Here's an updated Q version:

 

http://www.mediafire.com/download/xbxa9rz863om2zq/AISIDSTAR12Q.zip

 

Fixed:  AI departing after landing and turn around at a monitored airport were not being vectored correctly (e.g., AI was found wandering around airport on the ground often in a strange orientation).

 

Revised:  The parameter AIFinalApproachSlowDownScalar replaces the old AIFinalApproachSlowDownRate parameter.  To make the AI slower on approach, set the scalar less than one.  To make AI faster, set the scalar more than one.  This parameter only applies to AI on final approach.

 

Revised:  AICull (intelligent traffic density).  Added new parameter AICullHeight.  Default is 0 (off).  AI above this height (ft MSL) not subject to AICull. Set to avoid culling AI at  enroute flight levels.

 

Thanks,

-Roland 

Hey Roland,

 

Fantastic work!!  I have one question though - how do the AI aircraft controlled by your program respond to the default ATC commands?  So if the aircraft is on a star approach, is the ATC aware of their approach plan?  Yesterday I saw one aircraft being guided in by your program and then ATC told them to turn to some ridiculous heading and next thing you know they disappeared completely.

 

Thanks again.

 

Ilya

Hi Roland,

 

thanks for the update!

 

I am testing currently at EDDM, one thing i realized is that the handover from the IAF (at the time FSX controls the AI) to AISIDSTAR Final Approach-Mode is not very smooth at my setting. Sometimes the AI overflys the IAF at the specified altitude (5000ft, checked in ADE) and starts to descend, while AISIDSTAR-Monitor says Control Status FSX and "enroute". But before reaching the AISIDSTAR Final Approach Point 1700ft, it starts to climb again, still under control of FSX. It overflys the AISIDSTAR 1700ft Waypoint and then, about 1-2nm after this point, AISIDSTAR takes over the AI in Final Approach Mode and routes the AI back this 1-2nm to the first Final Approach Waypoint in the opposite landing Direction! After reaching this point again, the Final Approach Mode works like it should, i guess ;-). Any Ideas? ( I will go on with the tests, and try to describe it better in the next days, if needed)

 

Another point is the Touchdown Smoke effect of the tires. Landing Aircraft in Final Approach Mode touch down, slow down under control of AISIDSTAR and come to a full stop. After this, FSX takes over again and in this Moment, with a not moving Aircraft, the Touchdown Smoke Effect occurs. This looks not very nice. Any Idea how to prevent this?

 

Overall its a great piece of software, and once again Thank You very much!

 

Kind Regards,

Sebastian Beier

HI, i just noticed that aircraft departing KDCA from runway 01 are not following the SID. They continue straight through the restricted area over flying the white house. When I depart from DCA in the real world, you have to make the left turn as soon as practical to pick up the 328 radial. I noticed your custom departure has them follow the river but they continue runway heading. Also is there a limit to the amount of monitored airports?

how do the AI aircraft controlled by your program respond to the default ATC commands?

 

 

Yep, stock FSX ATC will complain when the AI uses realistic approaches, finals and departures. I recommend turning stock FSX voice off and (if you have not already) buying a more advanced ATC program, such as Radar Contact V4, which is compatible with AISIDSTAR.

 

 

Sometimes the AI overflys the IAF at the specified altitude (5000ft, checked in ADE) and starts to descend, while AISIDSTAR-Monitor says Control Status FSX and "enroute"

. . . .

about 1-2nm after this point, AISIDSTAR takes over the AI in Final Approach Mode and routes the AI back this 1-2nm to the first Final Approach Waypoint in the opposite landing Direction!

