GSX Pushback tug selection

[VHEBN]

Hi Umberto,

As I understand, GSX pushback tug selection seems to be based on parking radius defined by the AFCAD file.

I am curious as to whether there is a limitation that has forced this to be the case, or if this could instead be based on something like the max_gross_weight in the aircraft.cfg.

At the moment, GSX is perfectly happy to select a tug that is either ridiculously over, or under sized for the aircraft.
https://media.discordapp.net/attachments/490671051311808555/882467033080553473/2021-9-1_13-27-8-929.png?width=810&height=431

Using a different value such as max_gross_weight would solve this issue in almost all cases.

[virtuali]

As I understand, GSX pushback tug selection seems to be based on parking radius defined by the AFCAD file.

I am curious as to whether there is a limitation that has forced this to be the case, or if this could instead be based on something like the max_gross_weight in the aircraft.cfg.

No, it's not. It's based on the airplane weight, exactly as you suggested, which of course is the most realistic option, by default.

HOWEVER, there's an additional constraint that, if the parking radius is too small ( so you shouldn't probably park a big airplane there to begin with ), the largest truck won't be used, because they would have trouble maneuvering in a limited space.

Of course, GSX cannot possibly know if the parking radius defined in the AFCAD is *really* cramped, or it was made like that intentionally (perhaps to affect AI assignment), or it was just an oversight of the airport developer so, by default, GSX trust the AFCAD but, as many things in GSX, the parking radius IN GSX, can be customized, overriding the size read from the AFCAD, which in turn will also affect the Pushback selection.

And of course, the Pushback starting position can also be customized so, even if the parking IS somewhat cramped ( to the small radius wasn't an error after all ), by combining a larger radius in GSX ( overriding the AFCAD ) and a custom start position for the Pushback, even the largest vehicles might be able to maneuver there.

[VHEBN]

HOWEVER, there's an additional constraint that, if the parking radius is too small ( so you shouldn't probably park a big airplane there to begin with ), the largest truck won't be used, because they would have trouble manoeuvring in a limited space.

Let me preface this with the fact that I literally drive pushback tugs for a living.

Now,

This is potentially the stupidest piece of logic that has ever been written into any flight sim addon ever.

Pushback tugs, by design, are generally very manoeuvrable despite their large size and weight. This is thanks to features such as all wheel steering, reverse and side cameras, mirrors and hydraulically lifted cabins. Getting a large tug into a small bay has NEVER been an issue for me - and I've pushed 737s off a cramped terminal built 50 years ago with a Blissfox F1-300 (capable of pushing A330/787).

On the other hand, you would never even think about attempting to push back a widebody aircraft with an undersized tug. Not only would the engine not have enough power to start moving, but the tug itself would be far too light, in terms of weight, to get traction. It's a simply physics, something that weighs only 15,000 kg is going to have a lot of trouble moving something that weighs 300,000 kg.

For example - pushing a 737 aircraft plus payload, say 55,000 kg, with a tug that weighs 9,000 kg, it's easy to loose traction, especially when wet - and I wouldn't recommend it. Up that to a 20,000 kg pushback tug, and the issue is avoided.

Obliviously this is very much simplified, but lets say an A330. 200,000 kg. To get the equivalent of the 9000 kg pushback tug on the 737, our tug needs to be about 30,000 kg. Otherwise the lack of traction makes the pushback not only unsafe, but potentially impossible and it's almost guaranteed that you would damage equipment.

So with all that in mind, I come back to where I started, in saying that this is potentially the stupidest piece of logic that has ever been written into any flight sim addon ever.

[virtuali]

Let me preface this with the fact that I literally drive pushback tugs for a living.

So you are not a programmer, which explains how wrong ( and a tad offensive ) your subsequent comment is:

This is potentially the stupidest piece of logic that has ever been written into any flight sim addon ever.

It obviously isn't, in the context how ground vehicles in FSX/P3D work, something GSX must in some way adapt to, which you clearly doesn't seem to know anything about, yet this didn't stopped you from commenting in such way.

