Automatic Takeoffs Are Coming For Passenger Jets (cnn.com) 89
New submitter LazarusQLong shares a report: In late 1965, at what's now London Heathrow airport, a commercial flight coming from Paris made history by being the first to land automatically. The plane -- A Trident 1C operated by BEA, which would later become British Airways -- was equipped with a newly developed extension of the autopilot (a system to help guide the plane's path without manual control) known as "autoland." Today, automatic landing systems are installed on most commercial aircraft and improve the safety of landings in difficult weather or poor visibility.
Now, nearly 60 years later, the world's third largest aircraft manufacturer, Brazil's Embraer, is introducing a similar technology, but for takeoffs. Called "E2 Enhanced Take Off System," after the family of aircraft it's designed for, the technology would not only improve safety by reducing pilot workload, but it would also improve range and takeoff weight, allowing the planes that use it to travel farther, according to Embraer. "The system is better than the pilots," says Patrice London, principal performance engineer at Embraer, who has worked on the project for over a decade. "That's because it performs in the same way all the time. If you do 1,000 takeoffs, you will get 1,000 of exactly the same takeoff."
Now, nearly 60 years later, the world's third largest aircraft manufacturer, Brazil's Embraer, is introducing a similar technology, but for takeoffs. Called "E2 Enhanced Take Off System," after the family of aircraft it's designed for, the technology would not only improve safety by reducing pilot workload, but it would also improve range and takeoff weight, allowing the planes that use it to travel farther, according to Embraer. "The system is better than the pilots," says Patrice London, principal performance engineer at Embraer, who has worked on the project for over a decade. "That's because it performs in the same way all the time. If you do 1,000 takeoffs, you will get 1,000 of exactly the same takeoff."
more passengers? they are packed in to much right (Score:2)
more passengers? they are packed in to much right now!
Re:more passengers? they are packed in to much rig (Score:5, Funny)
Re: (Score:2)
Re: (Score:2)
Bots need vacations also. My CyberTruck wants to visit Australia in order to run over kangaroos, making them pancakes, and serving them in iHop.
Tickets for pilots seats (Score:2)
Maybe they'll sell tickets for pilots seats next. Branded as 'Thrillseeker Class' seats.
Re: (Score:2)
Re: (Score:2)
United used to let you listen to ATC radio traffic on one of the armrest channels. I always enjoyed that, following the flight as the pilots received their instructions.
Re: (Score:2)
will they use the same test plan as BOEING MAX? (Score:2)
will they use the same test plan as BOEING MAX?
Re: (Score:2)
The Boeing MaxHAL
"Sorry Dave, but I cannot take you back to your Country. And your luggage is sitting snuggly next to Bin Laden. Would you like to hear a song about bicycles?...Too bad, I'm singing it anyhow..."
Re: (Score:1)
No, because it will never actually be implemented. There's no need for this. Takeoffs are easy and there's no reason to automate them. Airlines will not pay for it since its not needed or wanted. This is just Embraer trying desperately to remain relevant.
Automated takeoffs are useful ... (Score:1)
... as a stepping stone to fully-automated (747-sized drone) passenger service.
Re: (Score:2)
... as a stepping stone to fully-automated (747-sized drone) passenger service.
Cargo planes will be automated first.
The technology already exists. The barriers are mostly legal and political.
get exactly the same - if many things are the same (Score:5, Insightful)
" If you do 1,000 takeoffs, you will get 1,000 of exactly the same takeoff."
If weather and atmospheric conditions are the same, if the plane is the same, if the the plane's configuration and loading is the same, if ...
Probably not (Score:5, Interesting)
" If you do 1,000 takeoffs, you will get 1,000 of exactly the same takeoff."
If weather and atmospheric conditions are the same, if the plane is the same, if the the plane's configuration and loading is the same, if ...
I used to work with one of the engineers doing automatic plane landings.
He pointed out that they had to put some random variation into the system, because the system was so precise it would always land in the same spot - quickly wearing out that specific spot in the runway.
The software uses feedback control based on position. I don't think atmosphere and weather conditions, or loading and configuration make as much difference as you suggest.
Re: (Score:1)
That sounds urban-legendy.
Re: (Score:2)
It sounds like a computer controlled airplane.
LOL. "But, computers, ..."
:-)
Where the computer controlled aircraft specifically touches down will vary upon when the ground crew last washed the aircraft.
Re: (Score:2)
Where the computer controlled aircraft specifically touches down will vary upon when the ground crew last washed the aircraft. :-)
Big crosswinds etc will still cause effects but the computer can detect undesired motion and correct it literally before a human would have even noticed it. The computer can correct for problems like excessive drag due to bug guts and it takes no more or less of what it substitutes for concentration and skill to do that than to fly the plane in ideal condition.
