Board index FlightGear Development AI Traffic

AI Landing Way before the Threshhold

Intelligent, computer controlled vehicles that drive/fly over the planet!

AI Landing Way before the Threshhold

Postby jtprophet » Thu Dec 02, 2021 10:50 pm

Good evening,

My reference airport is KCLT, upon arrival to 36L airplanes are landing in the trees and disappearing under ground and then finally appearing on the runway as if they did a perfect landing. Has anyone seen this behavior before with the AI traffic; not sure if only this runway is the one that they exhibit this behavior on because I'm still trying to determine exact specifics to get my ai to land on parallel runways as well for example 36R; normal operations use 36C for departures.

Thank-you,

John.
jtprophet
 
Posts: 205
Joined: Mon Apr 04, 2016 3:21 pm
Callsign: DAL1464

Re: AI Landing Way before the Threshhold

Postby BecOzIcan » Fri Dec 03, 2021 7:56 am

Hey John,

Testing with recent nightly build and terrasync I got proper traffic separation with departures from 18R, arrivals on 18C + 18L and touching down on runway (not in forest)

Image

Problems in touch down position are normally APT.DAT / ILS XML or THR XML related but the Terrasync data seems clean from what I can see,
any custom scenery override running on your machine ?

Cheers
Ian
Current Projects: AI Traffic, Models & Liveries
User avatar
BecOzIcan
 
Posts: 1302
Joined: Tue Oct 04, 2011 11:43 pm
Location: Sydney, NSW, Australia
Version: 2020.4.0
OS: Win10

Re: AI Landing Way before the Threshhold

Postby wkitty42 » Fri Dec 03, 2021 2:05 pm

jtprophet wrote in Thu Dec 02, 2021 10:50 pm:My reference airport is KCLT, upon arrival to 36L airplanes are landing in the trees and disappearing under ground and then finally appearing on the runway as if they did a perfect landing. Has anyone seen this behavior before with the AI traffic;

yes, i've been seeing in on next for a little while... not exactly sure when it started, though... i've seen it at KRDU... i've also seen it at other airports, recently, but brief testing with this morning's next build isn't showing the problem... i've spent 5 to 30 minutes each and checked EGLL, EGKK, EIDW, EGCC, KBOS, KRDU, KCLT so far... at KCLT, especially, the craft were landing on 18C and 18L with no problems...

one thing to note, though, is that by the time i got to KCLT, my sim (flying UFO) had been running with AI on for about 2 hours... i started at EGLL and let everything settle down for about five minutes... i watched a few AI landings and then i moved/flew over to EGKK and watched some there... i had seen a couple of flights headed to EIDW so i flew over there real quick and watched them land.. then back over to EGCC and to KBOS, KRDU, and now KCLT... none of those landings had problems... i went back to KRDU where craft were now landing on 23R and they were dropping into the buildings and below the terrain on approach... going roughly by the UFO altitude, the threshold for 05R is about 380ASL... i watched American 1862 KBNA-KRDU on the in-sim map with the Data enabled... it went all the way down to 0ASL and was onto the runway for a bit before it suddenly popped up showing 350ASL... since the altitude numbers in that display jump by +/-50, this is a reasonable number... then i watched Delta 837 KSLC->KRDU landing on 05L and it did the same thing...

so what i'm getting at is the calculation of the flightplan for AI craft... there may be some problem getting the terrain elevation for an area when the AI flightplan is calculated... early flights already in progress may have FPs that take them all the way to 0ASL but FPs calculated after the sim has been running for a while may be correct since the terrain elevation may now be available... i don't recall exactly when the FP is calculated, though... i don't know when these craft departed their origin airport and where i was in the sim at that time... i'm not saying this is the problem but it is a possibility if there's some bug in the AI flightplan code when it calculates the landing leg of the flight... according to the data i can still see on Delta 837, it left KSLC on 2021 Nov 27 at 23:55Zulu and arrived at KRDU on 2021 Dec 3 at 12:57Zulu... so yeah, there's a bug in there somewhere... probably a few of them ;)
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 9146
Joined: Fri Feb 20, 2015 4:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 20.04

Re: AI Landing Way before the Threshhold

Postby portreekid » Fri Dec 03, 2021 2:15 pm

There's more than enough bugs, yes :-D The FP is actually split into 9 Legs and each leg is calculated when reached. AFAIR the altitude used for Leg 6 (Landing) is the one from airport center. Leg 5 might use a different one. After spending at least a week hunting for an ATC bug blocking runways I'm working on Leg 9 (Parking) to align it better. What I found out was that if you start on the runway that will always be blocked for AI traffic. So always use start at parking if watching AI.
portreekid
 
Posts: 651
Joined: Tue Jan 14, 2014 4:36 pm
Location: Leipzig
Callsign: PORTREE
Version: 2020.2.1
OS: Windows 10

Re: AI Landing Way before the Threshhold

Postby wkitty42 » Fri Dec 03, 2021 2:43 pm

i hear ya!

when i'm in the UFO and watching AI, i generally spawn in hovering a few hundred feet above the top of the control tower... if i transition to another airport, i head to its tower location, too...
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 9146
Joined: Fri Feb 20, 2015 4:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 20.04

Re: AI Landing Way before the Threshhold

Postby jtprophet » Fri Dec 03, 2021 6:06 pm

I see so it could be my specified scenery folder that I use specifically for CLT; this also includes the KCLTgroundnet.xml file that doesnt pass all the internal checks of the airport taxiway infrastrastructure within the "kids" ; ) wonderful program; I'll try it without the scenery folder to see if it yields better results, thank-you.

John.
jtprophet
 
Posts: 205
Joined: Mon Apr 04, 2016 3:21 pm
Callsign: DAL1464

Re: AI Landing Way before the Threshhold

Postby jtprophet » Fri Dec 03, 2021 6:15 pm

While we are on the subject can someone analyze my runway use file and see if my content is correct; what I am seeing is if the wind is reporting to be 200@ 10 they still insist on using 36L only to land.

<?xml version="1.0"?>
<rwyuse> <!-- Times are in UTC -->
<com> <!-- Commercial Traffic-->
<wind tail="7"
cross="20"/>
<time start = "00:00"
end = "24:00"
schedule = "day"/>
</com>
<gen> <!-- General Aviation -->
<wind tail="7"
cross="20"/>
<time start="7:00"
end = "19:00"
schedule = "outbound"/>

</gen>
<!-- commercial schedule -->
<schedule name= "day">
<takeoff>36C, 18C, 36R, 18L</takeoff>
<landing>36L, 18R, 18L, 36R</landing>
</schedule>

</rwyuse>
jtprophet
 
Posts: 205
Joined: Mon Apr 04, 2016 3:21 pm
Callsign: DAL1464


Return to AI Traffic

Who is online

Users browsing this forum: No registered users and 3 guests