"Pilotless airliners safer" - London Times article

[MG23]

Originally Posted by Tourist

Fortunately, rarely triggered bugs are rarely triggered, so you have time to patch the software before the second smoking hole.

Unless it's something like a leap second bug, which happens all across the world at the same time, and affects every aircraft in the air at that time. There are many things that don't happen very often, but could affect multiple aircraft when they do.

Besides which, just debugging it may be extremely difficult if triggering it requires some complex set of circumstances that weren't properly logged on the aircraft that crashed.

And if it's hardware... oh dear. There's a few billion dollars gone to replace it all.

[RetiredF4]

None of this is relevant because nobody has designed any of them with the kind of safety and redundancy you would put in a passenger UAV.

A point i agree completely, and it is my argument, thank you that you start to follow. There is no proof what so ever that such system exists or that such system will be available in the timeframe you are talking about. Safety and redundancy in adequate way is necessary and all your mentioned systems do not have those yet. And because no such system exists your statement that those systems are safer than piloted systems is nothing more than an assumption. And assumption is the mother of all f*ck*p.

We might now start to define, what a safer than human piloted aircraft should be able to do, and it should not be less than it does today. And it should not cost more than today, and it should be as cost effective to maintain as today. And that all these necessary investments pay out, it should be safer than those today with pilots.

Would it not be wise to take the money such a new design would cost and improve the quality of those pilots who should fly those aircraft? You yourself stated, that you observed the knowledge and skill of pilots decreased over the years, which is not due to human inability but due to neglect. There had been a higher standard out there, why not regain it?

Start with a proper preselection of pilot candidates, train and evaluate them to high standards, pay and treat them as important persons in the company and give them chances to keep up with their knowledge and skills. And put the fate of the pilot corps in the hands of people who have been there, done that and known it.

I' ve said more than i intended to say, therefore it should be enough. Lets move on.

[Tourist]

Slash

All good points.

The reason I have mentioned military repeatedly is that in most of aviation the heavy lifting in terms or research, development and cost is done by the military.
The civil side follows the military once the money is spent.

This is not always the case, but generally, particularly when it comes to big changes.

The manufacturers on the military side are not just putting small amounts of money into it. As far as I can see, all the big players consider the next fighter programs to be unmanned.
This is I think where the big money will be/is being spent. The civil side will piggyback on it.

I think the public will demand it when years of USAF unmanned freighters have a much better safety record. Airlines will demand it when they see the profit.

You may be correct about takeoff distances, however there is a case to reduce the Takeoff distances required for unmanned due to the lack of thinking time required.
On a takeoff run, the distances are calculated assuming thinking time.
A human needs to try to gather the information and make a decision extremely quickly hence V1 was invented to simplify the decisions. Making decisions based upon multiple factors whilst working hard to takeoff is not a humans strong point.
A computer will have all the information it needs, or at least a lot more and can immediately take that decision. Computers are very good at concurrent calculations.

How fast are we at current position? How does that compare with expected?
Are we accelerating like I would expect? Better or worse than planned for?
What is the state of all my engines? Better or worse than min certified?
What is the wind speed? Better or worse than planned for?
Exactly how much runway do I have left? Better or worse than planned for?
What is the breaking action?

It can do these calculations 100 times per second.


It can also, without jeapardising safety, change its mind. It is not under pressure, it is just continually monitoring factors that affect its ability to fly. It can know if the best idea is to get airborne or take the overshoot.
We have had to invent things like V1 to try to mitigate the human habit of making poor decisions under pressure.
I see no reason to saddle a computer with V1
If the computer sees that more than one engine has taken damage in a twin above V1 or even nearly at rotate it might decide that aborting is the best option. It takes too long for a human to do that so sometimes aircraft get airborne that won't make it.

Your question is just a trap
You are asking me to say what will happen after the computer has failed because nobody could think of the scenario, yet you have just thought of the scenario.

[joe two]

forget about the coke , let's get some water for the tourist please.

[Tourist]

RF4

The safety and redundancy I was talking about was not related to the control system.

I was talking about more than one engine.
Fire suppressants
Backup actuators
Backup hydraulic systems.

These things will work just as well unmanned but are not worth the effort in a predator.
They would obviously be fitted to any unmanned airliner.

Lots of "shoulds" in your statement.

Let's take it back a few years.
The jet replaced the old prop airliner.
Do you think the first ones were cheaper?
Safer?
More reliable?

They happened despite not meeting any of these challenges.

Finally, yes we would all like a return to the golden age of well trained pilots selected for ability then kept in practise, but that really is your head in fairy land!

[KBPsen]

Originally Posted by Tourist

Finally, yes we would all like a return to the golden age of well trained pilots selected for ability then kept in practise, but that really is your head in fairy land!