Re: (Score:3)
I totally believe it. Runway wear is a real thing -- https://skybrary.aero/articles... [skybrary.aero]
Re: (Score:2)
Re: (Score:2)
Sounds like some of the creative variables used in "scientific math" that theoretical physicists use in their theories to explain the Universe ... in order to make all the various bits & bobs fit the theory.
Kindly elaborate on "scientific math" and "bits & bobs".
Theories are designed to fit observations, not the other way around.
Re: (Score:2)
Have a look at Einstein's "geometric approaches" to what is now considered "Classical Unified Field Theory".
He had to impose relaxations on certain theoretical aspects and his maths & abstractions are TRULY NOT his best work.
So whatever he observed he was clearly unable to form a working theory to explain it that would withstand close scrutiny.
Therefore, these Einsteinian ideas are generally disregarded in this line of thought & research.
Re: (Score:2)
I assume you mean this. [wikipedia.org]
Einstein can be forgiven for exploring a cul-de-sac in the latter part of his career. Many great scientists have been distracted by interesting but ultimately fruitless endeavors. But none of our thread-discussion has anything to do with applying "fuzz" to auto-landing software calculations in order to avoid wearing down runways in exactly the same place for every landing.
Re: (Score:2)
Re: (Score:2)
Glad to see we're back on topic.
Okian Warrior was talking about a feature in landing software that varied slightly the position on the runway for each auto-landing. How is that a threat to human lives? Arguably it's safer to add such variation, because it avoids premature failure of the runway due to repeated friction in the same spot(s).
Re: (Score:3)
I don't think atmosphere and weather conditions, or loading and configuration make as much difference as you suggest.
They do, for example HIGH HOT HUMID will require more runway to take off.
Re: Probably not (Score:2)
I've noticed recently that planes I'm on seem to be going further down the runway and not as aggressively powering the engines. I wonder if this saves a significant amount of fuel on takeoff or not.
Re: (Score:1)
It gives you a safety margin after take off, because you have a higher speed.
E.g. you lose an engine just after lifting off. The higher speed makes it more easy to cope with that.
Re: (Score:2)
Yes it's called flex temperature at just on Boeing planes. The pilot inputs a higher temperature into the performance calculator. See https://en.wikipedia.org/wiki/... [wikipedia.org]
Re: (Score:2)
performance limited take off are better for the plane and less noisy for neighbors.
Re: (Score:2)
I've noticed recently that planes I'm on seem to be going further down the runway and not as aggressively powering the engines. I wonder if this saves a significant amount of fuel on takeoff or not.
Sounds like a departure from safety. My understanding is that pilots want to take off quickly, leaving plenty of runway in case they need to abort before the wheels leave the ground.
As for saving fuel? Meh, I don't think you'd save that much by taking off more slowly, compared to fuel-consumption for the multi-hour flight.
Re: (Score:2)
Performance-limited take offs are standard procedure. They do save a lot of fuel and are more comfortable for passengers. Also they are quieter for the neighbors. Watch YouTube videos from Just planes. It's quite interesting how pilots set it up. Flex temperature.
Re: (Score:2)
They do, for example HIGH HOT HUMID will require more runway to take off.
In the context of the discussion, I don't think that "HIGH" varies very much for a specific airport.
Re: (Score:2)
They do, for example HIGH HOT HUMID will require more runway to take off.
In the context of the discussion, I don't think that "HIGH" varies very much for a specific airport.
Excellent point :-)
Re: (Score:2)
I don't think atmosphere and weather conditions, or loading and configuration make as much difference as you suggest.
I'm not sure if you've seen a plane land in bad weather, but I do genuinely wonder if a computer can cope with maneuverers like this one: https://youtu.be/Mk0xSrji8gc?s... [youtu.be] When you approach the runway looking out the side window rather than the front one, weather conditions have made a difference to how you are landing.
Re: (Score:3)
I'm not sure if you've seen a plane land in bad weather, but I do genuinely wonder if a computer can cope with maneuverers like this one
To the computer, that's no different from any other conditions under which it operates. It "knows" where it's trying to go, it "knows" how to adjust the throttles and control surfaces to make it happen, and landing like that is no different to it from making any other landing. It has a better picture of what the plane is doing at any given time than a human pilot, because its sensors and software react faster than our senses plus the brain that has to make sense of them.