 

 

I'll take a look at it, were you using the stock STAR files?  Also, what was the landing runway.  It seems like you are using add-on scenery, which one is it?  One quick suggestion is perhaps the STAR exit waypoint is too close to the IAF.  Delete the last waypoint from the relevant STAR files to give the AI a more normal angle to the IAF.  This can also happen when an AI already on a normal, FSX final and a new "autogen" final becomes available for that runway.  The AI will turn around and join the new final, but only for AI that were already on final and above 700 ft (i.e., temporary).  Subsequent AI will approach normally.  You can also write a custom, final approach file, tweak it, and see if that works better for you.  Finally, you can add AIImmediateForceLandAll=1 to the .ini file.  This will cause the AI to disregard the FSX-selected IAF.  This results in the best approach patterns, but unfortunately the AI won't have FSX taxi services available after it lands, so it must be removed from the area or it will clog up the runway.

 

Another point is the Touchdown Smoke effect of the tires. Landing Aircraft in Final Approach Mode touch down, slow down under control of AISIDSTAR and come to a full stop. After this, FSX takes over again and in this Moment, with a not moving Aircraft, the Touchdown Smoke Effect occurs. This looks not very nice. Any Idea how to prevent this?

 

 

Yes, I'm looking into this.  Nothing to report yet however.  In the meantime, you can lengthen out the roll  to perhaps make it harder to see.  Add these to your .ini file:

 

AIRollOutTime = 30

AIRollOutSlowDownRate1 = 1.025

AIRollOutSlowDownRate2 = 1.1

AIRecordRollOutSpeed = 10

 

-Roland

 

 

HI, i just noticed that aircraft departing KDCA from runway 01 are not following the SID. They continue straight through the restricted area over flying the white house. When I depart from DCA in the real world, you have to make the left turn as soon as practical to pick up the 328 radial. I noticed your custom departure has them follow the river but they continue runway heading. Also is there a limit to the amount of monitored airports?

 

Yes, this is a bug.  I increased the height at which the AI is handed over to AISIDSTAR.  Unfortunately when the AI reaches that height, it doesn't have enough time to intercept the SID entry points that are close.  I've fixed the problem, the new version should be available shortly.

 

To users generally, I've also fixed the bug where some AI were landing without the gear extended (cosmetic issue, the AI still landed and rolled out OK) and not all the normal landing lights were on.

 

-Roland 

Hi from here.

 

Now, after about 3 Days or 25 hrs of trying, changing observing for only one Airport i must say, i failed to get the tool running well.

First of all: I´m using P3D v2, so im unsure if some problems might be related to a different simconnect.dll.

In every point i´m taking about the EDDL (Düsseldorf) Add On from Aerosoft.

Further i use UT2, but about 70% of the FDE´s are modified to get a more realistic rollout (maybe 1200 m for Narrowbodys, 2000 m for the fatties)

After trying to let AISIDSTARS run "out of the box" i ran into dozends of issues, not a single approach went well.

There have been 2 main reasons:

The brake settings i did in the FDE now are out of funktion. So the Aircrafts stopped immediatly after 300ft.

You need to know that in EDDL the first exit is almost at the end of the 3000 m Runway.

So the taxi time is exorbitant, no chance for following traffic.

So i tried the Settings you mentioned above, but without a satisfying result.

 

More worse was that the AutoApproach won`t work. Out of the Box the STAR is too short, after passing the last Fix RONAD or BOTNB the Aircraft turns right to the ILS, but way to tight to the runways so that the overshooting couldn´t leveled out.

 

Now i made the descission to modify the STARS according to the Charts. Final Waypoint is DL016 for RW 05 and DL004 for RW 23. Both are exactly in Runwaydirection, distance about 15 nm

What happend now is that the Traffic follow the STAR, but instead of passing the FAF e.g. REGNO in case of RWY  23 they took a hook over a non existent Waypoint 30° from the RW heading and turning back in the same missangle, heading to the middle of the Runway lenght. There, the Aircraft have been droped like a helicopter during a emergency landing with autorotation anywhere in the Grass next to the RWY.

 

So i got rid of Autogenerated Approach and added my own approach files according to the RNAV Approach.

So this is how it looks now:

 

Not using altitudes leads to traffic overfly the RWY in 3000 ft, to the end of the Runway, making a very tight 180° turn and flying back to the FAF.

 

Using altitudes like recommended let the traffic doing a low pass in 500 ft, same end.