Yes, OF COURSE we obviously know the REAL truck are very maneuverable, have selectable steering modes including all wheel steering and such. We have their user manuals, we have their specs, we know what they are capable of.

Your mistake here, is to assume this could be replicated *precisely* or easily in the sim.

First, ground vehicles in the sim DO NOT support all wheel steering. There's not even anything indicating if a particular vehicle has front or rear traction, let alone all wheel steering. And, the whole concept of ground friction/traction, and the eventual effect of slippery surfaces based on weather is just not there.

Some of this is present, in a very rudimentary form, in the User Airplane simulation but, it's very rough and very inaccurate and, some 3rd party airplane developers even went as far as replacing or meddling with the flight model parameter ( in real time ) when the airplane is on ground, to at least try to add some realism which is just not there in the sim. And the Ground Vehicles simulation is even MORE SIMPLIFIED than the User airplane simulation.

The way GSX moves vehicles is somewhat an hybrid between using the default ground vehicles simulation and parameters and its own simulation, in some cases we use everything default ( for example when a vehicle comes from far away ), in other cases we move the object in a custom way, for example to line-up more precisely with the airplane doors, since the default simulation wouldn't allow that with the precision we require.

This because, using the default methods is BEST FOR FPS, since we don't have to constantly update the objects positions and calculate all their physics in real time, especially when there are many of them on screen moving at the same time. Well, it's not exactly and "fps" issue, because our code runs externally from the sim, so it's not capable to "slow down" the fps. However, to prevent stuttering, we should be able to complete all calculations for all vehicles in less time than the frame time so, for example, if you are running at 60 fps, in order to not see any stuttering, all calculations for objects that move MUST be done in LESS than 1/60th of a second, that is 16.67 ms.

This is usually the way GSX works for almost all its vehicles, EXCEPT PUSHBACK. During pushback there aren't many other things moving around ( the user airplane and the wingwalker, usually ), so we could afford a more customized algorithm, because it's a more complex system that needs a custom kinematics, so we are almost entirely custom at least in the actual pushback stage after the vehicle is positioned but, considering is still a standard Ground Vehicle, we designed the way it moves based on how the standard ground vehicles work, even adding things they don't support, like the distance between the wheel axles, but we still don't have all wheel steering.

Would be nice to have all wheel steering in the future ? Sure it is, of course this will complicate the math and the code quite a bit, especially considering tow truck can be in different steering and driving modes, such as RWD with 2 front wheels steering, AWD with 2 front wheels steering and AWD with 4 wheels steering so, ideally, we should be able to simulate them all.

But right now, we don't have that which means, OUR Pushback doesn't have the same maneuverability in the sim compared to its real life counterpart and, if we just ignored the parking size restriction, it would be more likely to see weird moves that most users would associate to a BUG.

Yes, of course, those would be easily fixed by tweaking the starting position ( which by default is driven by the parking radius, but can be changed ), but most users either don't know that, or don't want to do that, but they'll surely notice a strange or impossible turn, or an 180 turn, and that would look like a WORSE BUG than seeing a smaller vehicle in a smaller parking ( which, again, can be easily fixed by making the parking larger ).

So no, the parking size constraint is NOT "the stupidest piece of logic", quite the opposite, is the CORRECT logic considering how GSX will work.

Obviously, if we'll ever have all wheel steering in a future version, that "logic" can be easily removed, because is not even a "logic", it's a parameter in the SIM.CFG of some of the larger Pushback you can even change yourself:

Example of the file Addon Manager\Simobjects\PBR\FSDT_Pushback_02\SIM.CFG

condition = %(aircraftWeight)g >= 180000 and %(parkingRadius)g > 12

changed to this:

condition = %(aircraftWeight)g >= 180000

would remove the parking size "logic", leaving only the selection based on the aircraft weight

[VHEBN]

So you are not a programmer

Not by trade, no - but I certainly have some experience in the field

which explains how wrong ( and a tad offensive ) your subsequent comment is:

This is potentially the stupidest piece of logic that has ever been written into any flight sim addon ever.

