Metering on the Network

[Brandon Barrett]

Separation starts from the ground up. I've been seeing a lot of facilities lately (more than when I first join the network) put a dedicated C1 on TMU which is great. FNOs are tough because traffic comes from all different directions but I have seen facilities coordinate with other tower controllers in the adjacent ARTCC or even two ARTCCs over to slow guys down off the RY. In my opinion, that's where it needs to start so the centers and tracons don't get saturated. I agree with Dan Everman, holds should be a temporary 20 min fix, we shouldn't have 10 a/c stacked up waiting to descend!   

[Karl Mathias Moberg]

... I think VATUSA has to own it and do it week in & week out.  We have no fancy software and the number of variables we deal with exceed those in the real world.  Its going to take trial and error to build some proficiency.  It's also going to require some people sitting out of the fun of actual controlling.

Having each ARTCC attempt to do it 3-4 times a year isn't going to cut it, in my opinion.  I doubt anyone is going to volunteer to be the TMU for every event.

I think it depends on the size and activity of the facility. As Jeremy has pointed out previously - we at ZNY already do this on our own. We talk to our neighbors and have found a good working relationship that in most cases work very well. You don't need to get VATUSA involved - just talk to your neighbors and create an agreement on how you want to deal with the issue.

Yeah, it's not going to solve the aircraft flying cross country from LAX to JFK for a JFK FNO, but those we can plan for, it's the short pop-up flights that typically cause an issue and those your neighbors can deal with.

[Jeremy Peterson]

It's important to remember that TMU is a process, not just a function. It takes planning, familiarity of airspace, open lines of communication, and collaboration. I remember for the PHL FNO a few months back, I was TMU for that event, I made sure to be cognizant of what could go wrong before it actually hit us.

The constraining factors for this event were staffing levels and high anticipated volume. I anticipated two possible scenarios and proposed workarounds to their constraints. Here's the first possible scenario:



This scenario features heavy ZBW flow on J121 in combination with high DC Metro (PCT) volume. The solutions were pretty straightforward: (1) BOS ground stop for 1 hour and then a PCT ground stop about 15 minutes after, to end at the same time as the BOS one to stabilize demand; or (2) extensive MIT with ZNY, ZDC, and ZBW.
This is the 2nd scenario and it actually came to fruition:



Here, we have heavy ATL volume with little room to accommodate flows from ZOB or ZNY66 (flights from/through ZBW). Unfortunately, there is little coordination we do with ZTL (a major pitfall for ZNY since there's so much traffic between the 2 ARTCCs) so we would have to heavily rely on ZDC. The best possible action would have been an ATL ground stop for no more than an hour to clear the airspace and reaccommodate capacity to allow for that flow. Once it was released, we would implement MIT for ZDC overflights landing PHL (i.e., the ATL departures). Option 2 was to stop DC Metro departures and place ZBW on CFR with ZNY_TMU in order to optimize the arrival streams. The last option was to reroute aircraft through ZOB (which we ended up doing) to help out ZDC with the MIT. This required coordination with ZDC, ZTL, ZID, ZOB, and ZNY and was the least desirable option; we only used it because we were unable to get ATL to give us the ~40 MIT we needed to accommodate other arrival streams in time.

[Matthew Kramer]

This is a great discussion.

I've seen TMU work well on multiple occasions here out west, and the only thing I'd like to add to that is a small wishlist.

1) As Nick said earlier in the thread, pilot outreach. It would be great to work with the organizations offering pilot ratings on the network information and education on how a virtual TMU might function during a high traffic even. During CalScream we use slots, and Cross the Pond does something similar, slot wise. It would behoove us to make prominent the concept of traffic management as an increase in realism and a net benefit to pilots.

2) The FAA uses software that puts all aircraft inbound to an airport (or an airspace) on a cool timeline that makes it easier on TMU and lets those controllers see immediately where aircraft are bunching up. It would be cool to have some sort of VATSIM approximation of this for events.

[Shane VanHoven]

2) The FAA uses software that puts all aircraft inbound to an airport (or an airspace) on a cool timeline that makes it easier on TMU and lets those controllers see immediately where aircraft are bunching up. It would be cool to have some sort of VATSIM approximation of this for events.

Something along these lines are in the works. But I agree pilot education on this topic is a must, because TMIs only work with cooperation from the pilots.

[Daniel Hawton]

2) The FAA uses software that puts all aircraft inbound to an airport (or an airspace) on a cool timeline that makes it easier on TMU and lets those controllers see immediately where aircraft are bunching up. It would be cool to have some sort of VATSIM approximation of this for events.

