having their own life on the line is a safety feature. It is a good motivator of taking pre-flight checks seriously and avoiding risks when in the air.

So I don't believe in fully automated civilian transport any decade soon. Single pilot + remote control as backup now is different story. Especially for the currently uneconomical <30 passenger planes.

[Denti]

The article i quoted talked about total operating cost, which includes relief crew, stopovers and so on. It is still a very small part of the complete cost equation.

Small aircraft might be a possibility, however for those the cost of other parts become a lot more prohibitive which is why many major airlines are phasing out CRJ200 and other similar small aircraft in favor of larger ones (CRJ 900/EMB 190). The biggest sector around the 10 seat mark is the GA/corporate sector, it would be interesting to see if those clients like to fly without pilots.

Dunno about your part of the world, over here datalink becomes mandatory in february 2013 for ANS in core EC states mandatory for all of europe in february 2015. Every aircraft with a new CoA issued from january 1st 2011 on has to be able to use datalink services, there is a grace period without any flight restrictions until february 2013, after that flight restrictions will be imposed. A retrofit period for older aircraft until february 2015 is available, however they will still be subject to flight restrictions from february 2013 on. States could already impose flight restrictions since january 2011, none does though. Communication should be much easier with that, however if you cannot really speak english, reading it might be even more difficult.

It is interesting though that many cite the current operating military UAVs as an example for civil aviation UAVs. The loss rate is several magnitudes higher in that operation than is acceptable within the civil aviation sector. That is without taking losses due to military response into account. Of course in the military that is acceptable as they do not have to run a business among many other things.

Yes, we do have an increasing amount of automation in our planes. However that automation is mainly a tool for human beings and is not able to replace them. The step to replacement is huge, both in needed investment and increased complexity. The gain however is exceptionally small, just the replacement of a couple pilots.

[Piltdown Man]

1. Technically feasible
2. Economically viable
3. Politically acceptable

4. Legally acceptable
5. Then you have to sell it! Maybe this should be number One?

PM

[fireflybob]

When the first fully automated aircraft spears into the middle of a large city, who are they going to blame?

The software engineers, aircraft manufacturers.....etc?

Answers on a postcard please.

[KiloB]

Hopefully fully automated airliners will only come about when the 'Software Crew' can guarantee us a safe arrival in the Hudson when required!

I doubt it though. Currently aircraft are becoming more and more autonomous and Crews less and less 'hands on'. The question is, will this process stop before full automation is reached?

KB

[Shaggy Sheep Driver]

Currently aircraft are becoming more and more autonomous and Crews less and less 'hands on'. The question is, will this process stop before full automation is reached?

I think is a real issue. The likes of Sully are highly experienced in hand flying aeroplanes of many types. Later generations of pilots, as the aeroplanes do more, are less so. We are beginning to see (Colgan, AF) perhaps the beginnings of the problems that might lead to. We are at a 'halfway house' right now; highly automated aeroplanes allowing pilots restricted hands-on, so if they do have to take over (especially, as with AF, under very difficult circumstances) they might not cut it.

Perhaps the insurance companies might like the industry, eventually when the technology is up to it (AF demonstrates it isn't there yet), to go the whole hog?

[EEngr]

That UAV 'lost' over Iran must really have both sides of this argument in a conundrum. The military argues (the last I heard) that this was NOT a shoot-down, but simply a loss due to C&C or systems failure. Meanwhile, the proponents of commercial UAV use are hoping for some hostile action, as that won't screw up the reliability statistics.

After reading "The Black Swan" (nothing to do with Natalie Portman) by Taleb. I'm not certain I want a statistician deciding when an empty cockpit is safe.

[grounded27]

Yes, we do have an increasing amount of automation in our planes. However that automation is mainly a tool for human beings and is not able to replace them

Really now, how does automation not replace the human function? It has so far cut the FE out of a job (fact)!

As far as UAV's and military ops go the reliability has only gotten better and the attitude/environment of ground operated UAV's with people on board would be much different. I can see pilots working shifts on the ground handling several aircraft, only having to actually fly in an adverse condition.

[vapilot2004]

Computers are not infallible and any statement touting digital perfection is hubris. A human (or two) will always be needed to keep a watchful eye and take control when things go wobbly.

