FSX-MS AI Controller 1.4 (SID, Jet Route, STAR and Final Approach) (New) Beta

[liketofly1 11]

Anyone noticed how much support the author is giving his product? If Payware authors gave this much support, could you imagine how good addons would be. Thank you for all your hard work and dedication. You should seriously consider a "PayPal" donate button on your post. I would donate.

 

Totally agree..... absolutley outstanding.

[Norbo 0]

Totally agree..... absolutley outstanding.

 

I also agree!!!!

[chumley 0]

 

 

chumley, on 16 Apr 2015 - 10:57 PM, said:

I followed on the same route also taking the GILES SID. The AI aircraft followed the SID and after passing the GILES fix and the end of the SID I expected it to be assigned a Jetroute.
 
I'll take a look, probably something simple like the mode not getting reset after SID exit because it never departs the user's reality bubble.
 
EDIT: are you using enhanced approach mode (just asking for troubleshooting purposes).
 

 

Yes Roland, I have it set to Enhanced Mode (haven't touched that setting since installing).

[Clipper Ocean Spray 189]

Thanks guys.  I get a lot out of the project myself (not just being altruistic)!

 

 

 

Yes Roland, I have it set to Enhanced Mode (haven't touched that setting since installing).

 

Think I found the bug.  Should be in the next hot fix.

 

 

After all the most important is to keep the separarion. It won´t help if a AI that proceeding a GA come back at any WP and end up in a mess of more AI´s.

 

These are good points.  I'm going to go with the standard rectangular go-around pattern (which can be adjusted using the enhanced approach options) in the next hot fix.  It offers space to separate out the AI with the benefit of being easily discerned as a "pattern" (I still get comments about the go-around back to STAR as some sort of error - it isn't easy to recognize what is happening by just looking at it in FSX/P3D).  I was never a big fan of these FSX-like go-around patterns, but I appreciate it more now as a good least-common denominator solution.  I may eventually add support for missed approach procedures from navdata (as I'm sure you're aware from personal experience, AIController has already proven itself capable of executing published missed approach procedures...even when it wasn't supposed too  :blush: ).

 

-Roland

[FlyTweety 2]

  I may eventually add support for missed approach procedures from navdata (as I'm sure you're aware from personal experience, AIController has already proven itself capable of executing published missed approach procedures...even when it wasn't supposed too  :blush: ).

 

 

 

Yes, this is what i wrote one or two pages before. For example the Navdata of EDDK (Cologne). Each approach file already contains two or more waypoints for the missed approach. Therefore i had to edit these 20 or more approaches because every approach was continued at about 1500 ft as a missed approach. I don´t know how it can be possible to recognize this automatically and skip those waypoints despite there´s really a reason to Go Arround?

However, currently i´m not sure if it ever will work to follow "fixed" patterns. For a good reason in the real world the ATC commands "dynamic" routes, shortcuts or roundabout ways for seperation purposes, additional to the speed restrictions.

The enhanced approach is a appreciated concept, maybe it can be expanded and traffic will skip the following STAR WP but will commanded for a direct entry. Or in case exiting of the STAR  it continues heading in a usefull direction for another 5 miles before it catches the ILS. Holdings and Go Arrounds should allways be the last solution.

[Clipper Ocean Spray 189]

A new beta version of AIController v1.14B (full version including Hot Fix 6) is available at the following link:

 

http://www.mediafire.com/download/9gnv1mru9577917/AIController14BFullWithHF6.zip

 

Hot fix 6 by itself is available at the following link:

 

http://www.mediafire.com/download/hlw9oa4rrzz00am/AIController14BHF6Only.zip

 

Please do not mix with any version of 1.4A or earlier.  Instead, install into a separate directory for testing.

 

IMPORTANT:  Hotfix 6 now provides AI cruising speed data from the aircraft.cfg files, so be sure to perform a "Force Update" after installation.

 

Hot Fix 6 New Features:

 

NEW:  Each AI's individual cruising speed during jetroute portion of flight determined from AI's aircraft.cfg data.