Re: (Score:2)
Auto land can indeed cope with a cross wind landing like that. Probably better and smoother than a human. Provided of course that the crosswind is within the set limits if the auto land system. Otherwise it will kick out. Under such circumstances the human pilot would probably initiate a go around anyway, perhaps diverting to an alternate.
Re: (Score:1)
It does make an enourmous difference.
Higher cargo load means more fuel.
Additional fuel means eben more additional weight.
And that means the engines have to deliver more power to get the plane to take off speed.
Of course you could argue, the algorithm only needs to ramp up acceleration, aka power, until it is right.
But keep in mind: everything the plane is trying to do, is affected by its weight.
Re: (Score:2)
But keep in mind: everything the plane is trying to do, is affected by its weight.
It would be interesting to know if automatic takeoffs are better or the same if the weight is miscalculated. It has happened [theregister.com] in the past:
check-in software treated travelers identified as "Miss" in the passenger list as children, and assigned them a weight of 35 kg (~77 lbs) instead of 69 kg (~152 lbs) for an adult.
Mentour has an interesting video [youtu.be] on this (TUI Flight 7226). The flight was a ton heavier than the pilots were expecting.
Re: (Score:2)
That information is -already- programmed into the flight control computer by the pilots. Auto-pilot and auto-landing systems -already- account for these variables. Do you think they'd just -forget- about them when extending those exact same systems to support takeoffs as well? This isn't even remotely new technology. Pre-TFS, literally the beginning of the first line, auto-landing started in 1965. This shit has been around a very VERY long time, and some of the most battle-tested systems in engineering anyw
Re: (Score:1)
That information is -already- programmed into the flight control computer by the pilots. Auto-pilot and auto-landing systems -already- account for these variables. Do you think they'd just -forget- about them when extending those exact same systems to support takeoffs as well? This isn't even remotely new technology. Pre-TFS, literally the beginning of the first line, auto-landing started in 1965. This shit has been around a very VERY long time, and some of the most battle-tested systems in engineering anywhere in the world with the strictest safety regulations.
And manual or computer controlled, the aircraft flies differently depending upon external conditions. For example HIGH HOT HUMID will require more runway to take off.
Re: get exactly the same - if many things are the (Score:2)
Maybe so, but sensors for temp and humidity are available. In fact, you could plaster those and ambient pressure sensors along the control surfaces for a super-localized sense of the real drag and lift on them.
Computer systems can process many more senses that are butt-seat sensed by humans.
Re: (Score:2)
All of which (and more) you input into the flight management system right now so that it can tell you the appropriate thrust settings and important speeds.
Re: (Score:1)
All of which (and more) you input into the flight management system right now so that it can tell you the appropriate thrust settings and important speeds.
Yes, automated or manual, the physics of flight is the same. For example High Hot Humid needs more runway to take off.
Re: (Score:2)
Adaptive controls don't assume any of that. Pilots are behind many control systems already. It's no longer the Wright brothers' aircrafts where they moved their weight around and pulled ropes and levers to control the flight. In turbulent conditions, it's the onboard control systems that are doing the job. Same with many other rough weather conditions. Pilots give high level orders, sort of. That's why we had the Boeing Max crashes. Landing is yet another task for these control systems.
Re: (Score:2)
Adaptive controls don't assume any of that. Pilots are behind many control systems already. It's no longer the Wright brothers' aircrafts where they moved their weight around and pulled ropes and levers to control the flight. In turbulent conditions, it's the onboard control systems that are doing the job. Same with many other rough weather conditions. Pilots give high level orders, sort of. That's why we had the Boeing Max crashes. Landing is yet another task for these control systems.
Whether the flight is manual, assisted, or fully automated. Conditions like HIGH HOT HUMID make the takeoffs different.
F/A-18's been doing it for a long time (Score:3)
Now, nearly 60 years later, the world's third largest aircraft manufacturer, Brazil's Embraer, is introducing a similar technology, but for takeoffs.
US Navy F/A-18 Hornets have been doing automatic takeoffs for decades.
Re: (Score:3)
Re: (Score:2)
Fair point, although a commercial airliner takeoff isn't so much like a pebble in a slingshot. The airliner, in it's long, un-tethered voyage down the runway, has to propel itself, and steer, and stuff.
My understanding is that steering is needed from about the point where the flight deck ends and the catapult is no longer involved. The auto takeoff establishes a stable flight for the pilot, then the pilots takes their hands off the safety handles and touches the controls.
Re: (Score:1)
So have Ford Pintos.
Re: (Score:2)
Re: (Score:2)
So have Ford Pintos.