Firstly, I obviously did not intend any offense, it's just that that's the fact of the matter - having an additional constraint that disallows appropriate sized pushback tugs based on the size of the bay is stupid. I know enough about programming to know that
   a) that's really not a programing thing, and
   b) the person most qualified to say that in this conversation is me.

And to be clear, that point about it being stupid logic was in regards only to the

additional constraint that, if the parking radius is too small ( so you shouldn't probably park a big airplane there to begin with ), the largest truck won't be used

Now,

Yes, OF COURSE we obviously know the REAL truck are very maneuverable, have selectable steering modes including all wheel steering and such. We have their user manuals, we have their specs, we know what they are capable of.

Your mistake here, is to assume this could be replicated *precisely* or easily in the sim.

I understand why you wrote it but the 600 word essay on why the GSX steering animation logic is that way it is really wasn't required. I do completely understand that there's limitations with what you can do both with SimObjects and just generally in P3D, however, I don't think anyone was ever asking for an accurate simulation of all available steering modes on all pushback tug models. My point was more that they exist and therefore saying in regards to larger pushback tugs that

they would have trouble maneuvering in a limited space.

is really not valid - even despite the shortcomings of the steering animation.

it would be more likely to see weird moves that most users would associate to a BUG.

I'm not sure
https://cdn.discordapp.com/attachments/304906290399019008/883220406683729970/unknown.png
that it could
https://cdn.discordapp.com/attachments/304906290399019008/883220433057484810/unknown.png
be any more
https://cdn.discordapp.com/attachments/304906290399019008/883220460270125076/unknown.png
likely than
https://cdn.discordapp.com/attachments/304906290399019008/883220660405555220/unknown.png
it already is.
https://cdn.discordapp.com/attachments/304906290399019008/883220486765568000/unknown.png

but they'll surely notice a strange or impossible turn, or an 180 turn, and that would look like a WORSE BUG than seeing a smaller vehicle in a smaller parking

In my opinion, which I admit is an opinion, I think a vehicle doing something strange which as I've shown above happens all the time, is far less of an issue that a pushback tug that's physically incapable of pushing the plane.

Why? Because a vehicle doing a turn is something that happens once, takes less than a second, and most likely you won't even see it happen. A physically incapable pushback being used, not only looks stupid, but it's much more obvious to the user as the pushback tug sits on bay for the entire turn, and then performs a critical role in one of the primary functions of GSX.

But also these impossible turns happen to the tiny tugs as well, not just the big ones that you've ruled out from these bays. The most common impossible turn I see in GSX is the vehicle doing a 180 on the spot. That is no less impossible in a tiny baggage tug than it is in a 50,000kg TMX450-50.

Therefore, I really struggle to see validity in the idea that preventing or reducing impossible turns can be achieved by eliminating larger vehicles.

( which, again, can be easily fixed by making the parking larger ).

but as you say (and yes this is slightly out of context but the point still stands)

most users either don't know that, or don't want to do that

So no, the parking size constraint is NOT "the stupidest piece of logic", quite the opposite, is the CORRECT logic considering how GSX will work.

Yes, it's still stupid because you're still basing the choice of tug off a limitation that doesn't exist while ignoring a limitation that is much more important - physics. While choosing a pushback tug to use, which is something I do more days than not, my choice comes down to 2 things. Firstly, which tug is appropriate for the aircraft type I'm pushing and secondly, which tug do I prefer personally.

If for some reason I looked at what bay the aircraft was on and thought 'hmmm, yes, this bay is quite small, it might be a challenge to fit an appropriate tug here' that would most certainly NOT be a justification to attempt to pushback an aircraft with a tug incapable of doing so.

And also, in case the part about physics didn't make this point already, using a tiny pushback tug on a big plane is impossible. It probably won't move and if it does move, it certainly won't stop, and the tug will likely have some serviceability issues afterwards.

