Airbus studying Advanced TCAS

killerwhale65 · Post by **killerwhale65** » 21 Mar 2006, 18:41

Airbus is currently studying the development of Advanced TCAS (traffic collision avoidance system). In emergency situations, the system would revert controls from the pilots to the autopilot. The aircraft would then automatically fly away from a potential mid-air treat, without any pilot input.

“We are looking at the possibility that when you have a TCAS RA [resolution advisory], you don’t wait for the pilot to take avoiding action, but have the autopilot take control and avoid the traffic,” says Airbus vice-president training and flight operations Capt Jean-Michel Roy.

More info: http://www.flightglobal.com/Articles/20 ... ation.html

SOURCE:Flight International

Knight255 · Post by **Knight255** » 21 Mar 2006, 19:35

I can't wait for this type of TCAS to fly in RVSM airspace and start freaking out at every plane that passes making wild gyrations and rolls to avoid other aircraft. Too much automation!!!!!

vc-10 · Post by **vc-10** » 21 Mar 2006, 20:07

sometimes automation is good. I don't think that we will have planes suddenly shhoting upwards, instead a slightly more gradual system. RE the A320 thing, stopping controlled flight into terrain is a good idea, particularly in Europe/North America, which have frequent snowstorms in Winter.

earthman · Post by **earthman** » 21 Mar 2006, 20:22

This is mainly useful when the plane is flying on auto-pilot, and especially when both pilots are taking a nap (I heard they are allowed to take a nap at the same time, provided that they do it in their seats, and not longer than 45 minutes, or was it 20 minutes)

Avro · Post by **Avro** » 21 Mar 2006, 20:38

Knight255 wrote:I can't wait for this type of TCAS to fly in RVSM airspace and start freaking out at every plane that passes making wild gyrations and rolls to avoid other aircraft.

Why should this happen ?

When you fly RVSM today you only get a TCAS TA (traffic advisory) but no RA (resolution advisory). The planes are not on collision paths so the RA won't be triggered. And as far as the RA isn't triggered the new system won't be activated as well. So no need to worry about RVSM flights.

Chris

Knight255 · Post by **Knight255** » 22 Mar 2006, 00:46

(I heard they are allowed to take a nap at the same time, provided that they do it in their seats, and not longer than 45 minutes, or was it 20 minutes)

Not in the U.S.!!

FAR's prohibit a pilots sleeping while in the flightdeck.

When you fly RVSM today you only get a TCAS TA (traffic advisory) but no RA (resolution advisory). The planes are not on collision paths so the RA won't be triggered. And as far as the RA isn't triggered the new system won't be activated as well.

I take it you've never heard of the Dallas bump?? Due to the complexity of the airspace in the Dallas-Ft. Worth area, some of the departure and arrival corridors are extremely close, so close in fact, that a TCAS will give an RA. The descending aircraft are given a climb to avoid oncoming ascending aircraft that are on SIDs (hence why the name "bump" came about). What would one of these new Airbus's do in this circumstance?? You'd always have the auto-pilot taking over and making a ridiculous climb for no reason!!! Oh, and btw, there have been 0 accidents related to this departure/arrival corridor, so do not question the safety of it.

I'm not sure if there are any more of these anomalies, but I am for sure that FT. Worth Center would not like aircraft suddenly climbing, or doing whatever the computer thinks necessary, to avoid "collision" traffic.

earthman · Post by **earthman** » 22 Mar 2006, 01:06

The original article says that the system would only engage when the auto-pilot is turned on, because currently the pilots need to disengage the auto-pilot when they get a TCAS RA. With this system, they can continue with their crosswords (what else is there to do with the auto-pilot on?).

The idea is “very doable”, says a TCAS expert, who adds: “It makes sense if the autopilot is engaged to let the autopilot do it.” The pilot must now disengage the autopilot on receiving an RA and hand-fly the TCAS climb or descend commands. Roy says it is unlikely the system would take control if the aircraft is being flown manually, as “the autopilot should not engage unless it is done so by the pilot”.

Knight255 · Post by **Knight255** » 22 Mar 2006, 04:46

The idea of the computer making safety decisions and not the PIC is what disturbs me. What if the computer makes a wrong decision and decides to climb into another aircraft while averting another? Who is responsible? I think a CDTI coupled with a TCAS is the way to go. It allows the pilot to make the best and safest decision for the flight with a lot of available knowledge and experience.

earthman · Post by **earthman** » 22 Mar 2006, 13:39

My opinion is that when the auto-pilot is on, the plane should do whatever it 'thinks' is best. If the PIC wants to be the one making the decisions, he should not engage the auto-pilot in the first place! Besides, since when are humans any good at making safety decisions?

It's like the new generation of cruise control on cars, which measures the distance to the guy in front of you, and slows down if you get close. Or the automatic braking system being developed in Sweden (where else), which hits the breaks automatically the moment it determines that there is an obstacle in front of you and you could not avoid hitting it if you didn't start braking right this very moment. Of course this system is mainly for trucks, whose drivers tend to be preoccupied with changing CDs or other such nonsense and end up rear-ending traffic jams.

Knight255 · Post by **Knight255** » 22 Mar 2006, 17:45

Besides, since when are humans any good at making safety decisions?

