Inebriated ATC or Design Flaw?

[Bob Keeshan]

ADE Team,

Not sure this is an ADE issue, but when I added a new taxiway to an airport I'm working on now and tested, the ATC seems to give some rather odd taxi instructions.

The runway has two taxiways, one on both the north and south sides of the runway. The ATC would direct aircraft to taxi using the south taxiway, cross over the runway, taxi on the new north section of the taxiway I created, and then cross back over the runway to the south taxiway again. In reality, I would have thought the ATC would have directed the aircraft to taxi only using the south taxiway.

Is this a common ATC AI issue, or does this indicate my new section of taxiway is somehow not working properly? Also, does the taxiway direction impact how ATC issues taxi instructions or how AI uses the taxiway?

Thanks.

BK

 

[rfields]

We'd have to see the file to be sure, but it sounds like you have an unconnected segment of the taxiway.

However, FS does have some weird taxi instructions at times.

One of the worst things people do in FS2004 is 'clean up' airports where a taxiway might be in several segments with the same name - such as KJFK's taxiway A (10 sections) or taxiway B (11 sections).

Those segments are built to keep the AI aircraft moving and not stopping for an aircraft halfway across the airport. Cleaning them up creates problems.

No, we haven't yet figured out how to do that taxi segmentation ourselves.

 

[Bob Keeshan]

Another Taxiway Question

Thanks for the thoughts. I checked my nodes and things seem to all be connected. I watched the airport run a while tonight, and I didn't get any weird taxi instructions this time.

HOWEVER...

To your point above, I did notice that ATC will tell airplanes to hold on a taxiway that parallels the runway when an aircraft is on final or taking off. Is this because the taxiway is positioned too close to the runway. What is the required distance we should use between the runway and parallel taxiways to avoid this holding command?

Thanks.

BK

 

[jvile]

BK

ATC locks runway crossing movement and undesirable parallel movement based on codes once the arrival is inside the outer marker or a prededermine mile (hardcode) if there is no Outer marker defined in the bgl. The point of lock is the Tower controller when they say "Cleared to land"

The other problem is the taxiway that parallels the runway if not segmented (Reggie's post) can cause a plane to stop before it raches the holdshort node in some cases. Make sure the entrance to the runway holdshold is within the proper ft. distance as seen with ADE if you select show hold short limits. There are other factors but we start with the basic's.

We have a work around that will allow Planes to continue across any active runway (including 2 parallels) regardless of where the arrival is located on the approach.

If you want to pursue that work around I suggest you download my FSX KMIA from AVSIM.

http://library.avsim.net/esearch.php?CatID=fsxafcad&DLID=104602

Open it with ADE File dropdown menu that says "Open Airport from bgl. Path to the KMIA bgl and open into the grid of ADE.

Now look at many taxiways (not all) that cross a runway and you will see what we refer to as the diamond method.

If you have any questions on this technique please ask after you review how it is done.

 

[kagazi]

Sorry to hijack the post a bit, but something Jim said caught my attention.

When is this statement NOT true? "Make sure the entrance to the runway holdshold is within the proper ft. distance as seen with ADE if you select show hold short limits."

I have seen a couple of examples when the holdshort is clearly outside of the proper ft distance as shown by ADE, but the holdshort works and the runway is active. One example is KLAS 1L and the hold short point that crosses the overrun section but misses the runway. Without knowing why I just figured that the overrun was treated by fsx to be part of the runway.

Another example is with an addon KIAD by Blueprint Simulations. The holdshort to both ends of 1L and 19R are clearly outside of the proper ft distance as seen by ADE, but the runway works fine. I'm curious if this is more a thumb as opposed to a hard rule.

Thanks.
kagazi

 

[jvile]

kagazi


The entrance point ONLY of a runway MUST HAVE the hold short limit set correctly in a foot/meter dimension as seen with ADE.

This is part of the code that tells User/AI Planes based on their Model.mdl radius in FS9 and it appears to be their 1/2 wingspan value in FSX to stop and announce, change from Ground to Tower freq., get permission from Tower to take the active runway, etc.

If the holdshort is too far away from the runway AI Planes stall at that point and will not advance forward because they never get the trigger command to contact Tower. This is easily seen with the Traffic ToolBox utility where and why the advancement stall occurs.

We have a fudge factor built into ADE that purposely sets the holdshort outer ring closer to the side of the runway then what FSX actually needs. If the circle of the holdshort limit does not touch the side of the runway we still get advancement and the correct trigger point for the AI User plane.

Also there are some runways at certain airports that will work without a holdshort node for AI/User advancement on to the runway.