 

NEW:  Progress bar added for aircraft.cfg scan during update process.  Over a network connection, the AI aircraft.cfg scan may take longer than the actual scenery decompile process.  This discrepancy may lead the user to believe the "update" process has completed, when actually the aircraft.cfg data is still being written.  Thus, the new progress bar will help keep the user appraised.  Note:  If you had any problems with AI not getting assigned the correct gates in earlier versions, an incomplete aircraft.cfg scan might have been one of the causes.  Please install this hot fix and retry. 

 

NEW:  Specific support for My Traffic 6, including the non-ascii characters used in the ATC-IDs.

 

NEW:  Support for coordinated STAR-transition-approaches added (i.e., preference is given to selecting an approach having the same transition waypoint as the STAR exit waypoint the AI was assigned to). 

 

NEW:  Short-final go-around mode changed to standard left-base go-around pattern to make go-around patterns more coherent, reduce holding, and aid AI spacing. 

 

Hot Fix 6 Fixes:

 

FIXED:  Incorrect parking assignments and double-parking based on duplicate gate numbers at different areas (terminals) of airports.  The fix also modifies AIController and AIMonitor to provide more detailed parking assignment information.  

 

FIXED:  Gate radius data sometimes not processed correctly (AI-gate size mismatch).

 

FIXED:  Only one aircraft.cfg parking code being read (note: scenery gates typically have mulitple parkings codes, which was read by AIController, however the aircraft.cfg itself can also contain multiple parking codes, which is now being read and processed).

 

FIXED:  Parking alignment bug sometimes prevented AI from being designated "Parked at Gate" causing double-parking and preventing departure of same AI after waiting period (if auto-inject option enabled). 

 

FIXED:  Sometimes AI obtaining high/low altitudes after being released from jetroute.  

 

FIXED:  Turbojet Default Lower and Upper Jetroute Altitudes Bad.  Choose "Reset Options to Default" to activate corrected default values - OR - if you do NOT want to reset your customized options to default, go to "Options" - "Jetroutes" and change the "Turbojet Altitude Lower Range" to 18000 (or your own preferred value) and change the "Turbojet Altitude Upper Range" to 22000 (or your own preferred value), then press the "Apply" button.  Note: the bad default altitudes may have also caused-in-part the "AI obtaining high/low altitudes" bug above.

 

FIXED:  Scenery.cfg absolute-path TO UNC-network-path translator in Prepar3d mode isn't working.  Important for users running AIMonitor on a networked client machine because MyTraffic 6 places an absolute path (pointing to a folder shared by different FS programs) in the scenery.cfg.

 

FIXED:  Deletion routine for stalled taxi-out AI (i.e., AI not moving) fixed.  The routine was sometimes not deleting the longest waiting stalled AI (e.g., the stalled AI at the head of the line).

 

FIXED:  Sometimes AI not getting assigned jetroute after SID exit.

 

New for Version 1.4B (Prior Revisions)

 

NEW:  Manually delete AI (or a group of AI(s)) by highlighting them in AIMonitor.  Highlight the AI(s) in the AIMonitor window, right-click, then choose delete.  This is similar to the old AI Traffic Explorer utility.  Good for deleting pesky AI that you can't get to with the FSUIPC zapper, but you can identify on the AIMonitor. 

 

NEW:  Persistent highlighting of selected AIMonitor rows.  This feature was added to support the delete option above, but has actually been requested several times by itself!  I presume the persistent highlighting will allow the user to better track particular aircraft. 

 

NEW:  Specify AI install directories outside of the FSX/P3D path.  In the config menu, advanced options.  Useful if an AI package has been installed to location outside the SimObject path.

 

NEW:  Absolute path to UNC path translator for Scenery.cfg.  In the config menu, advanced options.  Useful for when AIController is running on a networked client.  You can tell AIController the corresponding UNC path for any absolute path (e.g., a path containing a drive letter) appearing in the scenery.cfg.  As always, relative scenery paths (relative to the main FSX/P3D installation directory) are not a problem for AIController, so I recommend using relative paths if possible.  However, absolute paths are very useful for sharing scenery between different simulators (e.g., FSX and P3D).