 

Adding the altitude for the touchdown more or less 0 ft AGL let the traffic digging through the woods half a mile before arriving the Runway. But at least he´s on the ground

 

So the final solution was an altitude of 185 ft for the touchdown Waypoint.

Usefull to know that the Airport Altitude is 153 ft, so that leads me to the opinion that the field asks for MSL instead of AGL? Worth to check this...

 

Also i need to use the settings:

 

AIRollOutTime = 30

AIRollOutSlowDownRate1 = 0.1

AIRollOutSlowDownRate2 = 0.1

AIRecordRollOutSpeed = 60

 

to stop close enough to the exit to shorten the taxi time to minimum .

You see, 0.1 is already the bottom line. May i request a adjustment of the default setting to get more air under the wings for even harder cases?

 

What is still bad?

 

Well, first of all the assignement of the STARS.

In my case it needs not only "a few minutes" after start until all incoming traffic have a matching STAR.

 

I use OPUS for the weather, my procedure is normally, i start on the default Airport, change the "aircraft" to BOB from ORBx for a better sight, wait until the weather is loaded.

Then i move to EDDL.

Normally the first half hour the assignement looks like given by chance 50:50 wrong/right.

Though the active runway is e.g. 23, at least the half of the NEW injected traffic will assigned to RW 05 STARS.

After 30-45 minutes it looks fine, all traffic assigned to RWY 23.

BUT: Finally i experienced after watching the traffic for another 2 hrs, the assignements moved to RWY 05 again! 100%, without exeption. Active and from P3D assigned Runway was still 23R.

I didn´t understood.

I hope, you can achive a major improvement here.

 

Also still to hard is the Slow Down Scalar while following the STAR.

I´m pretty sure, the STARs are long enough to get a propper separation.

Nevertheless AISIDSTARS tends to slow down the second Aircraft to hard  (220 vs. 144 kts) and forces a traffic jam in minutes because number 3, 4 and 5 aren´t able to get alternative vectors to get more space.

There are more than 30 nm left beginning from the first Waypoint, but this tool tries to archive the seperation urgently.

Maybe it would be better to ignore the missing separation longer: While using the STAR No. 1 keeps the Speed above 220-240 kts, No. 2 to "n" decellerate minimum to 200 kts. While No.1 "escapes" and is finally above the separation distance the former No. 2 speed up to the Speed of 220-240 kts and so on.

Also it looks like "following" traffic will released to late from slowing down. I found traffic fly down the ILS with 64 kts while no aircraft was on Runway or Taxiway. Finally at the Waypoint before touchdown it was released.

 

So overall, i´m highly thrilled about this tool, but it looks like there is still hard work to do to make it charming.

 

Thanks for reading

Hi Roland,

 

i am using the Aerosoft Megaairport EDDM, landing runway was 26 R and L. The STARs i used were self-created with AIConv.exe and fine tuned, so that the last Waypoint before the FSX IAF is about 7nm away from the IAF and gives the AI a good intercept angle. Also im using no AutoFinal Approach Files, the Final Approach Files were self-created. I will give the AIImmediateForceLandAll=1 a try.

 

Hopefully i will have a bit time this week for testing and will report back if something changes!

 

Thank you!

 

Kind regards,

Sebastian Beier

More worse was that the AutoApproach won`t work. Out of the Box the STAR is too short, after passing the last Fix RONAD or BOTNB the Aircraft turns right to the ILS, but way to tight to the runways so that the overshooting couldn´t leveled out.

 

 

I took a look at this.  The program probably recorded a "bad" AI approach for the autogen.  I'm revising the code to better detect AI approaches that should be filtered out and not recorded.

 

 

 

Usefull to know that the Airport Altitude is 153 ft, so that leads me to the opinion that the field asks for MSL instead of AGL? Worth to check this...

 

 

Yes, take a look at the example final approach files for KDCA and KJFK regarding touchdown elevations being non-zero MSL values.

