A350-1000 autonomy tests aimed at supporting dual-pilot cockpit: Airbus

[slast]

Item from Flight International: interesting twist on the manufacturer claimed objectives.
A350-1000 autonomy tests aimed at supporting dual-pilot cockpit: AirbusBy David Kaminski-Morrow13 January 2023

Airbus has trialled further autonomous flight concepts using an A350-1000 test aircraft, including emergency diversion and automatic landing, but insists they are aimed at supporting a dual-pilot, rather than single-pilot, cockpit.

Part of the airframer’s UpNext initiative the ‘DragonFly’ programme takes its inspiration from insects of the same name – using biomimicry of dragonflies’ combination of high-speed senses, spanning a wide field of view, feature recognition and precision flight control.

DragonFly has focused on three particular aspects of autonomous operation – diversion, landing and taxiing – including assistance during a simulated incident of pilot incapacitation.

The aircraft relies on automatic intelligent decision-making based on flight data obtained through a combination of sensors, including cameras on the A350’s nose, and backed up by vision algorithms and guidance calculations.

DragonFly head of demonstrator Isabelle Lacaze says the programme has been running for two years, but flight-testing using aircraft MSN59 commenced in July 2022 and the airframer is “enthusiastic about the first promising results” obtained towards the end of last year.

Among the achievements, she says, was the simulation of sudden pilot incapacitation over Limoges during a flight to Lyon, in a real air traffic environment.

“The system took over and diverted the [aircraft] safely to the nearest appropriate airport – Toulouse in that scenario,” says Lacaze.

“It was really amazing to see that all our stakeholders in this scenario [including test pilots and air traffic controllers] were really positive about the feasibility and the acceptability of the solution.”

She states that the diversion function can activate automatically or act as a support to the pilots.

https://d3lcr32v2pp4l1.cloudfront.net/Pictures/480xany/6/5/1/91651_dragonflycairbus_460240.jpg

Source: Airbus

Cameras mounted on the A350’s nose are part of the sensor array to gather external situation data

Addressing the possibility of incapacitation remains one of the contentious issues when considering a possible shift towards single-pilot operations.

But Lacaze insists that DragonFly is intended to develop and evaluate assistance to a two-pilot crew, to help manage workload during intensive phases of flight, such as diversion – with cases of incapacitation or medical emergency among the possible causes.

“We consider there is already huge value [in this],” she says. “We are in exploratory field of research. Our framework – to mature, to explore, see what we can do with the new technology – is really focusing on… assistance for dual-pilot operations, meaning the current crew composition.”

But DragonFly does envision support for extreme circumstances in which pilots are no longer able to control the aircraft.

The criteria for detecting incapacitation are “sensitive”, says Lacaze, and she does not elaborate on the techniques involved, but says that the system is designed to take over “seconds” after detection.

Lacaze says the DragonFly demonstration team has had to “fine tune” the implementation through a step-by-step process, gradually integrating different aspects, to ensure that safety was preserved as the aircraft took control. “It was not an easy walk in the park,” she says.

She says the programme is intended not only to offer diversion assistance but to support crews in taxiing at congested airports, through computer-vision obstacle detection, to maintain speed control and centreline tracking, as well as alerting pilots to the risk of runway crossing. The taxi functions have been tested at Toulouse airport.

https://d3lcr32v2pp4l1.cloudfront.net/Pictures/480xany/6/5/0/91650_dragonflytaxicairbus_488697.jpg

Source: Airbus

DragonFly also used vision systems to support taxiing assistance

Prior autonomous flight testing using an A350-1000, under a programme named ATTOL, demonstrated a fully-autonomous take-off using vision systems to track the runway, and carry out automatic rotation.

DragonFly has extended this to automatic landing, including in low visibility conditions using sensors and vision systems, and Airbus believes the capability could be broadened to approaches to any airport, even those not equipped with the ground technology to support landing automation.