 

NEW:  Manual Landing Runway Selection:  If the user for whatever reason dislikes the current landing runway, which is based on prevailing winds, the user can instead enter up to three runways in AIMonitor to override all wind-based landing runway selections.  The AI will immediately switch to the new patterns, there will be no gradual transition.  When two runways are entered (e.g., 32L,32R), the load-balancing will be approx. 50-50.  When three runways are entered (e.g., 17C,18R, 18L), the load-balancing will be approx. 33-33-33.

 

NEW:  Max. Time Departing AI Remains Stopped on Taxiway Before Deletion.  Maximum time that AI taxiing OUT to the departure runway will be allowed to remained stopped on the taxiway before deletion (default 60 seconds).  Occasionally (even when not using AIController), departing AI becomes "stuck" just before runway entry holding up that AI and all the AI line-up behind it waiting.  This parameter will delete the AI that has become "stuck", allowing the AI behind it enter the runway and take off at a consistent rate.  The parameter cannot be used to increase waiting time higher than the FSX default value.  

 

NEW: AI Rollout Performance and Parking Based on AI's Aircraft.cfg File and ADE:  The AI rollout performance (e.g., braking distance) and parking preference (e.g., parking codes) are based on performance data in each AI's config file.  The parking algorithm has been revised to conform to typical FSX/P3D - ADE design methods (e.g., first AIController looks to gate radius, second AIControler looks to airline parking codes in the gate order as listed in the ADE parking list).  Version 1.4B automatically reads AI aircraft.cfg data to get each AI's braking, radius, airline code, etc. data, no user action is required.  Thus, the old GrabAIParkingCode utility is being retired.  An interesting note:  AIController will scan more than 5 airline codes per gate, thus eliminating the old FSX restraint.

 

NEW: Improved Landing Animations.  Consistent touchdown smoke, wheels (including nose-wheel) make firm contact with runway at correct angle, no more smoke effects when AI exits the runway and enters taxiway.  Touchdown smoke may be less prevalent at airports near sea level.

 

NEW: Special WinXP versions of AIController.exe and AIConv.exe are included in the distribution:  If you're running AIController in a networked configuration on an older laptop running WinXP, please use the AIController_WinXP.exe and AIConv_WinXP.exe files in the WinXP folder (see the readme in that folder for further instructions).

 

NEW: Latest version of SIDSTARConverter included, which allows you to select your preferred version of AIConv.exe.  The latest AIConv.exe is included in the AIController installation directory.

 

Prior Fixes for Version 1.4B:

 

HOTFIX 5:  Not released (rolled into Hot Fix 6).

 

HOTFIX 4:  (1) Incomplete aircraft_cfg_data.txt file causing long rollouts and parking misassignments; (2)  AIController reading crosswind runways (Jim Vile-ADE technique) causing problems with final approach assignments after STAR exit.  Note: use force runway option in AIController to support crosswind and parallel operations; (3) High final approaches (AI was incorrectly and quickly transitioning from Jetroute to STAR to final leaving AI high on final close to airport); (4) AI parking alignment; (5) Not all AI parking codes being read causing parking misassignments; (6) AIController crashing reading some STAR/SIDS/Approaches converted from the latest navdata; (7) (Evaluating) AI start successively going-around for no apparent reason.

 

HOTFIX 3: (1) AI being deleted after rollout at some airports.  Problem was due to conversion error (parking radius stored in scenery files as meters, AI wingspan stored in aircraf.cfg files as feet); (2) AIController sometimes crashing because of bad taxi-graph read; (3) Multiple fixes for non-enhanced mode, including custom final approach/autogen files not being selected and AI disappearing/re-appearing in AIMonitor; (4) Waypoint updates can be made while FSX/P3D running.

 

HOT FIX 2:  (1) Expanded search for *.cfg files in all of the SimObjects subdirectories.  I don't have all AI add-on packages, so I suppose it is possible some package will have the AI reside outside Simobjects.  If so, let me know and I'll patch it up again; (2) Fixed continuing issue with repetitive "Scenery Change Detected" message and then AIMonitor attempting update waypoints at the same time it starts AIController; (3) Fixed bug where sometimes AIControllerPath.txt doesn't sometimes contain the "taxi=" paths even though the user just ran an update; (4) Revised code to support "My Document" paths to better support auto-injection of departing AI at monitored airport (to maintain departure rates if user remains at airport for long periods of time); (5) Extended "Force Landing Runway" to support 3 runways; and (6) Removed ability to keep the "Force Landing Runway" option to remain active across different AIMonitor sessions (to reduce the chance of accidentally keeping option enabled when switching airports). 