I've attempted this, but with such an unknown when it comes to wind aloft and who uses what, where does it change, etc. it was completely unreliable, so I scraped it.

[Matthew Kramer]

2) The FAA uses software that puts all aircraft inbound to an airport (or an airspace) on a cool timeline that makes it easier on TMU and lets those controllers see immediately where aircraft are bunching up. It would be cool to have some sort of VATSIM approximation of this for events.

I've attempted this, but with such an unknown when it comes to wind aloft and who uses what, where does it change, etc. it was completely unreliable, so I scraped it.

True, there are a lot of variables with sims and weather, not to mention I doubt most pilots (and I'm guilty of this myself) file accurate true airspeed. I wonder how far off the accuracy would be, though, if we made basic assumptions about winds aloft and built in some sort of padding. As shane said, accounting for roughly 80% of the airport's actual capacity.

That, and just knowing the limitations of such a tool could just make it easier to visualize.

Forgive my ignorance on the subject, but what are some actionable steps we can take to start educating pilots on the network? Is VATSIM currently reworking training programs or is it still a hodgepodge of 3rd parties with the ability to assign pilot ratings?

[Robert Shearman Jr]

Forgive my ignorance on the subject, but what are some actionable steps we can take to start educating pilots on the network? Is VATSIM currently reworking training programs or is it still a hodgepodge of 3rd parties with the ability to assign pilot ratings?

It is essentially that, yes, although a rather pessimistic way of looking at it in my opinion.  But that's a discussion for another day.

What sort of pilot education are you and Shane suggesting is needed?  Over and above what we already do -- such as helping people understand how to follow ATC instructions, how to fly holds, how to control their airplane when asked to divert from the programmed route, etcetera -- what exactly are pilots lacking in understanding that the Pilot Training organizations can help with?

The biggest limitation that the Pilot ATO program faces is not the decentralized aspect of it.  It's the fact that none of it is mandatory in any way, shape, or form.  Only 350 VATSIM pilots **worldwide** on the network hold the P4 rating.  Anything we do to further pilot education about IFR traffic metering is going to be a drop in the bucket until that changes.

[Ryan Parry]

Forgive my ignorance on the subject, but what are some actionable steps we can take to start educating pilots on the network? Is VATSIM currently reworking training programs or is it still a hodgepodge of 3rd parties with the ability to assign pilot ratings?

what exactly are pilots lacking in understanding that the Pilot Training organizations can help with?

Although this has never seemed to be a big deal with the network (and a little off subject), preflight planning is fairly important. I'm not talking about clicking a few buttons and having simBrief spit out whatever fuel load either, I am talking about know what you're planning in simbrief/ PFPX and the decision making that goes into preflight planning.

In my experience when we are hosting a major event and inevitably get jammed up, we go to holds and start getting people worried about fuel. If you're flying into a major event, it's best to plan some hold fuel and maybe some contingency fuel for possible reroutes/ excessive vectoring etc. It seems like the majority go to simbrief or fire up PFPX and just plan the flight straight and have no idea what they planned. The reality is they're taking the bare minimum into a very fluid and busy event.

I'd be willing to help out if you needed it and I know there are a few other dispatchers on here as well that may or may not be interested in giving some pointers.


But getting back on subject, we came up with a faux sort of GDP thing with ZLA. We haven't used it in a full scale event, but in the few tests we did it seemed like it may work. Basically, since we can't give an EDCT and expect the pilots to know what to do with it, we control the delay on the ground via gate holds.

Once a clearance is issued we instruct all flights to call for push back, then the ground controller can meter the outbound taxiing aircraft. Either they get to push and taxi, or they are told they have to wait and are put into the queue. It gets a little facility specific at this point because we are limited on how many aircraft we can have waiting at our runways, but basically the ground controller will send the aircraft waiting the longest once they are able to based on various factors.

When they get to the runway we use time to create MIT when we have successive departures to the same airport that has a high arrival demand. It's fairly involved, but it was something worth trying.

[Matthew Kramer]

The biggest limitation that the Pilot ATO program faces is not the decentralized aspect of it.  It's the fact that none of it is mandatory in any way, shape, or form.  Only 350 VATSIM pilots **worldwide** on the network hold the P4 rating.  Anything we do to further pilot education about IFR traffic metering is going to be a drop in the bucket until that changes.

Totally agree. Forgive my pessimism, what I was driving at was more along the lines of "what does a new pilot on the network get as far as training, and what sort of outreach does the network do to existing pilots?"

Forgive my ignorance on the subject, but what are some actionable steps we can take to start educating pilots on the network? Is VATSIM currently reworking training programs or is it still a hodgepodge of 3rd parties with the ability to assign pilot ratings?

what exactly are pilots lacking in understanding that the Pilot Training organizations can help with?