No, that is remembering a past that never existed. Nostalgia is remembering the past how you thought it should have been and is something you seem to be suffering from.

[Mr Optimistic]

Well you can see why the military might be interested as the crew plus seats and support are proportionately heavy and g limits could be stretched. The military business case would still be for human ground control fairly obviously. Key issue probably link capacity and integrity. For civilian freight or heaven protect us pax, never - it is not a technology thing primarily, just plain obvious commercial pay back.

Imagine an autonomous system having to deal with unusual attitude recovery. A successful and reliably performing system may well be a technical impossibility when you stretch to ten to the power of whatever. No doubt a big button which called up an automatic recovery sequence would be quite welcome but imagine the permutations of starting conditions, data reliability, aerocharacteristics out of envelope etc.

Anyway, as pax, I derive a certain comfort knowing that those in control are mortal and have skin in the game!

[Tourist]

Mr Optimistic

Is this a wind-up?

It is taking up enough time defending my point against superior numbers without having to prove the same point again and again!

Go to post 225
They have such a system currently operational
It runs off an android mobile phone.
This is not the future, this is now and mundane
It is designed to take control off incapacitated pilots and has been tested in vertical dives towards the ground.


And no, the military intention is autonomous not ground controlled as is demonstrated in the X47B and Blackhawks links I've posted.

[Tourist]

KBPsen

Not really sure where you are coming from with that.
This is not about me bemoaning anything, it is others who think the past should be the future.

[AirRabbit]

As Long As We're Thinking "Futuristically"...

Notwithstanding all of the previously posted opinions, facts, feelings, predictions, and arguments (including my own) … I think it just may be equally likely to see an alternative form of “transportation” to the kind of “automatic piloted,” 500+ passenger airplane equipped with one or more sophisticated computer systems, similar to or more advanced than those described in the 1966-era “Star Trek” television series, that so many here seem to believe is inevitable. This “alternative” system would mirror a transportation mode also first envisioned in that same iconic television/movie series, that would be more convenient and at least equally safe … and would be based on the “transporter” system that allowed transfer of animate and inanimate objects between orbiting space ships and virtually any location on the planet below … or to other space ships. Transportation time would be impervious to any climatic or physical barriers and transfer time would be almost immediate!

[neville_nobody]

12 Pages and noone has even touched on how you could ever certify a pilotless pax aircraft. The efforts required to prove the redundancy plus the test flying would almost be not worth it. Then once you've done that someone has to insure it. All that's assuming you had the technology available to do it in the first instance. The cost of the redundancy would also be an issue to consider.

Some of these tech guys on here have a website mentality to these things in that you build a half arsed product get it out to market ASAP then iron out the bugs later on. Can't do that in aviation i'm afraid.......Mind you Boeing had a good crack it with the 787!

[-438]

In reality, aircraft will become more reliable and more capable.
The first step would be to go from 2 pilots to 1.
This will take at least 20 years. As far as I know, neither A or B have any plans for certifying a single pilot airliner.
Pilotless would be at least another 20 years behind that.
I will be retired by then.

[Faire d'income]

Quote:

Tourist - how would your totally computerised aircraft have handled the AF447 scenario?

Ah, perhaps a bad example!!

If the autonomous programming was not allowed to hand back control to a flight crew, then it would have been programmed to do another level of fallback. Which would almost certainly be something like "power and pitch".

It would have had to decide which 2 of the following to ignore: Overspeed, Stall, and Unreliable Airspeed.

You appear to have suggested that it should ignore the Overspeed & Stall warnings by going to pitch and power (assuming it is programmed). If the aircraft was actually stalled, as in AF447, applying power immediately would probably exacerbate the problem and indeed airbus recommend lowering the nose first, if you are stalled. The aircraft has no way of knowing which was which was correct, even if it referenced the GPS it would still have to decide based on majority (e.g. 3 GPS say we are slow, 3 ADIRUs say we are overspeeding and slow at the same time). What does the logic do now?

[harrryw]

X-47B Fails Landing Attempt — Again | Defense News | defensenews.com
I guess two out of three not too bad.
Of course if the diversion airport programmed was out i guess it would only be one out of three......and they would have had to say the test got aborted while they build another. Too bad if these were passengers.

[goldfish85]

[Autopilot takeoff is trivial compared to autoland. I first took part in autolands with pax on board 50 years ago, but 3 aircraft generations later no-one has built a commercial aircraft with an auto takeoff system - there's no business benefit to doing it.]
Isn't the real reason, takeoffs are optional, landings are mandatory

Serously, just look at the accident record to date. Military drones are operating with about ten times the mishap rates as fighters. Neither are quite as good as general aviation.