 

HOT FIX 1:  (1) Repetitive repetitive "VARIABLE WINDS or TIMEOUT waiting for wind direction data" message in AIController; (2) Fixed repetitive "Scenery Change Detected" when AIMonitor started for first time; (3) Jet, prop, and VFR ONLY STARs/SIDs not recognized.  This has been fixed, but jet, prop and VFR only patterns are now specified by "pattern type" instead of "pattern name" to maintain consistency with final approach pattern types.  Any custom user files that user the "pattern name" to specificy jet-only, prop-only, or VFR-only files will have to be modified so that "pattern type" is used instead (this should be an easy modification, see the readme.pdf for further details); (4) Duplicate line added to AIControllerPath.txt file; (5) Excessive error messages when attempting scenery scan, such as when program paths incorrect.  AIController will now limit the error messages to five when an unsuccessful scenery scan occurs; (6) Sometime AI didn't park at gate at correct heading.  This has been fixed; (7) Generic left and right base pattern names reversed.  This has been fixed; (8) Pattern entry points (enhanced generic left-right patterns) sometimes not the closest to AI's present position; (9) Some approach files converted from nav data contained missed approach data causing AI to perform missed approach procedure (go-around) when landing.  This has been fixed.  (10) Addon/scenery folder not being scanned when active in FSX/P3D (this is where use-generated ADE files typically reside!); (11) AI's landing runway displayed more consistently on AIMonitor; (12) Departing AI not being auto-injected at some airports; (13) Locale issue continuing to cause errors when comma-deliminated error.  For those users who switched affected by this error, please give this fix a try; (14) Quicker assignment of wind-based runway selection, less incorrect STARs that have to be re-assigned later

 

Important Revision:  A registered version (4.9x or later) of Peter Dowson's FSUIPC utility is recommended to use the full functionality of AIController.  For example, AIController will not be able to delete problematic AI targeted for removal (e.g., parked at gate after arrival for the specificed parking time).  For networked configuration users, Peter's WideFS (ver. 6.9) is also recommended.

 

Various other fixes/revisions were also made.

 

Important New NOTAMS!

 

ENTERING FINAL APPROACH COORDINATES.  Do not use FSX when entering final approach coordinates for runway positions (e.g., touchdown, rollout) as the FSX in-game coordinate display (shift-Z) lacks the necessary precision (significant digits).  Use the freeware Airport Design Editor X (or similar) instead.

 

ENTERING RUNWAY DATA.   Ensure the runway number corresponds to the FSX runway description (for whatever airport scenery you are using).  Often, the latest FMS data will reference changed runway numbers that do not correspond to stock FSX runway descriptions.  For STAR files, if the runways do not match (STAR versus FSX), AIController will choose the STAR based on the waypoint having the closest entry point to the AI’s current position (i.e., not based on runway data), which often is satisfactory, but not always.  For Final Approach files, if the runways do not match, AIController might choose the final approach file having the incorrect runway data.

 

Please see the Original Thread Post for Additional Details Regarding Ver. 1.4 Including Installation Video.

 

Thanks!

-Roland

[alyzar 0]

Good Morning,

 

I seem to be having an issue with PFE (Pro Flight Emulator) and your great work.  If i start PFE and inject into FSX then run AIController I have no issues from what i can tell, it might be luck.   If i reverse the order AIController ( let it complete its start up) then inject PFE, I am told by PFE to "hold for traffic" at runway entrance.  There is no traffic inbound for departing runway, within reason,  and no traffic waiting behind me. If I am told to hold and i enter and take off PFE will then tell the non existent AI to line up and wait, even though there way nobody behind me.  I know its just a conflict in settings, but would "Inject departing AI at monitored airports...." be the culprit?

