TWS-Auto

[IronMike]

Fantastic update. The first thing I am going to try with TWS-Auto is intercepting multiple incoming AS-4 Kitchen at 80'000 ft. So far this has been extremely difficult with manual elevation control.

 

Having read the piece multiple times now, I still have troubles understanding the concept of the steering centroid. Let's analyze the following picture:

 

 

Why is the steering centroid to the left and what would be the suggested action?

 

I think I understand the weighting of the tracks. The friendly track has no weight. The unknown track is marked as Do Not Attack and has no weight either. Of the two hostile tracks, the left-hand one is marked as mandatory attack and is about to leave the scan volume limits, therefore it is weighted higher.

 

Now here is what I assume will happen: The scan volume is going to shift left in order to keep the higher weighted track within its limits, until right-hand track falls outside the scan volume. When the right hand track is no longer updated and is dropped, the left hostile remains the only track with any weight and the scan volume will move further left to center it (illumination centroid).

 

But why is the steering centroid offset to the left? No maneuvering is going to prevent the loss of a track when two tracks are more than 40° from each other and no maneuvering is required to keep tracking the left-hand hostile, as it is well within the radar gimbal limits.

 

 

 

 

It is being overthought here a bit. Don't look at where the target is, look at where it will be. the centroid here is imo trying to keep both illuminated as long as it can, but already positioned so that it will keep track of mand attack target above anything else. the target is going to exit the left cone if TWS-A does not adjust, or say you would fly straight in TWS manual without adjusting. Thing is: you are in TWS-A, you do not have to steer left, in fact you can steer right if you like. But as closer both targets get, the more likely the right bandit is going to drop out of the picture. The steering centroid does not equal the point you have to fly to, in order to keep the target illuminated. Nor does the steering T. They just help guide you if you would wish to do so.

[Naquaii]

I am sorry for being an idiot, but I simply not get it. Am I the only one that doesn't understand it?

 

In this situation, steering left will do nothing to help track both targets. One will be lost when they get more than 40° from each other. Steering left will also not help to continue tracking just the left-hand target, as it is nowhere near the radar gimbal limit. The radar will track it just fine for quite some time to come without any steering being necessary. What is the useful information the X is transporting here?

 

 

The steering cross doesn't mean that it will loose the tracks if you don't follow it. You have to look at where the targets with launch prios are heading. While one is heading more or less towards the radar the otherone is going to the left. This means that the WCS will suggest that the pilot should steer left as the targets will eventually exit left. Additionally it also means that the optimum launch attitude will also be slightly left as pointing your missiles towards where the targets will be at impact will improve your chances and with the targets going left that point will also be left.

 

As with the launch cues for missiles in all aircraft not following the cues will not mean you'll miss but you'll be less likely to hit.

[Dino Might]

What I am getting from this is that the steering cross is often where you should point, but in some cases will not be ideal, specifically because if you have two priority targets diverging from your steering axis, the weighted sum behavior won't decide it should drop one of the targets before it is actually out of the scan window and out of hold track memory.

 

 

These situations are probably very rare, and the pilot should already be doing something proactive in this situation anyways.

[Pally]

Dear Heatblur team:

 

Thank you for this update. Now, please redirect me if I misunderstood, that per topic post,

"[...] the computer takes over. The scan volume pattern (2-bar vs 4-bar) is re-evaluated every 4 seconds, while the centroids are re-evaluated multiple times per second. [. . .] The scan volume algorithm considers the future positions of all targets, and selects between which of the two options would give a greater total illumination weight. If they are equal, 4-bar ±20° is selected."

 

Does that mean the computer will switch from 4bar x 20 to 2bar x 40, and vice versa, if that deem more favorable to keep track of all targets of interest due to their (weightedness) within the tracking cone? Also, once the AIM-54 is launched, will it further influenced that behavior? E.g.: If two targets on the left of the TWS tracking cone limits were receiving their respective in-flight AIM-54s, would they weight more then a merely tracked target off to the right even when it is targeted with a "next launch" numeric identifier?

 

If that is the case, then it will be some smart algorithm as it will reduce a lot of a RIO's work load after Phoenix launches in a target rich environment.

Edited April 8, 2020 by Pally
Sentence structure for clarity.

[Naquaii]

Dear Heatblur team:

 

Thank you for this update. Now, please redirect me if I misunderstood, that per topic post,

 

Does that mean the computer will switch from 4bar x 20 to 2bar x 40, and vice versa, if that deem more favorable to keep track of all targets of interest due to their (weightedness) within the tracking cone? Also, once the AIM-54 is launched, will it further influenced that behavior? E.g.: If two targets on the left of the TWS tracking cone limits were receiving their respective in-flight AIM-54s, would they weight more then a merely tracked target off to the right even when it is targeted with a "next launch" numeric identifier?

 

If that is the case, then it will be some smart algorithm as it will reduce a lot of a RIO's work load after Phoenix launches in a target rich environment.

 

That's correct, the WCS will control azimuth, elevation and if to use 4 bar / 20 degrees or 2 bar / 40 degrees.

And as for launched missiles, it will always prioritize targets engaged by actual missiles in flight above tracks that have a prio as loosing those targets will likely scrap those missiles.

 

Wow, where did you get data as exact as this? From a former F-14 radar-/weapons-systems-engineer? :) Impressive!

Also, since when are you in Heatblur team? :) awesome! (can't find any data about the Heatblur team, not even on www.Heatblur.se)

 

Gyro is the main coder for the weapon systems and has been with us from the start more or less.

The data for the displays is from our cache of systems manuals for the F-14.

[Pally]

AWESOME!

 

Thank you for your answer! Glad to know the TWS-A will be as awesome as we imagined. Beautiful work, indeed.