[goldfish85]

Basil: "In order to maintain order on board there has to be an appointed Pilot in Command. At the very least a change in law would be required to appoint, say, the chief steward as aircraft commander."

In my first week at my airline (sadly no longer in business) the instructor wnet through the order of succession: If the Captain is disabled, then the First Officer is in charge. If the First Officer is disabled (dot) (dot) (dot). At one point, If the Flight Engineer is disabled, the the senior flight attended is in charge. At that point he stopped. looked at the class and said "and you're in real trouble."

[tdracer]

A couple of observations...

The current capability of UAVs is close to meaningless - no one is arguing that current technology would make pilotless aircraft safer. We're talking a future state (just how future being also open to debate). But computing power is growing exponentially. That means triply redundant systems will be cheaper in 5 years than single thread stuff is today. Project that forward 20 or 40 years?

The first 'pilotless' passenger aircraft isn't going to be a 400 passenger commercial airliner - it'll be a private 'bizz jet' general aviation aircraft (or even a prop), certified under far more relaxed regulations than the Part 25 FAA and EASA regs. The safety record for general aviation currently isn't all that great - orders of magnitude worse than the Part 25 counterparts. So you've got some businessman who can justify and afford his own aircraft, but doesn't want to either pay for a pilot or become one. You honestly think the idea of a pilotless aircraft - as safe or safer than the admittedly mediocre level of piloted general aviation - wouldn't be attractive? Sure, some will crash, people will die - just like what happens with all to much regularity in general aviation today. But the bugs will be worked out, and before long unpiloted GA aircraft will be way safer than their piloted counterparts.

HIRF/Lightning - REALLY? Fly By Wire and FADEC have been commonplace for 25 years - HIRF/Lighting are just as much a threat to FBW/FADEC as they'd be for pilotless aircraft. Further, I can't recall a single accident attributed to HIRF/Lightning affecting a FBW or FADEC. Yes, it needs to be accounted for, but we've already figured out how to do that.

As I said before - it's not around the corner, at least not by human standards. however, I've been working engines since 1977. I recall all too well the skepticism regarding FADEC - including people swearing they'd NEVER fly on a FADEC powered airplane. 20 years later they had conveniently forgotten they'd ever said that FBW was met with similar skepticism. Yet today, if you want to fly commercially on an aircraft that isn't ultimately controlled by computers, your options are going to very sharply limited. Look at the current Boeing and Airbus products - yes pilots are in the loop, but they are NOT ultimately in control - the aircraft is being controlled by electrons.

[Tourist]

Faire dincome

Leaving aside the fact that the computer would also have had the AoA info to hand, plus why on earth would the computer have only those inputs. There are vastly more parameters available nowadays.
Even a mobile phone has accelerometers. pitot tubes are not the only way of measuring airspeed etc etc.

It is beside the point I believe anyway.
Let's assume for a moment that the computer had got into the position the humans did. Let's assume that we limited the proposed new airliner to old school sensors.
I believe I am correct in saying that after the initial pitot problem, for the rest of the flight the instruments actually worked correctly?
So all the info was available to the pilots, they just failed to interpret it and act accordingly. All a computer had to do was lower the nose.

Your misunderstanding of pitch and power worries me if you are in fact a pilot.
Pitch and power is not a stall recovery technique.
It is a flying with unreliable instrument technique.
It does not mean pitch the aircraft and add power as you seem to suggest.
It means that there are pitch and power settings that will give known level flight speeds.
Yes, Airbus and the rest of the sane world recommends lowering the nose to recover from a stall in an airliner. This is not beyond the wit of man to program into a computer.

Goldfish85
Please read back a few pages. Drone survival rates are covered and irrelevant for obvious reasons. More relevant would be if you found the loss rate due to computer failure.

Harryw
If you are linking to the X47B hoping that it helps the "no" case then you don't quite understand the difficulty leap between airliner ops and naval combat aircraft operation. 2 out of three certainly isn't bad. Diversions being out affect manned aircraft just as much as unmanned, and plenty of human pilots end up going home.
Fortunately, nobody is planning or proposing landing airliners on carrier decks in UAVs.

[Mr Optimistic]

'Wind up' ? No, just giving my views on the matter, hope that's OK with you. Sorry my opinions aren't unique and are shared by others.

[Tourist]

The original point was answered with a link to emphatic proof that the opinion was fallacious which you would have known if you bothered to read the link I posted.


I was asking if the repeated asking of questions that had been put to bed was a windup.

I see nobody disputing the facts that the system exists to do just what you said was a distant pipe dream.

And it exists on an android phone.
No exactly cray territory.

It is not that hard to google claims that you are about to post to see if they are factually incorrect.
I am trying when able to post a link to provide evidence to support my case.
Do me the courtesy of reading them before assuming they are irelevant.