 Edit  There is an option in PFE "Hold for Jets and Hold for Props - sets the distance at which you will be told to hold for inbound traffic while waiting for takeoff clearance"  I bet thats it

 

 

Thanks for all your hard work

[daan_vb 37]

Hi Roland,

 

Can you tell me if your AI should follow AFCAD taxi path directions. I'm currently doing some spotting around Manchester EGCC, and I am seeing that the AI get a bit confused during taxi in. I'm seeing the AI stop on the runway and rotate really slowly to exit and spinning around nodes trying to find their way to the gate and generally taxiing around quite erraticaly (like double backing on themselves!) before finding their way. I'm wondering if the cause could be due to the direction of the pathways in the AFCAD?

 

Any ideas?

[Speedbird 217 126]

Hi Roland,

 

Can you tell me if your AI should follow AFCAD taxi path directions. I'm currently doing some spotting around Manchester EGCC, and I am seeing that the AI get a bit confused during taxi in. I'm seeing the AI stop on the runway and rotate really slowly to exit and spinning around nodes trying to find their way to the gate and generally taxiing around quite erraticaly (like double backing on themselves!) before finding their way. I'm wondering if the cause could be due to the direction of the pathways in the AFCAD?

 

Any ideas?

 

Just finished a flight to Munich and saw the same thing. AI was going all over the place, taxiing in and out of taxiways without clear direction and doing 180s and 360s a lot.

[FlyTweety 2]

This happens often after a new Hotfix because the settings are resettet to default settings. In the Taxi In section there´s a value that sets the maximum error margin to the next Waypoint. Default is "0", this was causing this behaviour in my opinion. I set it to 3 degrees, now i didn´t notice any 360´s anymore (exept on the Runway :( ).

 

Despite that, double parking keeps a huge problem. Seems, both P3D and AIC prefering the same Gates as long as they "think", the Gate is free.

I can observe it when i take a long time observation in EDDL. I assigned 6 parkings in a row for DLHX (Eurowings), Numbered V91 - V96.

 

Clearly new traffic will allways put on the parkings V91 and V92 first also if V93 - V96 are free. Also when V91/V92 are unrecognized occupied, never it happens that the next incoming AI will be assigned to V95, allways the lower numbers are prefered.

 

 

Further the time to release the Runway is too long.

I have trouble because the Traffic stops to early at the Runway, wait a few seconds, taxiing too slow (only 14 kts) to the exit which is far away, at the exit it stops again, wait again a few seconds and finally is out of the way.

I don´t know what you can do/ want to do to accellerate this process. The full stops are very harmfull, the long taxi to the exit as well.

Can you eleminate the full stops and add a faster taxi speed at Runways?

Though i would prefer a "dynamic" braking to the assigned exit, but i don´t get positive signals from you in this concern :ph34r: .

 

However, groundtraffic is already fine as long i set the values right. Unfortunatly i didn´t wrote down my best settings, and now, after installing HF6 they are gone again, so i have to figuere them out again :blush:

[Silver_Dragon 411]

Clipper Ocean Spray, has some method to enter great quantity of directories on SID and STAR section without  enter one to one? I have "all earth" countries and if I intent add them to a directory with all of them and put in the SID and STAR directory section, AIController "frezze", not read the scenery or similar, the bar don't show progress any progress (I wait ten minutes). I intent make them manually and add SID and STAR parameters copy of the IAConv to the AIController ini file but the program deplete all changer when I start the program and revert the the previous INI file configuration. any help about them?

[Clipper Ocean Spray 189]

Also it seems that approaching AIC-controlled aircraft  don't detect the user-aircraft at ground. Yesterday, I was "Cleared for take-off" alredy lined up on the runway, as there was touched down a 737 just before me.

 

Turn the runway anti-incursion feature on in the final approach section.  If user's landing and strobe lights on less than 10 kts (that is, the user entering the runway), AI on short final will be deleted if less than separation distance from user.

 

This suite of runway anti-incursion features was on by default, but I turned it off (default) because I was getting questions mostly about the "freeze" feature (user's aircraft on short final freezes AI from entering runway).   I really don't feel like breaking the suite into more sub-options, perhaps I'll relabel the option in AIMonitor to something more clear.

[Clipper Ocean Spray 189]

There is no traffic inbound for departing runway, within reason,  and no traffic waiting behind me. If I am told to hold and i enter and take off PFE will then tell the non existent AI to line up and wait, even though there way nobody behind me.  I know its just a conflict in settings, but would "Inject departing AI at monitored airports...." be the culprit?