Example of the file Addon Manager\Simobjects\PBR\FSDT_Pushback_02\SIM.CFG

condition = %(aircraftWeight)g >= 180000 and %(parkingRadius)g > 12

changed to this:

condition = %(aircraftWeight)g >= 180000

would remove the parking size "logic", leaving only the selection based on the aircraft weight

On the topic of SIM.CFG files, in speaking to a friend of mine who is perhaps your biggest fan, we discovered that the logic used on the "FSDT_Pushback_Small" uses an OR, not an AND, like the other tugs, which seems inappropriate. In saying that, after reading this post I'm honestly not sure if that's actually an issue or if it's intended.

But anyway, I hope that you can now understand why any link between the pushback tug and the parking radius makes no sense, adds nothing to the user experience, and is completely unrealistic. I really don't see the limited turning animations as a justification for such a big oversight in realism, especially when such issues are so common in GSX regardless.

Hopefully this issue can now be resolved.

[virtuali]

On the topic of SIM.CFG files, in speaking to a friend of mine who is perhaps your biggest fan, we discovered that the logic used on the "FSDT_Pushback_Small" uses an OR, not an AND, like the other tugs, which seems inappropriate. In saying that, after reading this post I'm honestly not sure if that's actually an issue or if it's intended.

The Pushback Small logic use an OR, and that's exactly what is supposed to be like, to have at least that vehicle always available, considering the constraints in the OTHER vehicles.

If it wasn't like that, considering the other constraints, if you were in an heavy airplane on a small parking, no pushback would come out at all.

But anyway, I hope that you can now understand why any link between the pushback tug and the parking radius makes no sense, adds nothing to the user experience, and is completely unrealistic.

That's just your own opinion, which I don't agree to.

I really don't see the limited turning animations as a justification for such a big oversight in realism, especially when such issues are so common in GSX regardless.

Nowhere I ever said such thing cannot happen, but they would be even WORSE if we didn't restrict the large tow trucks in the smaller parking spots.

Hopefully this issue can now be resolved.

There's nothing to be "resolved", and I thought my explanation was abundantly clear about the reason why it was intentionally made like this, in addition to indicating how EASY if for users to change it to their liking using one of these TWO methods:

- Acting on the parking radius in the GSX scenery customization page. This would be the more correct approach, because the reasons why a parking radius are set are not the same for each scenery and for each parking on the same scenery.

OR

- Acting on the vehicle constraints in the tow truck SIM.CFG. This for those who don't agree to the very idea of restricting tow trucks depending on parking size.

And note that, the current constraint threshold for the parking size is set to 12 mt of radius. Considering half wingspan of a 737 is 17 mt., the real issue here is that your are trying to park larger planes in a place with a radius 5 mt. smaller than a 737 half-wingspan so, the problem is ( assuming the scenery DOES have enough physical space to park there ), the parking radius in the AFCAD has been set way too small for some reason.

[VHEBN]

Sticking your head in the sand and saying no issues here, nothing needs to be fixed is really not the solution here.

If it wasn't like that, considering the other constraints, if you were in an heavy airplane on a small parking, no pushback would come out at all.

But had you not created the issue in the first place, this solution would not be required, thus avoiding this specific issue.

That's just your own opinion, which I don't agree to.

It's not an opinion, it's a fact. The only opinion you don't seem to agree to here is whether GSX should realistically simulate ground services or not.

Nowhere I ever said such thing cannot happen, but they would be even WORSE if we didn't restrict the large tow trucks in the smaller parking spots.

I don't really see how they could be worse though. The smaller vehicles are already performing strange moves anyway - using a realistic model of tug would not change this.

And note that, the current constraint threshold for the parking size is set to 12 mt of radius. Considering half wingspan of a 737 is 17 mt., the real issue here is that your are trying to park larger planes in a place with a radius 5 mt. smaller than a 737 half-wingspan so, the problem is ( assuming the scenery DOES have enough physical space to park there ), the parking radius in the AFCAD has been set way too small for some reason.