Airbus says its demonstrator takes into account a range of external factors – including terrain, weather and military flight zones – as part of its assessment of where to land, and generates a new trajectory.

It adds that the aircraft also uses constant communication with air traffic control and the airline operations centre to co-ordinate a safe approach.

Several Airbus divisions and a number of partner companies, including Collins Aerospace, Honeywell, Thales, Onera and Cobham, have co-operated with the DragonFly programme, which has embarked on its last three months of testing.

“These tests are one of several steps in the methodical research of technologies to further enhance operations and improve safety,” says Lacaze.

She says the UpNext initiative is launching a project to advance landing and taxi assistance with a view to taking advantage of new-generation computer-vision algorithms.

[172_driver]

Is this the software developers having run out of ideas for yet another feature?

Isn't it already doing automatic TCAS maneuvers and emergency descents?

[Less Hair]

It sounds like that Garmin general aviation autoland feature scaled up to airliner sized aircraft. This will be what the new push button at door 1L is for.

[ATC Watcher]

Is this the software developers having run out of ideas for yet another feature?

Isn't it already doing automatic TCAS maneuvers and emergency descents?
No , it is how Airbus works, it let their brightest engineers develop new ideas and support prototyping. Not all their bright ideas or prototypes are kept , look for instance at the 2 or3 urban previous VTOLs, or their 2 electric aircraft, in each case a prototype was build, flown then abandonned.
Dragonfly here is another of thsose , but the long time aim is autonomous aircraft for sure. Wether some of hese features will see production , will it be standard or as an (expensive) option , time will tell.

As Less hair said the basic technology already exsists, ( Garmin for the auto diversion / land and Gulfstream for the recognaissance cameras ) putting iall this into practice and having it certified for a 300 seat aircraft will be the challenge I guess, But you have to admit it is very interesting .

[Bergerie1]

Airbus is right to research such things and, remember, in all R&D there will be some projects which reach dead ends and some which will open up unexpected avenues.

However, while I don't think we will see pilotless airline operations in the near future, I do think it will be possible, in the not too distant future, to design and certificate aircraft for single pilot operations in the cruise so that, on very long flights, it would be possible either to reduce or even eliminating the need for 'heavy crews'. This would enable one pilot to take his statutory rest while the other is left alone at the controls. There would then be no need, depending on the length of the flight, to roster extra pilots.

You could then have, on ultra long flights, two at the controls for take-off and landing but only one at the controls during the quiet parts in the cruise with the other readily available in the crew rest area - much like the watch system at sea. I am sure this could work and would offer substantial financial benefits to the airlines.

It would also allow for co-pilots to be trained and to build up sufficient experience so that they could take sole control in the cruise. In practice, I think there would need to be an extra co-pilot or apprentice pilot period (call it what you will) for a long time until the necessary experience has been gained. Experience is not something that is acquired quickly and takes several years. As we all know, good piloting skill is not just about flying the aircraft but also about knowing what situations to avoid.

Finally, I think that long solo duty periods would be so ruddy boring I doubt any sensible person in their right mind would ever want to do the job. I know I wouldn't!!

[Less Hair]

Two guys are needed to talk to each other so one can learn from the experiences of the other guy and see how things are done.

[Innaflap]

For sure this is the type of feature that could help in the event of an incapacitated crew or nefarious forces

[VHOED191006]

For sure this is the type of feature that could help in the event of an incapacitated crew or nefarious forces

As long as it is used for things like that, and nothing further. I've always liked Airbus for what they're doing with technology. Hopefully they don't consider using it to remove the humans who are meant to fly it.

[Less Hair]

"Welcome to your airline flight safety service center. For engine No. 1 fire extinguisher discharge press star one. To divert to the next available field press hash 911."

[atakacs]

Autonomous & automated flying is coming, sooner or later, for the better or the worse. These are all small steps in that direction.

[Kennytheking]

Airbus today suffers from a LOC* anomaly where it sometime cannot properly capture the localiser. This is something aeroplanes have been doing for, like, 90 years and Airbus still can't get it right. I'm not too sure that their autonomous project is going to pass muster.