 

 

Also i need to use the settings:

 

AIRollOutTime = 30

AIRollOutSlowDownRate1 = 0.1

AIRollOutSlowDownRate2 = 0.1

AIRecordRollOutSpeed = 60

 

to stop close enough to the exit to shorten the taxi time to minimum .

You see, 0.1 is already the bottom line. May i request a adjustment of the default setting to get more air under the wings for even harder cases?

 

 

I received several complaints about the rollout distance.  I'm redoing the code.  For autogen, I'll detect the final runway waypoint to be the point the AI just starts turning off the runway onto the taxiway.  For custom final approach files, the user specifies the last waypoint.  The code will automatically de-accelerate the AI from touchdown speed to zero at the last waypoint (which is near a taxiway exit).  Thus, the AI will not spend a lot of time in taxi while the rollout will be longer.  For users who like a shorter rollout, I'll add a scalar to increase so that the AI stops well before the last waypoint (and perhaps takes advantage of earlier taxiway exits).

 

 

Finally i experienced after watching the traffic for another 2 hrs, the assignements moved to RWY 05 again! 100%, without exeption. Active and from P3D assigned Runway was still 23R.

 

 

I haven't looked at this yet, but I just detect the landing runways assigned by FSX to incoming AI.  There will be a lag during the changeover.  For example, departing AI were assigned runway 23, but during taxi, the wind shifted.  These AI will still depart runway 23.  However, departing AI just waking up from sleep will be assigned runway 5 to depart.  In addition, some arriving AI will have already been assigned a STAR corresponding to runway 23.  However, newly incoming AI will be assigned runway 5 to land.    This I've seen before.  If the wind shift sticks, then eventually all AI will start using runway 23 to depart and land.

 

 

 

The STARs i used were self-created with AIConv.exe and fine tuned, so that the last Waypoint before the FSX IAF is about 7nm away from the IAF and gives the AI a good intercept angle. Also im using no AutoFinal Approach Files, the Final Approach Files were self-created. I will give the AIImmediateForceLandAll=1 a try.

 

.  

 

 

Sound good, I need to correct something I wrote.  I think AIRecordRollOutSpeed = 10 is too low if anyone is trying that.  Stick with a value of 30 or more.  Regardless, this parameter will be deprecated in the rollout code rewrite I mentioned above.

 

-Roland

Yes, take a look at the example final approach files for KDCA and KJFK regarding touchdown elevations being non-zero MSL values.

 

 

 

Damn, than i missed it in the manual. Well, at least i figured it out by try and error :rolleyes:

 

 

 I'm redoing the code.  For autogen, I'll detect the final runway waypoint to be the point the AI just starts turning off the runway onto the taxiway.  For custom final approach files, the user specifies the last waypoint.  The code will automatically de-accelerate the AI from touchdown speed to zero at the last waypoint (which is near a taxiway exit).  Thus, the AI will not spend a lot of time in taxi while the rollout will be longer.  For users who like a shorter rollout, I'll add a scalar to increase so that the AI stops well before the last waypoint (and perhaps takes advantage of earlier taxiway exits).

 

 

 

Jeez, if this works, the FS-Community will kiss your feet :im Not Worthy:

"Brake to Vacate", will fix one of the worst things regarding to AI-Traffic.

 

I haven't looked at this yet, but I just detect the landing runways assigned by FSX to incoming AI.  There will be a lag during the changeover.  For example, departing AI were assigned runway 23, but during taxi, the wind shifted.  These AI will still depart runway 23.  However, departing AI just waking up from sleep will be assigned runway 5 to depart.  In addition, some arriving AI will have already been assigned a STAR corresponding to runway 23.  However, newly incoming AI will be assigned runway 5 to land.    This I've seen before.  If the wind shift sticks, then eventually all AI will start using runway 23 to depart and land.

 

 

 

Hmm, but is there a possible solution?

In this case i couldn´t see any assignement to RWY 5 at any time ( i used your Monitor-Window on a seperate Display).

Is it possible to improve the readout of the active Runway?

It destroys the whole chain of incoming Traffic when Aircraft follow the entire STAR for RWY 05 and finally  pass the Runway at 3000 ft to get to the FAF of RWY23.