 

Thank you, again, Hearblur.:thumbup:

[TLTeo]

 

 

Just posted by Jabbers!

[draconus]

One important question - are there any roll limits or roll rate limits to keep TWS-A tracking?

[Schmidtfire]

Can Jester operate the TWS-Auto?

 

Feels like Jester AI drop locks quite often and communicates poorly during BVR engagements.

Maybe this feature can make him a little more reliable :)

[SkateZilla]

//Thread Cleaned,

 

In the Future gentlemen we do not need to:

A. Question the Value and/or the reasoning for a member's participation in closed testing.

B. Have Public Disputes about administrative actions whether they are for ED Forums or another location.

[TLTeo]

Can Jester operate the TWS-Auto?

 

Feels like Jester AI drop locks quite often and communicates poorly during BVR engagements.

Maybe this feature can make him a little more reliable :)

 

 

Once you enter TWS-Auto by definition whoever's in the backseat doesn't control the radar, the computer does, so whether it's Jester or a human RIO doesn't matter.

 

 

HB have also repeated multiple times that before putting effort into improving Jester, they want to see how TWS-Auto works out.

[IronMike]

One important question - are there any roll limits or roll rate limits to keep TWS-A tracking?

 

 

The same as in TWS-M, as long as you do not break your radar gimbals, it will track and ajdust no matter how much you roll. During testing I tried to be as extreme as possible and was quite impressed how reliable it is. Ofc, if you lose gimbals, you lose the target - however another very benefitial treat: the track gets held for up to 2 min, which means going fully defensive will leave the track file and recommits become less guess work especially in combination with ground stab. Now, ofc the track file can be wrong or dated, but still helpful.

[IronMike]

On a sidenote: we included Jabbers' video into the original post. At which point I would just like to say thank you to all our testers and community content creators and media partners. They do not do this, to get access sooner, we do not do this to sell more copies, or because we favor some and some not. We do this, because we need to and, luckily, there are people in this community who bring both experience and the willingness to voluntarily help us improve our product. Which we all do together: for you.

 

I think this is common sense for most of us, but in case some forgot, I just wanted to kindly remind us all of it. Testers put an incredible amount of time into helping iron out bugs, get a first feedback and support us also morally, they cheer and fear together with us, which is really a massive additional backup. And our entire team is very greatful for that.

 

On top of that, community creators allow us to provide you a first look video, when we ourselves have no time to make it. And with Jabbers - among others - we have an excellent relationship and we're extremely happy of his support.

 

Thank you all for your kind consideration.

[Scotch75]

^^^^

 

Well said IronMike. Thankyou to all of the testers for helping to improve an already remarkable module.

Cheers!

 

 

Sent from my SM-G950F using Tapatalk

[AG-51_Sabot]

//Thread Cleaned,

 

Thank you. For a moment there I thought I was watching MSNBC

[near_blind]

The same as in TWS-M, as long as you do not break your radar gimbals, it will track and ajdust no matter how much you roll. During testing I tried to be as extreme as possible and was quite impressed how reliable it is. Ofc, if you lose gimbals, you lose the target - however another very benefitial treat: the track gets held for up to 2 min, which means going fully defensive will leave the track file and recommits become less guess work especially in combination with ground stab. Now, ofc the track file can be wrong or dated, but still helpful.

 

Are tracks going to be more resilient with this refactor? I'm not sure I've ever seen the AWG-9 in DCS recover a track once it's gone into extrapolation. Usually if a target is lost, even if it's for a single frame, the computer will continue to extrapolate the current track along the last observed path while creating a new track when the target is found again, often within a mile or two of the new inaccurate old track.

 

Really plays hell on the shoot order :D

[sLYFa]

Out of curiosity, will the steering T be available only with master arm on or also in the training position?

[Eagle7907]

Great write up guys! I can see why a complete revamp was decided upon. Very complex!

 

Question though. When can we expect it next?

 

 

Sent from my iPhone using Tapatalk Pro

[Neuse]

How will Jester handle this? Will he use it on its own or are there additional menu points in the wheel?

[draconus]

Both - options and abilities.

[Zergburger]

Awesome to see this feature finally being rolled out in a finished capacity.

 

 

Question for the devs: Currently in TWS-M it is easy to lose tracks while cranking, even if the RIO keeps the scan volume centered on a track. I have not done specific testing for causality but i notice that it seems to lose the track easier if the pilot cranks faster, and it seems that once you get past about 40 degrees off boresight the tracks drop regardless of how gentle the crank is.

 

 

Is this a bug or some aircraft limitation I am failing to comprehend?

Will TWS-A have this same limitation?

Whether its a bug or WAI, can someone explain why this is happening?

Edited April 14, 2020 by Zergburger

[draconus]

Zergburger, if you mean Jester - he does not move the radar cone unless you command him to. Currently you can crank to near 40 and then command Jester to move scan azimuth. Only then crank more while having radar azimuth moved to the side.

Edited April 14, 2020 by draconus

[MobiSev]

is this coming in the next update?

[Xenovia]

is this coming in the next update?

 

I really hope so

[gyrovague]

Awesome to see this feature finally being rolled out in a finished capacity.

 

 

Question for the devs: Currently in TWS-M it is easy to lose tracks while cranking, even if the RIO keeps the scan volume centered on a track. I have not done specific testing for causality but i notice that it seems to lose the track easier if the pilot cranks faster, and it seems that once you get past about 40 degrees off boresight the tracks drop regardless of how gentle the crank is.

 

 

Is this a bug or some aircraft limitation I am failing to comprehend?

Will TWS-A have this same limitation?

Whether its a bug or WAI, can someone explain why this is happening?

 

I think that is one of the issues that was fixed while implementing TWS-A, so it should be better for TWS-M too. Let us know how it goes!