[megan]

Autonomous & automated flying is coming, sooner or later, for the better or the worseNever happen, pilot error would just be replaced with errors in software code and architecture design of the black boxes, we've had an A320 that was written off because the black boxes wouldn't allow the crew to flare for the landing, an A330 that did its best to crash the aircraft, and a 777 out of control because the black boxes had a failure that wasn't supposed to be able to occur. Thank God in all cases there were two aviators in row 0 who knew their business.

[KRviator]

And how many times have those humans - because they sure as hell weren't 'pilots' yet alone aviators - flown perfectly serviceable aircraft into a hill, an ocean or elsewhere because they screwed up, yet alone committed suicide and decided to take their passengers with them?

You've cited three examples of tech errors, and I'll grant there are likely many more - but "pilot error" outstrips technological failure especially these days, by an order of magnitude. A computer isn't going to shut down the wrong engine, like the crew did at Kegworth or Taipei. A computer isn't going to fly a 777 into a sea wall on a CAVU day that my 10 year old can (and does) land in and a computer isn't going to panic and hold full back stick while falling 35,000' into the Atlantic. There'll be new and novel failures but they'll still be rarer than the number of crew-induced errors today.

That being said....What scares me about the pilotless airliner is the same thing that scares me about those robot trains around the world. I'm paraphrasing here but, there are known unknowns - things we know we don't know, and there are unknown unknowns - those things that no one knows they don't know. And those are the things that are going to limit the introduction and deployment of autonomous airliners, because "It's not supposed to do that" doesn't cut the mustard when things go pear shaped in the flight levels.

[megan]

Just picking a year at random 2019 there were 12 crashes with the loss of 261 lives, 157 of those were the Ethiopian MAX. I very much doubt pax would accept no crew, I certainly won't, particularly with QF 32 in mind.

[vilas]

Just picking a year at random 2019 there were 12 crashes with the loss of 261 lives, 157 of those were the Ethiopian MAX. I very much doubt pax would accept no crew, I certainly won't, particularly with QF 32 in mind.
WIth more reliable autonomous aircraft when it has already proved it's reliability for cargo operations it will be accepted, especially with significant price difference. Indonesian max on an earlier flight was landed safely by acting differently. But this time MCAS completely confused the pilot. So two humans in front reacted differently to the situation at different time. So it will be a matter of what makes you more uncomfortable the machine or the man.

[ATC Watcher]

Just picking a year at random 2019 there were 12 crashes with the loss of 261 lives, 157 of those were the Ethiopian MAX. I very much doubt pax would accept no crew, I certainly won't, particularly with QF 32 in mind.
The Ethiopain MAx is probably not the best example to take, there are still a lot of knowledgeable people out there that believe that had the crew followed strictly the new MCAS poscedure the aircraft would not have crashed. aAfully automated aircraft would have followed the procedure, cannot be programmed otherwise.
But to be even more the Devil's advocate. let's take the latest incident covered here : the AA near ground collsion at JFK. Had AA been fully automated , the ATC computer would have instructed dep 4L. cross 31L at K, and the aircrfaft would have just followed that and not end up crossing 4L. But that said. when dealing with full automation ( and ATC my line of work we have the same debate ) I keep on reminding engineers one of the good old Al Wiener law : " Whenever you solve a problem you usually create another one. You can only hope that the one you created is less critical than the one you eliminated."
Fully automated aircrfaft will have their own issues . It won't be easy , and it will take time . But one day . especailly after a big pilot blunder taking many lives.. who can say never..?..

[Innaflap]

As long as it is used for things like that, and nothing further. I've always liked Airbus for what they're doing with technology. Hopefully they don't consider using it to remove the humans who are meant to fly it.

Mankind isn't ready for autonomous cars - and such cars are not ready for mankind

[Busdriver01]

the AA near ground collsion at JFK. Had AA been fully automated , the ATC computer would have instructed dep 4L. cross 31L at K, and the aircrfaft would have just followed that and not end up crossing 4L.