I'm not sure any cars launched from carriers ever established a stable flight, just an unpowered ballistic trajectory to the water.
:-)
Did I miss a JATO video?
Re: (Score:3)
Re: (Score:3)
Throwing a Hornet off a carrier the standard hands-free way isn't close to the same thing as all of the checks and variables involved in getting a jumbo jet airborne down a long runway.
True, it's more difficult in the F/A-18. F/A-18 manual control during launch and early flight is considered too dangerous, unlike a jumbo jet takeoff on a long runway. The F/A-18 auto launch is like traction control in a car, beyond most human's performance. Auto takeoff in the airline is like cruise control, a convenience for something within human abilities.
Re: (Score:1)
That is nonsense.
Every plane launched from a carrier is launched by a catapult.
And that means the front wheel is locked into a sledge.
And the pilot is not even holding the stick, because of the absurd acceleration.
The only thing automatic perhaps, is that after release from the catapult some assist system keeps the plane flying automatically until the pilot can grab the stick.
And how long might that be?
Re: (Score:2)
And the pilot is not even holding the stick, because of the absurd acceleration.
Actually, it's because if you were to try to give positive feedback to the plane, you would crash it. This has been true for catapult launched planes for a long time.
You trim them into a take-off configuration so that they'll take off essentially automatically until it's safe to touch the stick.
The only thing automatic perhaps, is that after release from the catapult some assist system keeps the plane flying automatically until the pilot can grab the stick.
Correct. Though, not "until they can", it's "until it's safe to".
A catapult takeoff is a high-G maneuver, but it's nowhere near the highest G maneuver those jets make. It's not like there's a physical inability to
Re: (Score:2)
That is nonsense. Every plane launched from a carrier is launched by a catapult. And that means the front wheel is locked into a sledge. And the pilot is not even holding the stick, because of the absurd acceleration.
I am not taklking about the catapult shot. I am talking about the earliest moments of flight when the aircraft leaves the catapult. That is where the Navy beliefs there is enough instability to make initial computer control to stabilize the aircraft warranted. The pilot is not holding the stick at this point either because pilot inputs are not wanted yet.
The only thing automatic perhaps, is that after release from the catapult some assist system keeps the plane flying automatically until the pilot can grab the stick.
And how long might that be?
The Navy specifically states initial instability as the motivation.
Re: (Score:2)
Flying a 747 is fucking grade school compared to keeping an F18 in the air after you've fired it in a fucking slingshot.
It's so aggressive that they don't dare let a pilot try to do it. If you don't have the precise AOA at the end of the deck, you're fish-food.
now get an AI to buzz the tower! (Score:2)
now get an AI to buzz the tower!
Re: (Score:2)
Saw This One (Score:1)
Kirk: "Does M-5 know this is just a drill?"
Daystrom: "Of course."
Ask Sully (Score:1)
Re: (Score:2)
The Airbus he was flying already had a computer controlled fly-by-wire system with flight envelope protection built in - the Airbus A320 family has had that since the 1980s.
Sully praised the system during that flight, as it took away a lot of the hard work around angle of attacks etc and allowed the crew to concentrate on things like where to actually land.
Re: (Score:2)
If AI had been around and overrode his decision to ditch in the Hudson, they could've fell short of a runway and likely killed everyone on board.
I think the point of AI would be to allow a pilot with less experience and skill than Sully to safely ditch in the water. The AI would know just as well as Sully that the aircraft could not reach the airport, but unlike the many pilots who fail in their return attempts and crash (obviously not many commercial airline pilots in that count).
Mode 4 Landings (Score:5, Interesting)
The plane I worked on in the Marine Corps had mode 4 landing capability (the ability to land itself, even when approaching a pitching aircraft carrier deck in rough weather). The story going around at the time was that the mode 4 was so good some error had to be introduced because the aircraft wheels and tail hook were hitting the exact same spot on the carrier deck every time. This caused the non-skid on the surface of the deck to be worn off in record time. This was back in the late 1980s. I wonder why it took so long to develop an automatic takeoff mode? The pilots of the plane told me they avoided mode 4 unless absolutely necessary for fear of losing, or eroding, their pilot skills. This would probably be a genuine concern today for pilots just sitting back and letting CPUs do their jobs as well.
Re: (Score:2)
I think we are at the point where the cost/benefit analysis say more soldiers will be saved due to greater accuracy with scopes than those lost due to unfamiliarity with iron sights.
At some point it will be the same with automated landings.
Re: (Score:2)
I wonder why it took so long to develop an automatic takeoff mode?