When you are on a Ground freq taxing out to a runway you should not be able to switch to or contact Tower until you are within the limits specified for the holdshort. You will find some runways and airports don't hold true to this rule based on 2 different unrelated problems.

One is the runway just simply does not have a code for the holdshort.

The other is the parallel taxiway could be within the invisible bounds that help in deciding what Runway is the active for takeoff based on parking spot position.

SO 1st rule using the basic is the only holdshort that is important to advance the plane forward onto the runway is the entrance of the runway being used. That is normally the 1st taxiway at the end of the runway but does not have to be.

2nd rule is all other holdshorts at lets say a taxiway used to exit a runway do not have to be in the proper draw circle. But it is imprtant that a holdshort exsit somewhere after the plane departs the runway to reverse triggers used for taxi to the Parking Stand.

 

[rfields]

One example is KLAS 1L and the hold short point that crosses the overrun section but misses the runway.

KLAS 1L has an offset threshold - not an overrun.

7R does have an offset threshold and 400 foot blastpad, but no overrun.

Now 19R does have a 775.0 ft blastpad.

While the charts and runway information show the blastpad - the aerial photos / satellite photos I find show the area marked as a displaced threshold.

The yellow chevron marks are on blastpads. It appears that FS2004 incorrectly drew yellow chevron marks on both overruns and blast pads.

AFCAD2 codes both blastpads and overruns the same. AFX perpetuates this bad practice. FSXPlanner and ADE both understand that these are two different things with different properties and usage.

I haven't tested to see if there is a difference in behavior of the blastpads and overruns in FSX - but will add that and KLAS to the list.

 

[scruffyduck]

In terms of what is seen as runway (and I don't know how FSX really treats overrun/Blastpad in this context) but offset thresholds are part of the runway and overrun/blastpads are not.

 

[rfields]

After reviewing some FAA documentation - I see that FS does things wrong at KLAS.

While the areas may be blastpads - they should NOT have the yellow chevron markings. Anywhere an aircraft is expected to taxi must not have yellow chevron markings.

Those markings mean no aircraft normally use that area - period, end of discussion.

We will probably have to make that part of the blastpad into an offset threshold to have it appear correctly in FS.

 

[scruffyduck]

We will probably have to make that part of the blastpad into an offset threshold to have it appear correctly in FS.

or use an apron?

 

[Bob Keeshan]

BK

ATC locks runway crossing movement and undesirable parallel movement based on codes once the arrival is inside the outer marker or a prededermine mile (hardcode) if there is no Outer marker defined in the bgl. The point of lock is the Tower controller when they say "Cleared to land"

The other problem is the taxiway that parallels the runway if not segmented (Reggie's post) can cause a plane to stop before it raches the holdshort node in some cases. Make sure the entrance to the runway holdshold is within the proper ft. distance as seen with ADE if you select show hold short limits. There are other factors but we start with the basic's.

We have a work around that will allow Planes to continue across any active runway (including 2 parallels) regardless of where the arrival is located on the approach.

If you want to pursue that work around I suggest you download my FSX KMIA from AVSIM.

http://library.avsim.net/esearch.php?CatID=fsxafcad&DLID=104602

Open it with ADE File dropdown menu that says "Open Airport from bgl. Path to the KMIA bgl and open into the grid of ADE.

Now look at many taxiways (not all) that cross a runway and you will see what we refer to as the diamond method.

If you have any questions on this technique please ask after you review how it is done.

Thanks for the tips. I downloaded your KMIA last night, but haven't had the opportunity to look at how you handled parallel taxiways yet. The issue I noticed with my airport was the ATC issuing hold commands while the plane was moving along a parallel taxiway. The plane wasn't try to cross the runway.

I'll take a look at KMIA tonight and see what you did.

BK

 

[kagazi]

Thank you Jim/Regie and Jon. Your explanations would explain why KLAS appears to be unique when in reality it's not, and also why possibly the addon at kiad appears to not abide by the distance rule when in actuality it does since you can't get clearance until you're close to the hold short and the hold short circle as drawn by ADE is close to the edge of the runway.

 

[rfields]

For KLAS - one of the INTENSELY frustrating things the Aces team does for some airports is put a special fix code for that airport only into some of the .DLL files.

Rather than fix the system, they make a one-off exception.

 

[jvile]

BK

The issue I noticed with my airport was the ATC issuing hold commands while the plane was moving along a parallel taxiway. The plane wasn't try to cross the runway.

Keep us posted on this one. A plane should not be told to stop along a parallel taxiway if the arrival is still airborne. If the arrival has touched down and slowed, at that point it is getting updated ATC data before you hear anything on the radio.

In some cases this will cause the plane taxing on the parallel taxiway to stop movement until ATC has updated all plane movement data.