So is ATC to be fully automated too? No reason to think human ATC is safer than human pilot...

[atakacs]

So is ATC to be fully automated too? No reason to think human ATC is safer than human pilot...
ATC and piloting will eventually be fully automated. Obviously a very long process, fraught with valid and irrational issues. But assuming a continuation of our technological civilization it will happen withing the next 20 to 50 years.

[Chiefttp]

Autonomous & automated flying is coming, sooner or later, for the better or the worse. These are all small steps in that direction.

So the fact that 25% of people are still wearing masks in airports in 2023, suddenly these brave, intrepid, pax will board an airliner with no pilots? And I’ve stated this before, if my company wants me to fly a single piloted airliner, I will ask for my current salary, plus my F/O’s current salary, plus a bonus for the additional responsibility that will fall into my lap…So where are the cost savings? And safety enhancements?

[safetypee]

Recent posts focus on the emotive no pilot autonomous operation, whereas the Airbus research is considering dual-pilot operations.

Instead of stating what we do not want, consider what we might require in a dual pilot cockpit, when, why, where. For what purpose, workload, safety viewpoint, or new or changed risks.

Is the type of automation being considered what crews will expect; do we sufficiently understand current operations, ourselves, in normal, high workload, or abnormal situations before considering the hypothetical.

What are the current critical situations; how are these identified and understood.

[ATC Watcher]

So is ATC to be fully automated too? No reason to think human ATC is safer than human pilot...
That is still te long term plan , yes. the technology is not new, was tested already in large simulations both in the USA and in Europe, It works but does not deliver capacity , so does not pass cost benefit analysis, yet . One paragdim was that fully automated ATC needs a fully automated aircraft too. Basically ground computers talking constantly to airborne computers using high broad band data exchange. . That data link however does not exists yet.
Then will come what they call the "emotional aspect" : basically acceptance by the public. Some believe if the price of the ticket is right , i.e. much lower than manned flight , they will go for it . Not sure of that but here you go.
@ Chieftp
if my company wants me to fly a single piloted airliner, I will ask for my current salary, plus my F/O’s current salary, plus a bonus for the additional responsibility that will fall into my lap…So where are the cost savings? And safety enhancements?
I am sure they will then most probably recruit and train youngsters to do this for half you current Capt salary. As to the safety enhancements, technically they do not have to be enhanced, it is still 10 to the minus 6 , we are currently well below that, so maintaining current saftey is enough. ( to avoid nasty remarks : I am beeing sarcastic here :suspect: )

[threep]

Freight doesn't get a say on who is piloting the aircraft, so that's where they will start with autonomous/automated flights. It does surprise me how willing some members of the public are willing to put their safety in the hands of "full self driving" cars despite such autonomous driving systems not being certified for autonomous driving.

I think we are likely to see full autonomous vehicles on the roads within the next 10 years, but they too will be carrying freight between distribution centres.

[KRviator]

So the fact that 25% of people are still wearing masks in airports in 2023, suddenly these brave, intrepid, pax will board an airliner with no pilots?A mate of mine still wears a mask when he's out and about and he's probably one of the smartest blokes I know. Not because he's scared of Covid, but because he has a legitimate reason he has to avoid getting crook. To ridicule is cheap and easy, and the oldest trick in the book...
And I’ve stated this before, if my company wants me to fly a single piloted airliner, I will ask for my current salary, plus my F/O’s current salary, plus a bonus for the additional responsibility that will fall into my lap…So where are the cost savings?Yeeaaahh, about that....How'd that work out for those pilots who took on the Navigation duties when they got rid of the Nav? Or those 767 Driver's who took over from the F/E back in the 80's? Or all those train driver's after they aboloished the Guard on the back of the train, or the Coey beside them whem they went to Driver-Only Operation?!?