Currently we are tasked with managing the system and as we watch the system, the system watches us.

[aviatorhi]

While I'm sure that someday fully automated aircraft with only one pilot/operator on board will become a reality I don't think that will be in my lifetime. The reason? Well, every civilian transport category aircraft which is on the drawing board and/or is intended to be produced over the next several decades has at least two people up front. Given that these aircraft will then be operated for several more decades I can safely say that someone who starts off in aviation at age 20 today would not see a civilian UAV by the time they retire at 65. Remember, it's already been a while since a new build aircraft required an FE, yet there are still many of those aircraft and those guys around, I'd be willing to say you'll see FEs around for another 20 years, even though the position is effectively "extinct".

[Groaner]

Interesting thought, generally radical new solutions start at the bottom and migrate up. As mentioned, cruise missiles and then UAVs have complete automation, is it really out of the question for GA/non-RPT aircraft to eventually be fitted with remote control as backup to the (single) pilot, and operate for quite a few years, and increasingly migrate to remote control as the main mode operation mode?

It may be proved after a (considerable) while that the remote control is safer than manned operation.

And then have the RPT manufacturers and regulators asking the question "if X number of aircraft have flown Y million unmanned hours, why not cut back RPT flight decks to a single pilot and the remote system?", followed later by "there hasn't been that many incidents when the (backup) pilot was needed, and by the way, she also caused Z number of unwarranted incidents due to interfering with the remote control, so why not get rid of her as well?".

Of course, this would be in the very long term, but I can see a path that makes some possible sense. Technical, political, legal hurdles etc are just that - hurdles. Don't think that statistics rule out the migration - just about every significant part on a modern aircraft is designed with an implied or explicit statistical failure rate.

May not be popular here, but don't imagine lack of popularity automatically rules it out.

Have you heard the one about the computer, the pilot and the dog?

[MathFox]

Computers are not infallible and any statement touting digital perfection is hubris.
Humans are not infallible and any statement touting human perfection is hubris.

In my experience computers can perfectly repeat boring tasks... It has been shown that computers are better at counting cells in microscopic samples than human analysts. But don't ask a computer to do anything outside its area of programming... :} :eek: :sad:

Humans are more adaptive and capable of taking action in situations they have not met in training. Quite often they are capable of saving the day... sometimes they mess up a perfectly salvageable situation. (Being distracted enough to cruise past TOD. ;) )

The more engineers make flying routine, the less need for pilots in the cockpit. The UAV observation is: the plane can be smaller if there is no need for a cockpit. Which allows development of an "unmanned small plane" niche. Utility planes for areal photography and surveillance with payloads of 50 kg or less. Technology might grow to unmanned mail-runners, replacing the piloted Cessna's that do that job now. It will take 20-30 years to see those appear on the market though.

[vapilot2004]

Humans are not infallible and any statement touting human perfection is beyond hubris and into the realm of farce or folly.

Agreed. :}

[Flight Safety]

Mathfox and Vapilot2004 both make good points. Computers are good at the boring and repetitious (via programming), while humans are good at the adaptive and unpredictable (via insight and intuition).

The reason A320s don't have auto throttle for ground maneuvering is because use of throttles on the ground in too unpredictable to program. The reason we don't have autos that can drive themselves from one location to another, is because the journey from source to destination is too unpredictable to program, with far too many variables. Maintaining a speed, a heading or a climb rate while airborne, is child's play compared to an auto journey.

The human mind can deal successfully with the unpredictable and the unexpected, but the computer can only deal with the unpredictable and the unexpected, to the degree that the programmers predicted and expected the unpredictable and unexpected. which will never fully happen.

It's not about hardware, it's about software and its limited decision making. No form of AI ever developed now or in the foreseeable future, can accurately handle the unpredictable and unexpected the way a human being can (when the human is properly trained).

[grounded27]

The reason A320s don't have auto throttle for ground maneuvering is because use of throttles on the ground in too unpredictable to program. The reason we don't have autos that can drive themselves from one location to another, is because the journey from source to destination is too unpredictable to program, with far too many variables. Maintaining a speed, a heading or a climb rate while airborne, is child's play compared to an auto journey.