Separation gets lost, some Traffic turns back at the ILS, lots of confusion.

Maybe there´s a better way to check more frequent what is the right STAR and maybe it is usefull to re-assign another STAR when it turns out, that the majority of the traffic still uses e.g. RWY 23.

 

5 Stars would be, if the wind changes, all landing Traffic will send into Holdings until  the last Aircraft finished its "taxi out" from the "old" Runway.

 

Looking forward for v 2.0 :p0311: :lol:

New Rev. 1.2Q (beta) can be found here: 

 

http://www.mediafire.com/download/uh7gf48ndmfnmkv/AISIDSTAR12Q.zip

 

New:  Ground rollout reworked.  For autogen final approaches, AI will complete the rollout about 20 feet short of a runway exit onto the taxiway.  Same for custom final approach files (if the user specifies the last waypoint at the desired runway exit).  Thus, AI will not spend a lot of time in taxiing on the runway.  For users who like a shorter rollout, I've added the scalar AIRollOutSlowDownScalar.  Decrease the scalar (e.g., AIRollOutSlowDownScalar = 0.5) and the rollout will be shorter (perhaps taking advantage of earlier taxiway exits if they exist).

 

Revised:  Autogen final approaches.  Revised code to better prevent abnormal AI approaches.

 

Fixed:  AI missing initial SID waypoints.  Occasionally, AI misses SID entry waypint on climbout and starts slowly circling the departure area.  

 

Important Revision:  A registered version (4.9x or later) of Peter Dowson's FSUIPC utility is recommended to use the full functionality of AISIDSTAR.  For example, the new AICull function requires FSUIPC.  For networked configuration users, Peter's WideFS (ver. 6.9) is also recommended.

 

Please consult the following video documentation (Note: for Rev. M, but still mostly applicable to the latest revision).

 

Video 1 - Working with the Included Sample SID/STAR Files and Creating New Files Using AIConv.exe:  

 

http://www.mediafire.com/watch/mj3hgm33w24thcv/AISIDSTAR_Video_1_Sample_SID_STAR_Files_and_Creating_New_Ones_Using_AIConv.mp4

 

Video 2 - Custom Editing SID/STAR Files - 

 

http://www.mediafire.com/watch/t9irz80c497beo7/AISIDSTAR_Video_2_Custom_Edit_SID_STAR_Files.mp4

 

Video 3 - Operating Parameters.

 

Part 1: http://www.mediafire.com/watch/urlrrdp09v5y5l4/AISIDSTAR_Video_3_Operating_Parameters_Part_1.mp4

Part 2: http://www.mediafire.com/watch/74e3kqyt4sndhwd/AISIDSTAR_Video_3_Operating_Parameters_Part_2.mp4

Part 3: http://www.mediafire.com/watch/24s4ek5x7xtiuf8/AISIDSTAR_Video_3_Operating_Parameters_Part_3.mp4

Part 4: http://www.mediafire.com/watch/24s4ek5x7xtiuf8/AISIDSTAR_Video_3_Operating_Parameters_Part_3.mp4

Part 5: http://www.mediafire.com/watch/24s4ek5x7xtiuf8/AISIDSTAR_Video_3_Operating_Parameters_Part_3.mp4

 

 

Thanks!

-Roland

Mr Roland, I got to say, why am I just finding this program now? I needed something like this for a long time. Just watched your tutorial and right away I created some procedures for mhtg. As you may know, this airport is notorious for its RNAV runway 2 approach due to the airport sitting in a valley. Of course fsx doesn't care and had planes fly straight in through the mountains. I have mangaged to get one of the approaches to work, but on short final the plane still hasn't gotten landing clearance. I may have to abandon the Melvo approach transition and use the TNT transition approach. The TNT transition takes the plane over the TNT vor next to the airport which is in the direction of runway 2. I have yet to try this but I theorize that this approach will fool fsx to give the clearance but since the plane is too high for fsx to control, your program can still vector it around the valley to the runway center line and then land.