And so then given how prevalent this seems to be, wouldn't that be even MORE of a reason not to use it as a factor? If it's completely irrelevant and often incorrectly set then it's not a surprise that using parking radius causes issues. Given how easily demonstrable and common the issues are, it's amazing to me that you don't want to even admit they're issues, let alone fix them.

And so the issue still stands. After speaking with some people I know, literally everyone agreed that the tug should be chosen based on the aircraft type or weight - maximum parking radius has nothing to do with it.

But hopefully this issue can now be resolved.

[virtuali]

Sticking your head in the sand and saying no issues here, nothing needs to be fixed is really not the solution here.

You AGAIN used a derogatory term ( like the first "stupid" ) to make your point. Attacking won't make your point any more effective. You have your opinion, I have mine, I'm sorry if they don't match.

But had you not created the issue in the first place, this solution would not be required, thus avoiding this specific issue.

We haven't created any issue. We tried to have the program reasonably guess if an airplane is using the correct spot, using the data at its disposal, which can be tweaked without changing the scenery.

It's not an opinion, it's a fact. The only opinion you don't seem to agree to here is whether GSX should realistically simulate ground services or not.

Saying "it's a fact" doesn't make it any more factual than it is. It's still your opinion, which I don't agree to. Because, in FACT, it all boils down what thing is seen as the worse bug by the average user. My option is the average user would see bad moving vehicles everywhere as a worse bug, compared to seeing "not appropriate" vehicles ONLY at small parking spots when using a large airplane.

I don't really see how they could be worse though. The smaller vehicles are already performing strange moves anyway - using a realistic model of tug would not change this.

The smaller PUSHBACK vehicles perform reasonably well, except in a very few borderline cases. If we allowed large vehicles there, without any change to their start position, which should be done by the user, because it's too dependent on the actual parking spot, they WILL PERFORM WORSE, this is out of the question.

Since what you call the "problem" CAN be fixed with the parking customization anyway ( even easier, just make the parking spot bigger than 12 meters ), you won't gain anything if we simply removed the constraint in the SIM.CFG ( which, again, you can remove yourself, as I already showed ): you would STILL require tweaking the starting position or the parking radius to let the vehicle move more better.

And so then given how prevalent this seems to be, wouldn't that be even MORE of a reason not to use it as a factor? If it's completely irrelevant and often incorrectly set then it's not a surprise that using parking radius causes issues. Given how easily demonstrable and common the issues are, it's amazing to me that you don't want to even admit they're issues, let alone fix them.

I really don't follow you. I gave you a precise, exact, number. You think it's reasonable to park something like an A330 or 787 in a parking that is too small for a 737 ?

And so the issue still stands. After speaking with some people I know, literally everyone agreed that the tug should be chosen based on the aircraft type or weight - maximum parking radius has nothing to do with it.

As I've said already in my first post, the tug IS normally selected by airplane weight, that's the default. The only time it's based on the parking radius and you would find it inappropriate for the airplane type, in on those very small parking spots less than 12 mt of radius used with an heavy airplane.

But hopefully this issue can now be resolved.

As I've said, several times already, there's no issue to be "resolved" because:

- the "problem" is caused ONLY when you try to park a large airplane in a parking smaller than 12 mt.

- the "problem" can be easily fixed in TWO ways:

1) acting on the parking radius, which is what you should probably do ANYWAY (even if we allowed larger vehicles there), to give the tug better approach paths, and this is of course the more realistic approach, since in real life each parking spot is different.

OR

2) acting on the SIM.CFG of the tug, which will "fix" the non-existing "problem" GLOBALLY, if you don't want to customize the parking size. Of course, doing this, you will have proof of what I said, that larger vehicles will move worse than smaller ones on such small parking ( now free to use for every tug ), so you'll find you WILL have to use method #1 anyway...

[Sholay]

Guys,
Let me tell you this - and this is not a sarcasm, really.
I love this discussion, I learned a lot from it.

Thanks for keeping it to the point, very informative.

&