 Edit  There is an option in PFE "Hold for Jets and Hold for Props - sets the distance at which you will be told to hold for inbound traffic while waiting for takeoff clearance"  I bet thats it

 

I don't think the inject option is affecting you.  However, AIController sometimes forces AI without "FSX" landing clearance to land in order to implement realistic STARs and final approaches (FSX requires the AI to intercept a generic FSX-generated IAF before landing clearance is granted).  Thus, if PFE is reading FSX status, it might get confused.  I'm glad there's a PFE option to increase the distance you wait for clearance before taking off.  Try increasing that as much as possible and make sure to enable the user-AI runway anti-incursion option in the final approach section of AIMonitor.  That should keep things clear at the runway start area.  

 

 

This happens often after a new Hotfix because the settings are resettet to default settings. In the Taxi In section there´s a value that sets the maximum error margin to the next Waypoint. Default is "0", this was causing this behaviour in my opinion. I set it to 3 degrees, now i didn´t notice any 360´s anymore (exept on the Runway ).

 

Thanks....when you find your preferred taxi-in settings, can you post them here?  I'm always interested in what other people think are better default values.  I'm beginning to think my default taxi values may be too aggressive.

 

 

Despite that, double parking keeps a huge problem. Seems, both P3D and AIC prefering the same Gates as long as they "think", the Gate is free.

I can observe it when i take a long time observation in EDDL. I assigned 6 parkings in a row for DLHX (Eurowings), Numbered V91 - V96.

 

Clearly new traffic will allways put on the parkings V91 and V92 first also if V93 - V96 are free. Also when V91/V92 are unrecognized occupied, never it happens that the next incoming AI will be assigned to V95, allways the lower numbers are prefered.

 

Yes, perhaps I can add an option for a reverse gate order scan.  The problem is the freeware/commercial scenery designers really put a lot of thought into their (ADE) gate order in order to maximize realism.  At least now AIController will recognize (according to user adjustable period) when an AI in injected into a gate that is already occupied and delete the previously parked AI avoiding double-parking.

 

 

I have trouble because the Traffic stops to early at the Runway, wait a few seconds, taxiing too slow (only 14 kts) to the exit which is far away, at the exit it stops again, wait again a few seconds and finally is out of the way.

I don´t know what you can do/ want to do to accellerate this process. The full stops are very harmfull, the long taxi to the exit as well.

 

I'm working on a fix for the tendency of AI to make sharp turns to find a runway exit.  That will help a lot I think when the AI doesn't have to make a sharp turn to get off the runway.  You can also increase the roll-out slow-down scalar in the Final Approach options page, which will cause the AI to slow-down more quickly across the board.

 

 

I'm seeing the AI stop on the runway and rotate really slowly to exit and spinning around nodes trying to find their way to the gate and generally taxiing around quite erraticaly (like double backing on themselves!) before finding their way. I'm wondering if the cause could be due to the direction of the pathways in the AFCAD?

 

Regarding sharp turns or runway exit, I'm working on a fix for that.  

 

Regarding the other issues, see FlyTweety's post above regarding increasing degrees margin of error to next taxi waypoint in the Ehanced Mode (Taxi-in) options screen.  Also, if the AI seems too "nervous" decrease the turn scalar to 2.0 or so.  "Scalar" is like a gain setting, too high and the AI acquires the heading quickly but tends to overshoot a lot (i.e., too much lead --> oscillation).  Set it too low and the AI overshoots less but may sometimes be too slow to acquire heading (i.e., too much lag).  All of the AIController "scalar" values work the same way.  You also might try reducing the max off angle before slowdown to 30 or so to make the AI more serious about slowing down when it starts to veer off target.  All of these taxi-in options can be applied "on-the-fly" while observing AI movement - you do not need to restart AIController or FSX/P3D.

 

Of course, when observing AI movement do not overlay a second window (I like using the Views-View Mode-AI Traffic instead), but I have so many warnings about that already in the program I think most people realize the issue.