You can ask, but you won't get. And all they will do is redefine your role s they have done to countless other industries to reduce numbers.

And safety enhancements?Really? Do you really want to go there?

Asiana 214
Air France 296
American 965
Korean 801
GermanWings 9525
SilkAir 185
EgyptAir 990
Air France 447
Transasia 235
BMA 92
Eastern 401
China Airlines 006
Singapore 006
Tenerife

And countless others besides where the crew have flown a perfectly serviceable aircraft into an obstacle, or grossly mishandled a relatively minor fault and the aircraft was lost. As I've said above, there's likely multiple times the tech crew have saved the aircraft from errant technology - however do not confuse pilots up front as being indefensible in terms of preventing fatalities, particularly when pilot suicide is a significant cause of fatalities in recent years. There'll also be new and novel failures experienced with autonomous operations, but I proffer that even with these losses, safety will be improved overall by removing the human element.

[davidjpowell]

A mate of mine still wears a mask when he's out and about and he's probably one of the smartest blokes I know. Not because he's scared of Covid, but because he has a legitimate reason he has to avoid getting crook. To ridicule is cheap and easy, and the oldest trick in the book...
Yeeaaahh, about that....How'd that work out for those pilots who took on the Navigation duties when they got rid of the Nav? Or those 767 Driver's who took over from the F/E back in the 80's? Or all those train driver's after they aboloished the Guard on the back of the train, or the Coey beside them whem they went to Driver-Only Operation?!?

You can ask, but you won't get. And all they will do is redefine your role s they have done to countless other industries to reduce numbers.

Really? Do you really want to go there?

Asiana 214
Air France 296
American 965
Korean 801
GermanWings 9525
SilkAir 185
EgyptAir 990
Air France 447
Transasia 235
BMA 92
Eastern 401
China Airlines 006
Singapore 006
Tenerife

And countless others besides where the crew have flown a perfectly serviceable aircraft into an obstacle, or grossly mishandled a relatively minor fault and the aircraft was lost. As I've said above, there's likely multiple times the tech crew have saved the aircraft from errant technology - however do not confuse pilots up front as being indefensible in terms of preventing fatalities, particularly when pilot suicide is a significant cause of fatalities in recent years. There'll also be new and novel failures experienced with autonomous operations, but I proffer that even with these losses, safety will be improved overall by removing the human element.

What is not reported there, because no crash occurred where human intervention has prevented crashes.

[Less Hair]

I think the step should be either two pilots or no pilots. However, we are a long way from the reliability and safety needed for commercial passenger operations without pilots. The flying public will not accept cheapo bot flights.
Two are needed for redundancy and educating future captains. No pilots might work for freighters electronically hooked up to a manned formation flight leader or similar. And then the freighter can leave the remote monitored formation to autoland garmin style at some remote desert airport or similar.

[FullWings]

I think the step should be either two pilots or no pilots. However, we are a long way from the reliability and safety needed for commercial passenger operations without pilots. The flying public will not accept cheapo bot flights.
Two are needed for redundancy and educating future captains. No pilots might work for freighters electronically hooked up to a manned formation flight leader or similar. And then the freighter can leave the remote monitored formation to autoland garmin style at some remote desert airport or similar.
That’s my take on it as well. Beyond a certain flight duration, you have to assume a single pilot aeroplane will have no-one at the controls for periods of time (or for the rest of the flight if they become incapacitated), therefore the aircraft will need to be able to deal with anything the pilot could. At that point, why have a pilot? You’ve already paid for an autonomous aircraft...

[Easy Street]

Someone mentioned MCAS as an example of technical failure. My take on that is that it was a technical failure which only came about due to the requirement to make the aircraft feel the same to the pilots. No pilots = no MCAS...

When Airbus gets automatic optically guided flight sorted out, which it inevitably will, the clock will be ticking for professional piloting as we know it. If the cost is that the next Miracle on the Hudson turns into a disaster because the computer couldn't work out what to do, the benefit of having avoided the dozens of pilot error accidents that would have happened in the meantime will have been worth it.

