The download hangar is currently disabled. We're doing our best to bring it back as soon as possible.

Sinking AI at RJOI

Since Matrix support is now terminated, this forum is now locked and will soon be archived.
User avatar
clickclickdoh
MAIW Developer
MAIW Developer
Posts: 1568
Joined: 03 Mar 2009, 03:04

Re: Sinking AI at RJOI

Post by clickclickdoh »

Firebird wrote: 18 Jun 2019, 15:47If this is indeed the case I stand enlightened.
I can't really prove it, but in general it tends to prevent the AI approach error where aircraft will do a lazy missed approach, pitch up then climb away for ever. Runway patter altitude appears to be the other factor in that particular issue. I say I can't really prove it, because obviously sometimes the runway works fine without them in place. Of course, most of the runways that I'm doing this on are all of the 1 to 4 foot wide variety, so the Start Location may be fixing a slightly misaligned runway line that was causing the AI to miss the runway.
User avatar
TimC340
Lieutenant Colonel
Lieutenant Colonel
Posts: 1321
Joined: 07 Mar 2015, 13:18
Version: P3D
Location: Hadleigh, Suffolk
Contact:

Re: Sinking AI at RJOI

Post by TimC340 »

CCD, thanks for that. Yes, I've modded the elevation for both and it's now fine. Adding the starts certainly seems to have improved the reliability of landing traffic, but without having yet found the exact traffic files (been busy doing not much today, but I'll get to it) I can't see what the traffic was intending to do. I haven't seen any problems with the helo/VTOL traffic taxiing in or out, though - unlike on the main airfield, where traffic is holding at ridiculous distances from their conflicting traffic! I've added a few nodes to see if I can free it up, but that doesn't seem to have worked.
User avatar
clickclickdoh
MAIW Developer
MAIW Developer
Posts: 1568
Joined: 03 Mar 2009, 03:04

Re: Sinking AI at RJOI

Post by clickclickdoh »

Which aircraft appear to be getting stuck? I can take a look at it on my system and see if I can generate the same result
User avatar
Firebird
MAIW Admin
MAIW Admin
Posts: 12132
Joined: 11 Aug 2006, 21:04
Version: FS9
Location: EGLL

Re: Sinking AI at RJOI

Post by Firebird »

I maybe getting confused again with Atsugi, but I noticed today that traffic taxying out would hold well away from the hold short lines because of a P-3 I was following was on finals. They three aircraft all held on taxiway F heading for runway 01.

The reason is the same that I have seen at other places , KDFW and KLAS to name two, where aircraft taxying towards an active runway will stop if an aircraft is landing on it. Sure enough once the P-3 I was watching clears the active the 3 aircraft start taxying again.

Is this similar to what you mean?
Steve
_______________________________________________________
Image
Quid Si Coelum Ruat
_______________________________________________________
User avatar
TimC340
Lieutenant Colonel
Lieutenant Colonel
Posts: 1321
Joined: 07 Mar 2015, 13:18
Version: P3D
Location: Hadleigh, Suffolk
Contact:

Re: Sinking AI at RJOI

Post by TimC340 »

Hi guys. No, not exactly Steve. I have seen that behaviour, but not much of it (bear in mind we're only talking about a couple of hours out of the last couple of days!).

What I'm seeing is mainly C130s coming off the runway at B3 and holding indefinitely - or at least until the sim removes them - wherever any other traffic may be. It seems to happen whether the other traffic is on the approach, or anything else is moving on the ground. From B3, I think they're intending to route B, E to their ramp. If they pass B3 on the landing roll, there doesn't seem to be a problem. They come off at F and route all the way down F to the ramp. It does get a bit more complicated in that traffic taxiing out to 02 will hold on B short of B8, waiting for the traffic on B3 to move. Of course, then nothing moves and you have to wait for timeouts.
User avatar
clickclickdoh
MAIW Developer
MAIW Developer
Posts: 1568
Joined: 03 Mar 2009, 03:04

Re: Sinking AI at RJOI

Post by clickclickdoh »

The B3 thing should be a simple fix. Here's what the point looks like in ADE:

Image

The problem is caused by the use of taxiways in early versions to create the taxiway shoulder. In later versions I replaced those with polygons for the shoulders. You can see one of those taxiway links as a shoulder right below the hold short node. That's what is causing the problem. What is happening is that once the AI is cleared from the hold short point, it's trying to get to the taxi point that is closest to it. Unfortunately, it's picking the shoulder point since it's closer than the actual taxiway point. Since the AI can't get to that point, it just sits there for ever.

Thankfully the fix is easy. In fact, there are two ways to fix it. The first way is to replace the shoulder taxi link with a polygon. The second, and easier way is to simply place a node a couple of meters up the taxiway from the hold short.
User avatar
TimC340
Lieutenant Colonel
Lieutenant Colonel
Posts: 1321
Joined: 07 Mar 2015, 13:18
Version: P3D
Location: Hadleigh, Suffolk
Contact:

Re: Sinking AI at RJOI

Post by TimC340 »

Ah, of course. I did try adding a few nodes in the area, but probably not close enough. As it happens, I got into a conversation about Iwakuni on the forum last night with someone else who’s using the same combination of P3Dv4 and Orbx Global Vector, and he put up a screenshot which prompted me to look a bit more into the place. I’m going to have a play with it today, starting from the the stock airport, to see if I can come up with something that better fits that combination. After I’ve done that, I’m going to stop sightseeing in parts of the virtual world where I’ve no business - it causes too much work!

Ah, I see you’ve found that conversation!
Locked