[FlyTweety 2]

I don't think the inject option is affecting you.  However, AIController sometimes forces AI without "FSX" landing clearance to land in order to implement realistic STARs and final approaches (FSX requires the AI to intercept a generic FSX-generated IAF before landing clearance is granted).  Thus, if PFE is reading FSX status, it might get confused.  I'm glad there's a PFE option to increase the distance you wait for clearance before taking off.  Try increasing that as much as possible and make sure to enable the user-AI runway anti-incursion option in the final approach section of AIMonitor.  That should keep things clear at the runway start area.  

 

Thanks....when you find your preferred taxi-in settings, can you post them here?  I'm always interested in what other people think are better default values.  I'm beginning to think my default taxi values may be too aggressive.

 

Yes, perhaps I can add an option for a reverse gate order scan.  The problem is the freeware/commercial scenery designers really put a lot of thought into their (ADE) gate order in order to maximize realism.  At least now AIController will recognize (according to user adjustable period) when an AI in injected into a gate that is already occupied and delete the previously parked AI avoiding double-parking.

 

 

 

I'm working on a fix for the tendency of AI to make sharp turns to find a runway exit.  That will help a lot I think when the AI doesn't have to make a sharp turn to get off the runway.  You can also increase the roll-out slow-down scalar in the Final Approach options page, which will cause the AI to slow-down more quickly across the board.

 

 

Sounds good.

Today i failed totally to find best settings, morely i was totally lost because AIC finally din´t response to any setting as expected.

 

My initial intension was to adjust the break settings one more time.

So my first choice was to choose a small Airport in the vincity, EDLW (Dortmund City), released from Aerosoft long time ago, likly a ported FS9 Airport, but it has a  2000 m Runway, that is a good benchmark for Aircrafts like B737/A320.

 

I used "Airport EnhancerX" to generate more traffic than normal to reduce wait time.

 

So, first i tried the toe_break scalar_settings for different Models without AIC and noticed, they are way to short, compareable to EDDL (my common playground).

Surprising, finally i was down to 0.4 without remarkable improvements.

Now i was courious and wanted to compare it with EDLP Stock Airport, only a few miles away, (both Asphalt Runways), but also there almost the same decceleration.

 

Now i activated AIC, again the rollout was still to short at both Airports, additional i adjusted the roll out scalar in AIC down to 0.3, no reaction.

But with every restart of the P3D, with every new start of AIC less and less traffic came to the ground.

AIMonitor reports the assigned Waypoint, no Holdings, Traffic  controlled by AIC , but the Traffic was wandering arround, not heading to the assigned Waypoint, i found 2 Aircrafts flying back and forth between two "Enhanced Autogen Approach" Waypoints without any reported reason (but obviously triggered from MY apperance with my user Aircraft because loss of separation),  they never found back on track wherever i moved to and also with "0 nm" setting.

 

Completly weired, so rather again i will set ALL settings to default and start with a white paper.

 

I only don´t know when, frequent my job keeps me off from home for the whole week.

 

 

Yes, perhaps I can add an option for a reverse gate order scan. The problem is the freeware/commercial scenery designers really put a lot of thought into their (ADE) gate order in order to maximize realism. At least now AIController will recognize (according to user adjustable period) when an AI in injected into a gate that is already occupied and delete the previously parked AI avoiding double-parking.

 

By the way, wouldn´t it be better to assign a new gate to the taxiing Aircraft instead of culling the occopied gate?

[Clipper Ocean Spray 189]

 

 

But with every restart of the P3D, with every new start of AIC less and less traffic came to the ground.
AIMonitor reports the assigned Waypoint, no Holdings, Traffic  controlled by AIC , but the Traffic was wandering arround, not heading to the assigned Waypoint, i found 2 Aircrafts flying back and forth between two "Enhanced Autogen Approach" Waypoints without any reported reason (but obviously triggered from MY apperance with my user Aircraft because loss of separation),  they never found back on track wherever i moved to and also with "0 nm" setting.

 

Wow, obviously you've gotten decent results before, so something must have caused this.  When you get a chance to look at it again, please let me know what triggered this effect.

 

There isn't any chance you loaded more than one instance of AIController while tweaking with the different settings?  That reminds me, I need to add code in the next hot fix to prevent

accidental running of more than one instance of AIController.