Since when did computers become ultimately responsible for the safety of flight?! It is the PIC's duty to make decisions that regard the safety of the flight, if a computer makes the decision and makes a mistake, who do you blame!??!!? I hate to break it to you, but computers are not perfect either!!!

earthman · Post by **earthman** » 22 Mar 2006, 18:29

A properly designed computer system will perform better than a human, time after time. The problem then becomes designing a good system, which is not easy, but that doesn't mean it can't be done.

Knight255 · Post by **Knight255** » 23 Mar 2006, 03:31

A properly designed computer system will perform better than a human, time after time.

Name a computer system that can learn and grasp concepts. Unless you've got a computer that has AI, no system will out perform the diverse capabilities of the human brain, which is why the safety of flight should still rest on the PIC's shoulders.

earthman · Post by **earthman** » 23 Mar 2006, 04:24

Knight255 wrote:
A properly designed computer system will perform better than a human, time after time.
Name a computer system that can learn and grasp concepts. Unless you've got a computer that has AI, no system will out perform the diverse capabilities of the human brain, which is why the safety of flight should still rest on the PIC's shoulders.

A system doesn't need to learn and grasp concepts to be able to fly safely. Expect the cockpit to get increasingly crowded with various intelligent computer systems offloading the pilots. Automatic TCAS RA execution while on auto-pilot is a no-brainer really.

dat-tech · Post by **dat-tech** » 23 Mar 2006, 16:38

something to think about...
computers located in the airbus are windows-based.

never had a blue screen???

earthman · Post by **earthman** » 23 Mar 2006, 16:43

And which computers would those be?

TWR · Post by **TWR** » 23 Mar 2006, 17:35

Due to the complexity of the airspace in the Dallas-Ft. Worth area, some of the departure and arrival corridors are extremely close, so close in fact, that a TCAS will give an RA. The descending aircraft are given a climb to avoid oncoming ascending aircraft that are on SIDs (hence why the name "bump" came about).

the same for EBBR...

fokker_f27 · Post by **fokker_f27** » 23 Mar 2006, 17:51

My opinion is that when the auto-pilot is on, the plane should do whatever it 'thinks' is best. If the PIC wants to be the one making the decisions, he should not engage the auto-pilot in the first place! Besides, since when are humans any good at making safety decisions?

Many problems are caused by human errors, that's true. But what scares me is what would happen if the system is malfunctioning.

cherdt · Post by **cherdt** » 23 Mar 2006, 18:16

dat-tech wrote:something to think about...
computers located in the airbus are windows-based.

never had a blue screen???

For the clarity of some readers: The windows PC's are not used to control the airplane!

They are used to replace information on paper by electronic documents.

They are also used in the B-777 I believe.

Avro · Post by **Avro** » 23 Mar 2006, 22:50

Sorry for my late replies but wasn't online a lot the last few days.

Knight255 wrote:
I take it you've never heard of the Dallas bump?? Due to the complexity of the airspace in the Dallas-Ft. Worth area, some of the departure and arrival corridors are extremely close, so close in fact, that a TCAS will give an RA. The descending aircraft are given a climb to avoid oncoming ascending aircraft that are on SIDs (hence why the name "bump" came about). What would one of these new Airbus's do in this circumstance?? You'd always have the auto-pilot taking over and making a ridiculous climb for no reason!!! Oh, and btw, there have been 0 accidents related to this departure/arrival corridor, so do not question the safety of it.

No I've never heard of the Dallas bump before. Sounds as an interesting situation. And don't worry I won't question the safety of the SIDs and STARs.

Now back to the advanced TCAS. For those specific airports where such TCAS RA's would pop up during normal operations due to the other traffic being close and on a collision course, a special feature will have to be implemented and tested. There would be a couple of solutions but I'm sure they'll think of them before installing those advanced TCAS systems.

What I was talking about is in RVSM flights as you mentioned earlier in the topic. And as far as the planes are not on collision course you will only get the TA and no RA !!

The idea of the computer making safety decisions and not the PIC is what disturbs me.

Do you really think the PIC can do much better decisions ?? History has told us that quite some collisions happened due to pilot error. So why not improve it by an automated system. Such systems have already been tested here in Holland (in cooperation with the TUDelft) as far as I know and the results where positive.

Since when did computers become ultimately responsible for the safety of flight?! It is the PIC's duty to make decisions that regard the safety of the flight, if a computer makes the decision and makes a mistake, who do you blame!??!!?

I don't agree with you once more. Why being against automation if it can enhance the safety of the flight ? Pilots are not superman either. So if we develop a computer system which can do the same job as the pilot (for the TCAS) but with a smaller rate of failure, I don't see why this shouldn't be implemented.

As for the windows computer: The day the real nav displays and so work on windows I won't step inside a plane anymore

Chris

Knight255 · Post by **Knight255** » 24 Mar 2006, 02:13

Do you really think the PIC can do much better decisions ?? History has told us that quite some collisions happened due to pilot error. So why not improve it by an automated system. Such systems have already been tested here in Holland (in cooperation with the TUDelft) as far as I know and the results where positive.

Yes I do believe PIC's can make better decisions. Airline flying has not been around for that long, and better yet, accident investigation is getting better in determining the causes of accidents. With proper training, pilots will make better decisions, and history has proven this. CRM work has decreased accident rates significantly. Something to think about, the automated TCAS will only perform what it was programmed to do, a pilot can use decades of experience and training to make a decision.

Mid-air collisions are not all that frequent which is why an automated TCAS seems overkill to me. I think the best path is an advanced CDTI which would give the pilot MUCH more information than just a TA or RA.