The US Navy has a non-zero acceptable casualty rate. Commercial airliners have an acceptable rate of zero. :-)
Re: (Score:2)
The US Navy has a non-zero acceptable casualty rate. Commercial airliners have an acceptable rate of zero. :-)
Airlines have fatality crashes and then get to continue to operate so long as lots of people with deep pockets can agree that it was the pilot's fault.
Re: (Score:2)
The US Navy has a non-zero acceptable casualty rate. Commercial airliners have an acceptable rate of zero. :-)
Airlines have fatality crashes and then get to continue to operate so long as lots of people with deep pockets can agree that it was the pilot's fault.
It's not about whether fatalities occur, it's about whether there is non-zero acceptable casualty rate. The US Navy is a military organization that recognizes the need for certain risky behaviors and the casualties that will result from that decision. Civilian airlines do not have to make such decisions.
Re: (Score:2)
The US Navy is a military organization that recognizes the need for certain risky behaviors and the casualties that will result from that decision. Civilian airlines do not have to make such decisions.
This is a lot of hot nonsense. Every flight is a risk. Commercial aviation maintenance schedules are a calculated risk.
Re: (Score:2)
The US Navy is a military organization that recognizes the need for certain risky behaviors and the casualties that will result from that decision. Civilian airlines do not have to make such decisions.
This is a lot of hot nonsense. Every flight is a risk. Commercial aviation maintenance schedules are a calculated risk.
You seem to be having problems with the word "acceptable". As a military organization that must take some risk to train, the expected and acceptable casualty rate is not zero. Civilian airliners have to accept no such enhanced risk so their acceptable casualty rate is zero. Civilian airliners can, and will play it safe, a military cannot always do so even in peacetime. In the last decade there are about 20-30 aviation deaths a year in the US military, for US Air Carriers its typical zero or one.
Re: (Score:2)
Re: (Score:1)
You are an idiot.
Unions have no say in what a pilot does and what not.
About time. No more Northwestf 255 crashes... (Score:3)
Such systems should be good at avoiding incidents like the Northwest flight 255 crash at Detroit Metro airport:
1. Crew turns off breaker to silence annoying redundant audio voice alarms/warnings.
2. Forget to turn it back on before takeoff.
3. Get interrupted by a radio call from the tower during checklist and miss lowering the flaps and slats.
4. Plane is pretty much fully loaded and nearly fully fueled. With flaps and slats not in takeoff configuration it doesn't have enough lift to achieve flight by the end of the runway with sufficient altitude to clear the fence, cross the adjacent interstate, and pass above the several (heavily occupied) multi-story hotels just on the other side.
5. Pilots try to take off with flaps and slats still up. The think things are a bit sluggish because of the load, and get far past the takeoff point of no return before realizing the flaps and slats aren't deplooyed. It's far too late to stop or change their setting. (They do figure it out: Nearly the last thing on the voice recorder is "Flaps! Flaps!")
6. To avoid killing the people in the hotels they fly the plane UNDER the freeway, through the airport entrance underpass. This tears off the wings, stops the aircraft body, and kills the crew and all the passengers except one four year old child.
Personal experience: Not having heard details of the crash, I was planning to drive to and from the Worldcon, which was at Phoenix that year. But I checked the airfares and was quoted a REALLY low price, so I booked the flight instead. Was puzzled when the plane only had something like three passengers. Turns out it was the (renumbered) replacement for the flight in question, a week later. Found out about it at the Con, but wasn't worried going back. I figured that if there was ANY flight they were REALLY being careful about, it was that one. B-b
Re: (Score:2)
That's a curious one indeed. Why didn't they reject the takeoff when it was clear the airplane could not fly? We'll never really know of course. Would more computers have helped with that? Maybe maybe not.
Re: (Score:2)
Past V1 probably.
Re: (Score:2)
Would more computers have helped with that? Maybe maybe not.
The computer program would not have forgotten to configure the flaps and slats for takeoff.
For those who aren't control engineers... (Score:2)
..."exactly the same" is a bit of a lie
Everything about the environment can change. A good control system adapts well, and may in fact outperform human pilots, but it's never "exactly the same"
Question (Score:2)
Anybody knows why is take-off more "difficult" than landing? Having as a matter of fact, no idea of flying a plane, I'd say it should be the other way around.
FSP – only human input is no-input at all (Score:2)
Embraer is positioning its product
most affordable
most safety appointed
most flight control handsfree
most integrated advanced technology
most augmented against human error
Initially, wanted to attack the crew reduction in the cockpit. But ten year development cycle is a huge sunk cost. But systems have a maintenance vector independent
L-1011 did it already (Score:2)