[Chiefttp]

Kraviator,
There will always be exceptions to the rule, you’re cherry picking incidents and accidents, and by the way, as far as I know 100% of Airliner accidents have occurred with a manned crew, so stats are irrelevant in this discussion. Here’s a statistic I was privy to, but most folks aren’t. Military drones crash, a lot….I’m Glad your friends wears his mask,sounds like he has a valid reason to do so, however, I’d bet a lot of money that 80% of the people I see wearing masks in airports in 2023 are wearing them for psychological security and not medical reasons. Masks have become the Adult equivalent of the baby pacifier. All I’m saying is these types of passengers will be terrified to board an aircraft without a pilot. As far as Freight ops, as a freight pilot, I’m not worried, because after the first 747 unmanned freighter crashes into a farmer’s field, or, God Forbid, a populated city, unmanned flights will cease.Hell they could barely fly the Concorde in the 1970’s because of a small noise issue, and you really think people will be OK with a 650,000 pound aircraft loaded with 100,000 lbs of jet fuel flying overhead with no pilot.?

[Captain Dart]

I still cannot envisage how a robot is going to pick its way through the ITCZ. Or thunderstorms in a terminal area. Or avoid cloud layers for a smoother ride. The pax will just have to keep belted in and stay terrified.

Recruiting cabin crew may also be a problem.

[ferry pilot]

Teachable autopilots have been considered and likely quietly developed for several years.
Would the integration and perfection of this technology allow for a gradual acceptance and smoother transition to autonomy over time?

[MissChief]

Think of salary structures. Most Long Haul operators promote pilots from their Short/Medium Haul Ops. Would SH/MH folk take a big drop in earnings to work as a back-up cruise pilot on the LH fleet? Obviously not. I cannot see where the airline's cost savings would come from given the salary structures of SH and LH being roughly comparable. A well-paid SFO on SH is far more likely to want to move to the left seat on SH than step way down in earnings to right-seat/sleeper seat on LH. And how would the new and lowly-paid FO/SFO in Long Haul ever get trained properly? While they are asleep? And does the Captain stay awake for the whole flight?

[Klauss]

Think of salary structures. Most Long Haul operators promote pilots from their Short/Medium Haul Ops. Would SH/MH folk take a big drop in earnings to work as a back-up cruise pilot on the LH fleet? Obviously not. I cannot see where the airline's cost savings would come from given the salary structures of SH and LH being roughly comparable. A well-paid SFO on SH is far more likely to want to move to the left seat on SH than step way down in earnings to right-seat/sleeper seat on LH. And how would the new and lowly-paid FO/SFO in Long Haul ever get trained properly? While they are asleep? And does the Captain stay awake for the whole flight?
I think that the real airline bosses are not even after cost savings: they are after power. They don´t want anyone to interfere with whatever they deem to be the best operation, the best PLAN . So, if there are no pilots, no-one is going to say ´no´ , no one is going to go on strike... and the bosses can rule.
Incientally, they save the $ for the pilots, too, but that´s a secondary consideration.
Pilots´ responsibility ? Will be legislated away from the airline, just as the risk of rockets falling onto earth-dwellers has been legislated away. Look to the Virgin rocket : it almost fell onto Teneriffe !
Thus, no-one in the airline hierarchy is personally responsible anyway...and the auto-pilots can do their job, one way or another.

Slightly scary thoughts, I admit. But cost can´t be the real point of getting pilots out of cockpits. You need to pay for a ton of research, electronics and extra maintenance to make that auto auto auto-mated cockpit happening.
Yes, technically, it can be done - as long as those auto auto auto-planes are the only objects in the sky and nothing goes wrong. Russia flew their space-shuttle around the world decades ago - no humans on board.
Many military organizations fly UAVs halfway around the world.
But: look at the safety record. Not good. (( mentally subtract those ´shot down´): https://dronewars.net/drone-crash-database/