You bring up a good point. For years now airports and airlines have been trying to eliminate taxi all together. It has been not much more than a cost savings debate but would w/o a doubt provide an answer to the a320 problem you present. There are other solutions such as prox sensors embedded in the taxiways. I would want to enhance the T/O roll with something like this as you would all ways need a protected runway.

[FullWings]

I don't doubt that eventually we will get to the point that AI-operated aircraft are safer and more efficient than human-piloted ones. At that point in history, though, you will be able to say the same about virtually every other human endeavour. Anything we could do would be better done by machine intelligences.

In the cockpit automation is on the rise but the human part of it is still necessary for unforeseen situations and interacting with other humans. I don't see this changing a lot in the short to medium term, very much like fusion power and strong AI were always 20 years in the future, whether the predictions were made in the 70s, 80s, 90s, etc.

This is without adding in the cost and the problem of consumer acceptance: we are still only a few steps down the road :p of making a reliable self driving car Google Car (http://jalopnik.com/5828101/this-is-googles-first-self+driving-car-crash). Who wants to be first up in the Google Plane?

[Flight Safety]

Full wings, did you read update #2 of the Google car link you posted? It's interesting.

[awblain]

Just a couple of years after the wrinkles are fully ironed out in the auto-lawyer.

[barit1]

Happened already (http://travel.usatoday.com/flights/post/2011/03/air-france-women-crew/146694/1) :rolleyes:

[Sassy91]

^^^ LOL

Those who think that eliminating the pilots will save space/weight, think of all the equipment that they would have to put in so the aircraft can be remotely controlled.

[MathFox]

Those who think that eliminating the pilots will save space/weight, think of all the equipment that they would have to put in so the aircraft can be remotely controlled.
Think of all the equipment you can remove (all of the cockpit dials and controls; pilot chair, oxygen... crash axe :ouch: )

[EEngr]

In my experience computers can perfectly repeat boring tasks... It has been shown that computers are better at counting cells in microscopic samples than human analysts. But don't ask a computer to do anything outside its area of programming.And therein lies the problem. Particularly in applications such as avionics/flight controls, the software has to be developed and tested extensively to ensure its proper operation under all foreseeable circumstances. Translation: You are depending on a bunch of code monkeys and engineers (like me) to anticipate the problems that you (the pilots) will see in operation. And sadly, engineers love to fall back on statistics to eliminate cases that they believe to be highly improbable. Which, in some cases turn out not to be.

Worse yet, examples of engineering breakthroughs (like self driving cars) depend heavily on machine learning systems. These are effectively trainable computers that 'make up' their algorithms on the fly (a dramatic over-simplification of the technology) based on training cases which need to be selected and validated by those same engineers (often a degenerate case, in which oversights in the training models propagate into the trained agents' behaviors). Or use actual operational data which can lead to some big holes in the algorithms' state space decision paths given that nobody is going to let the computer 'play around' during a revenue flight just to fill in some parameters*.

Back in my day, software certification processes did not allow for applications that, over time, can generate new decision paths not present in the original product and therefore untested. Has the FAA changed its position on this?

*Call me when they build a flight controls computer that can log onto PPRuNe after hours and shoot the bull with real pilots to pick up handy advice from experienced practitioners.

[MathFox]

*Call me when they build a flight controls computer that can log onto PPRuNe after hours and shoot the bull with real pilots to pick up handy advice from experienced practitioners.
Logging onto PPRuNe is the easy part... a GPRS (GSM) modem will even allow in-flight bull-sh*t-ting on the forums. A properly designed solid state computer can withstand 30-100g, so most bad landings will be survivable for the AI (that will be able to brag on on PPRuNe until consumed by the post-crash file.) Picking up good advice... I will need to discuss that topic with a natural language processing expert. (I am fairly sure that the AI will be programmed to avoid PPRuNe :p there)

With drones already flying around (with incentive to improve performance) there will soon be a choice of "robotic flight controllers", all too poor for passenger transport. However, some controllers will be suitable for (small, Cessna sized) unmanned freight transport. How much domestic use will the FAA allow? That's the major question that will ultimately decide upon acceptance.