Cant wait to try 2Q tonight

Hi Roland,

 

 

you are optimizing your really wonderfull tool very hard, but with every new update, i will have less aircraft get landing. Exactly nil ?! Some do go arounds and the other didn´t reach the final course. A lot of will be deleted, due to " aircraft is over no vector altitude and didn´t check in for two minutes ".

 

I don´t know what my mistake in the ini file is. Could you please be so kind and post yours in this forum !!!

 

And what is the meaning of " problem to open fsuipc link " ??

 

 

thanx in advance

 

klaus

An update of my progress on MHTG. result: failure. The only thing that works is the SID. The STARS are problematic because when the plane finishes it, ATC still doesn't pick it up so the it continues into the ground and disappears.

Just watched your tutorial and right away I created some procedures for mhtg.

 

 

Wow, you're going for the ultimate test!  I think you might need to add AIImmediateForceLandAll = 1 to your .ini file.  Without using this parameter, AISIDSTAR waits until the AI intercepts a FSX-generated IAF waypoint in order to get landing clearance from FSX.  Normally, this isn't a problem (e.g., the KDCA river visual approach begins following the river at approximately the FSX IAF waypoint).  However, for MHTG (Toncontin) it seems from your description the FSX IAF might be inside a mountain.  AIImmediateForceLandAll tells AISIDSTAR to skip the FSX-generated IAF intercept.  The drawback is the AI lands without landing clearance, so it doesn't get taxi service and must be removed after it finishes its rollout.  I don't think it is a big deal personally, you get the most realistic AI patterns on approach with the only drawback being you lose AI taxiing to the gate after landing (AI taxiing to the runway for departure still occurs).

 

If you're having problems, please send you STAR and final approach files to my support address.  I'm interested to see if we can get something working at this airport.

 

 

I have yet to try this but I theorize that this approach will fool fsx to give the clearance but since the plane is too high for fsx to control, your program can still vector it around the valley to the runway center line and then land. 

 

I think you may also need to add AICircleUntilAssignedRunway = 0 to your .ini.  With this parameter enabled (=1 by default) if the AI doesn't get a runway assignment it goes into a circular holding pattern (because sometimes FSX is late assigning a runway).  A circular holding pattern probably doesn't work at this airport though (although you can control the holding altitude).  For the next version, I'm going to automatically disable AICircleUntilAssignedRunway if AIImmediateForceLandAll = 1.  There's no point holding for a runway assignment if the user doesn't want the AI to intercept the FSX-generated IAF.

 

-Roland

An update of my progress on MHTG. result: failure. The only thing that works is the SID. The STARS are problematic because when the plane finishes it, ATC still doesn't pick it up so the it continues into the ground and disappears.

 

You're not using a final approach file, correct?  That means the AI is flying into intercept the FSX IAF (you know...the one inside the mountain) after it exits the STAR.  You are going to need to write a custom approach file.  I don't understand how anyone would realistically expect otherwise at this airport.  They're not that difficult.  I've provided examples for KDCA and KJFK.  If AISIDSTAR can make AI perform the Parkway Visuals 13L/13R at KJFK, I'm fairly confident it can do MHTG.  

Hi Roland,

 

 

you are optimizing your really wonderfull tool very hard, but with every new update, i will have less aircraft get landing. Exactly nil ?! Some do go arounds and the other didn´t reach the final course. A lot of will be deleted, due to " aircraft is over no vector altitude and didn´t check in for two minutes ".

 

I don´t know what my mistake in the ini file is. Could you please be so kind and post yours in this forum !!!

 

And what is the meaning of " problem to open fsuipc link " ??

 

 

thanx in advance

 

klaus

 

I was looking at EDDM and everything seemed to be running ok, but I was just using the stock STARs and autogen finals.  Can you send your STAR and custom final approach files to my support address?  I'll take a look.  Thanks!  

 

The FSUIPC message means the program failed to establish a connection with a registered version of FSUIPC.  AISIDSTAR should still run ok, however it can't perform certain functions like AICull and deleting misbehaving AI.