Runways do have a invisible bounding rectangle shape around them but it is not a constant shape size. There could very well be a minimum distance between the parallel taxiway and the runway but I have not encountered that problem as of yet.

You could save your airport (work from a BU) and then do a test by moving the taxiway further away from the runway and see if it changes anything.

 

[Bob Keeshan]

Runways do have a invisible bounding rectangle shape around them but it is not a constant shape size. There could very well be a minimum distance between the parallel taxiway and the runway but I have not encountered that problem as of yet.

You could save your airport (work from a BU) and then do a test by moving the taxiway further away from the runway and see if it changes anything.

I will work on this tonight. There are two parallel taxiways for the runway, one is closer than the other. The hold command was issued for a plane on the closer taxiway, while ATC allowed a plane on the other taxiway to continue to taxi until it reached the hold point. Like I said, I'll test this again tonight to confirm.

BK

 

[glottis]

Jim
what you are saying does have a lot of merit,
here's what I've often seen, many a time i have seen a aircraft push back from a gate, turn then stop! no other aircraft are moving,
no support vehicles nearby (another reason for aircraft to stop anytime) and a aircraft on final, the aircraft lands,
taxi's past the still waiting aircraft to the next vacant gate and then the waiting aircraft will taxi for takeoff, if it hasn't timed out,
my thinking here is a vacant gate is already allocated to it even before the aircraft has landed

Ray

 

[jvile]

There are two parallel taxiways for the runway, one is closer than the other. The hold command was issued for a plane on the closer taxiway, while ATC allowed a plane on the other taxiway to continue to taxi

BK

Now that is a true statement because you paralleled taxiways rather then series a taxiway.

ATC has a pecking order it uses for who gets the first takeoff slot. When a taxiway is in series the pecking order normally starts after 2 planes pushback and which one gets ground clearence first. ATC will then hold one of the planes at a node if it needs to be sure the correct plane gets to the holdshort first. We would like to think that ATC is random but it is not. Because the lineup of takeoff occurs way back at the pushback point we are lulled into thinking ATC has some sort of human intellegence but it does not and it is all based on a pecking order code.

Now when 2 taxiways are parallel, that gives ATC more choices to work with. However there will come a point when it must hold one plane to give the other time to get to the holdshort first because it was slotted first out of the GATE.

 

[colin_d]

Hi Ray,

Jim
my thinking here is a vacant gate is already allocated to it even before the aircraft has landed
Ray

I've been watching a lot of traffic movements over the last few weeks and it seems that ATC plans well ahead in as much as it reserves airline coded gates for aircraft with an appropriate code that are due in the next hour (maybe more) and probably does the same for traffic to uncoded gates but on a lower timescale. Based on my observations your thinking seems to be quite correct. This might explain some of the 'why the hell is it doing that' type holdups which often crop up on large, busy airports.

Colin

 

[jvile]

Reggie has a lot of the test data on when ATC assigns a specific from one trigger to the next based on his knowledge of the TrafficToolBox.dll usage that he works with.

He will probably be back in here in due time and hopefully list some of that data to help understand the pecking order of assigned gates going in and assigned runways coming out of the Gates.

 

[Bob Keeshan]

BK

Now that is a true statement because you paralleled taxiways rather then series a taxiway.

ATC has a pecking order it uses for who gets the first takeoff slot. When a taxiway is in series the pecking order normally starts after 2 planes pushback and which one gets ground clearence first. ATC will then hold one of the planes at a node if it needs to be sure the correct plane gets to the holdshort first. We would like to think that ATC is random but it is not. Because the lineup of takeoff occurs way back at the pushback point we are lulled into thinking ATC has some sort of human intellegence but it does not and it is all based on a pecking order code.

Now when 2 taxiways are parallel, that gives ATC more choices to work with. However there will come a point when it must hold one plane to give the other time to get to the holdshort first because it was slotted first out of the GATE.

I understand what you're saying here, but in my situation the aircraft was heading toward parking using the parallel taxiway when it was directed by ATC to hold, so I'm not sure the pecking order theory applies here. Also, it wasn't held at a node, but between nodes.

I noticed on your KMIA that you have a taxiway L that parallels rwy 8/26 and is almost the same distance from the runway as mine (62m). Have you ever noticed ATC issuing hold instructions for aircraft using that taxiway when another aircraft is either landing or taking off?

I did get FSX to repeat the hold instructions to another aircraft that was using the taxiway to get to parking, so I moved the taxiway out farther from the runway (from 62m to 72m). The maddening thing is I can't set up the situation myself (at least I don't know how) to see if the problem repeats itself. Instead, I have to be lucky and wait for aircraft to be in the right position at the right time...