AI Traffic lands off the runway

[martinlest2 58]

I have noticed that at some airports (e.g. SEGU) some (but not all - weird) AI traffic lands some way before or after the actual runway: at SEGU (segu_4.zip, flightsim.com) I just saw one a/c descend the length of the runway and land in the fields outside the airport. Obviously I have checked for duplicate AFCADs - there aren't any (according to ScanAFD, AFCAD2, my folder searches): just the 3rd. party airport AFCAD and the default FS AP******.bgl file. At SEGU the start points for both of these are almost the same - no reason I can see for some a/c to land past the runway. At one or two other airports I saw traffic land in the sea, short of the airport. Again. Not all ...Could this be something in an AI aircraft.cfg file?? I can't see how this is happening, so can't begin to see how to fix it either. Anyone any ideas?Martin

[Tim_Sims 0]

Yes I've noticed this a fair few in various places. Two things that seem to be associated with it:- very busy airports, with a high volume of inbound traffic that's created when starting FS9. So for instance I've seen this quite often at EGLL and EHAM. Not sure that SEGU would fit into that category though.- using flyable models for AI purposes. Not a clever thing to do anyway, what with the impact on frame rates etc. Sorting out the second of these in particular seems to have solved the problem for me. No idea why though.

[martinlest2 58]

Odd. I don't think I have any flyable a/c set as AI traffic - short of opening every .air file I am not sure how I would check: I will note, next time I see this, which a/c has landed off-track and see if I can locate the file.Martin

[N60JBravo 27]

This is not uncommon when you first start the sim.All AI will load at cruise altitude and some have problems getting down to an altitude where they can successfully land.They will "dive for the deck" and not be able to stop their descent. They will then land short of the airport and taxi in the rest of the way.It can take up to 15 minutes for AI to calm down.regards,Joe

[ja99 2]

You fellas are spot on i don`t think there`s much you can do!JA99 FAPS

[martinlest2 58]

Oh well, I'll just look on when it happens and smile then... Thanks.Martin

[speedbird2001 0]

When I am writing my own Flightplans I have sometimes created the odd rogue flightplan that has been compiled incorrectly and have found aircraft nose diving or landing way short of the designated runway. If a trip from Airport A to Airport B that should take two hours, has been programmed to take-off at say 9:00 and arrive at 10:45, the aircraft does begin to behave strangely i.e nose diving or landing short of the runway.

[kiwiflyer45 123]

I remember a thread about a year ago that went on for several pages. A bunch of us were trying to solve this issue and it really seemed to come down to so many variables that there just doesn't seem to be any simple solution. There are definitely some flight models that don't work well as AI. There does seem to be a corellation between traffic density and it sometimes seems worse if the airport is near a body of water. (Go figure!) Flight plans wrongly compiled can cause problems. But when we tried to put it all together, there just wasn't any single pattern...Maybe there's a genius out there who could develop a utility that would check all the variables - but I won't hold my breath! LOLYou can probably tweak it for a particular situation but I doubt you will solve the problem completely. But if I'm wrong, I'd love to hear the solution!!!!!Good luck mate

[speedbird2001 0]

I remember a thread about a year ago that went on for several pages. A bunch of us were trying to solve this issue and it really seemed to come down to so many variables that there just doesn't seem to be any simple solution. There are definitely some flight models that don't work well as AI. There does seem to be a corellation between traffic density and it sometimes seems worse if the airport is near a body of water. (Go figure!) Flight plans wrongly compiled can cause problems. But when we tried to put it all together, there just wasn't any single pattern...Maybe there's a genius out there who could develop a utility that would check all the variables - but I won't hold my breath! LOLYou can probably tweak it for a particular situation but I doubt you will solve the problem completely. But if I'm wrong, I'd love to hear the solution!!!!!Good luck mate

And as A.C Clarke once said..'The truth will be stranger'

[jvile 1]

I have noticed that at some airports (e.g. SEGU) some (but not all - weird) AI traffic lands some way before or after the actual runway: at SEGU (segu_4.zip, flightsim.com) I just saw one a/c descend the length of the runway and land in the fields outside the airport. Obviously I have checked for duplicate AFCADs - there aren't any (according to ScanAFD, AFCAD2, my folder searches): just the 3rd. party airport AFCAD and the default FS AP******.bgl file. At SEGU the start points for both of these are almost the same - no reason I can see for some a/c to land past the runway. At one or two other airports I saw traffic land in the sea, short of the airport. Again. Not all ...Could this be something in an AI aircraft.cfg file?? I can't see how this is happening, so can't begin to see how to fix it either. Anyone any ideas?Martin

MartinThere is one aspect of the AI Plane behavior that is not very well known. The problem starts with the Flight Dynamic Developer who does not follow any certain type rule when reworking the .air and .cfg file of the AI Plane.Even after the simulator settles down (as previosly stated in this thread) you will see some AI Planes continue to have problems landing. Since you say in your post

I just saw one a/c descend the length of the runway and land in the fields outside the airport.

I will assume that was Runway 3 since a field and water is at the end of it. The question now becomes, what is different about Runway 3 vs. Runway 21. The difference is Runway 3 has no ILS approach code or for that matter any approaches and Runway 21 does have a ILS approach code. Many AI Planes have tweaked FD's that are very poorly done. The one thing that covers this problem is the AI Plane gets help with its FD's because it is flying a ILS approach coded runway.Take that same AI Plane that has poor FD's and no ILS approach code in the database for a runway and the AI planes approach instructions (coded flight behavior) only has a hard .dll code that is a standard for every single airport /runway in the world.In other words when a ILS approach exist for the runway that the AI Plane is approaching, then its flying instruction behavior is assisted by the XML coded approach. This helps tremendously in getting the poorly FD's to fly the approach and land with better visual results.Take that same AI plane on a clear day to a airport like TNCM that does not have a ILS approach code to Runway 10 and once again you will see the plane in many cases not land correctly or not land at all. Many AI Plane developers use a large farvorite airport to visually tweak their flying characteristics but the runway they test on has a ILS approach code. Again, that can cover up poorly designed AI Planes Flight Dynamics. Some AI Planes will actually fall to the ground when the gear comes down. This happens when the sink rate at geardown can't keep the AI Plane on a hardcoded .dll approach when the ILS approach does not exist.The Airport Design Utility (ADE9) will check all the airport variables and let you see what approaches do exist at airports. This is a start in helping to understand why some runway ends don't work as well as others and why many airports close to water do not have a Lead-in ILS approach for AI Plane assistance.hope this is helpfuljim

[martinlest2 58]

Thanks for the interesting replies. I am not sure now about the past few times I have seen this - what the variables were. As I say, I think that the best thing is for me to note down as much information as I can the next time I see this happens and see if I can locate something that may account for it. I am reasonably adept at decompiling stuff and looking for coding problems (flightplans, bgl files etc.) so I will see what I can see (and post back of course if there is any thing noteworthy to report).Martin