Although this has never seemed to be a big deal with the network (and a little off subject), preflight planning is fairly important. I'm not talking about clicking a few buttons and having simBrief spit out whatever fuel load either, I am talking about know what you're planning in simbrief/ PFPX and the decision making that goes into preflight planning.

When they get to the runway we use time to create MIT when we have successive departures to the same airport that has a high arrival demand. It's fairly involved, but it was something worth trying.

I've seen the gate holds work at ZLA pretty well. Next event I plan to monitor the ground controllers a little more closely to see what sort of pilot friction (if any) there is. I know we got one or two negative feedback reports recently, but I suspect it was more about traffic over saturation for a single controller. We should work on formalizing some of these programs.

As for flight planning, I agree there, too. Not just fuel, but also what a good route looks like. Robert and I discussed it a bit during a VATSAR event we had at ZLA on Sunday.

[Matt Bromback]

This has always been something I wish most ARTCC's managed better was TMU. You see it over and over again traffic levels far outpace the controller capacity, network capacity (voice), airport capacity, etc...

A lot of people aren't mentioning this, what about staffing levels the event ARTCC has? If an ARTCC can only staff one guy on finals for a position that can fit 3 final controllers, what is his limit? I personally feel anytime you get more then 10 pilots on your frequency for a very busy sector (vectoring, finals, arrivals, etc..) due to voice and network limitations you can quickly lose control over sequencing. The pilots are NOT professional pilots and thus add to more radio calls being necessary also.

Many times when working ZTL (non events) I and a few other controllers would simulate re-routes due to weather in the ATL area. You would be surprised on how many pilots participated and LOVED the fact we were simulating that! We occasionally would get the pilot who did not want to participate and we would just send him on his normal route no questions asked. So while many in this thread are afraid pilots will not want to sit on the ground, or just disconnect, yes some will do that, however I think you will be surprised on how many would actually like that.

Also how many times have you seen a FNO where they have 4 center controllers online but only 1 is super busy why the other 3 are twiddling their thumbs??? ARTCC's need to be able to recognize this and take initiative on opening up further sectors, rerouting traffic, something. Again using ZTL as a example since I am the most familiar with it...They use 1, 2, 3, next to their position name when staffing a position. Why you may ask?? So that if you have 4 center controllers online they can very quickly switch frequencies to open/close different sectors without having to log off and change callsigns. To this day I really do not understand why other ARTCC's have not adapted this policy, it works wonders.

[Maius Wong]

2) The FAA uses software that puts all aircraft inbound to an airport (or an airspace) on a cool timeline that makes it easier on TMU and lets those controllers see immediately where aircraft are bunching up. It would be cool to have some sort of VATSIM approximation of this for events.

You mean something like this?  This is something that I've been working on for the last little that will hopefully help us better manage traffic flow.  We'll see how well this works for our next event.

[Matthew Kramer]

2) The FAA uses software that puts all aircraft inbound to an airport (or an airspace) on a cool timeline that makes it easier on TMU and lets those controllers see immediately where aircraft are bunching up. It would be cool to have some sort of VATSIM approximation of this for events.

You mean something like this?  This is something that I've been working on for the last little that will hopefully help us better manage traffic flow.  We'll see how well this works for our next event.

You are the hero this world needs. Exactly this. Looking forward to seeing this develop further.

[Shane VanHoven]

I want to thank everyone for their input, I hope the discussion continues.

Maius has been an absolute boss at doing the website end for ZMP, and he decided this TMU project was something that was doable. Here's an idea of how deep of a discussion we went into about this:

I were looking at stats from our last event and found the busiest 30 minute period, which we all agreed was beyond our max (factoring in network lag and bad pilots). So we figured a max, took 20% off of it, and decided that it's the "AAR" than we can use to shoot for during events. We just have to figure out the different rates for the other runway configs.

Holy crap we're dorks.

[Jeremy Peterson]

I want to thank everyone for their input, I hope the discussion continues.

Maius has been an absolute boss at doing the website end for ZMP, and he decided this TMU project was something that was doable. Here's an idea of how deep of a discussion we went into about this:

I were looking at stats from our last event and found the busiest 30 minute period, which we all agreed was beyond our max (factoring in network lag and bad pilots). So we figured a max, took 20% off of it, and decided that it's the "AAR" than we can use to shoot for during events. We just have to figure out the different rates for the other runway configs.

Holy crap we're dorks.

MSP Rates:
http://www.fly.faa.gov/Information/west/zmp/msp/frames.htm

Also, I'm very interested in your TGUI you made for MSP! PM me please, I'd like to know more!