Flight Scheduling

[Guest EverettP]

I am currently working on my masters thesis in engineering management and trying to lock onto a topic right now. My advisor developed a heuristic for process scheduling based on priority versus deadlines and wants me to help him see if it apply's to and how how well it works in non-manufacturing related processes. So here are my preliminary questions...

1) What method does the AF use to schedule it's flights (software,etc)?

2) Is the method standard for both training environments and real missions?

3) Are there any significant problems that arise with the way scheduling is currently done?

Once I go AD in June I'll get to see this a little better but I need to get the ball rolling now. I want to get my thesis done before I start UPT in November. Thanks for the help in advance, I'm sure I'll have more questions later.

Edited January 20, 2009 by EverettP

[slacker]

I scheduled for years. I used homegrown excel products, C2IPS, PEX and G2. I still think the excel files worked great. Some of the giant-gazillion dollar programs are just too big and too "one size fits all." As gearpig said, you're tackling a giant from the outside. It would be easier once you have been around for a while. Even then, it's an enormous task, different units, different challenges and processes.

good luck.

edit- isuckattypin

Edited January 21, 2009 by slacker

[stract]

1. plenty of people talk bad about PEX (Patriot Exalibur), but once you take some time to get familiar with the program and set up your defaults properly (as a scheduler), it's actually a great aid to scheduling. Yes there are some things that could be better, such as when it notifies you there is a conflight, it should pop up with the conflict instead of telling you there is one, so you don't have to navigate away from the page you're on just to see what the conflict is and if it's ok for it to be there.

You can set it up to automatically add 12 hrs of crew rest prior to a flight; you can check people's currencies, and in fact filters by grounding currencies to see who should be the priority for a particular event. It's a great SA-building tool. At my previous base, MX was also doing things in PEX, like pulling up the schedule on their own, which made the pen and ink meetings go a lot faster, and plugging in tail numbers against lines.

My current unit actually put the whole FHP in PEX, which my previous unit never did. Both use PEX for all Stan/Eval functions. There's a nice Go/No-Go display in PEX with green or red dots, so instead of having to look down a printed sheet from ARMS to figure out what someone is red for, you just click on the red dot and it tells you that person is overdue an instrument approach, or is missing a date for water survival training, or they still need to sign off FCIF 09-03B.

Then on the other side of the house, those who are not schedulers can login to PEX and easily see what they are scheduled for, can input commitments they have upcoming (outside of 2 weeks is our rule) such as medical appts or TDYs or Leave. "If it's not in PEX, it doesn't exist" is our motto for commitments and appts. It makes accountability a breeze. You can see with a few clicks of a mouse, who should be here and who shouldn't.

Jokes about Skynet aside, PEX is a great tool, just like all things, you have to actually take the time and/or get sent TDY to the week-long training at Eglin to use it to its full potential.

2. There's talk of bring PEX online downrange, but currently there are compatibility issues with currency information managed by ARMS on NIPR and the inherent classified nature of the ATO being on SIPR. In my community, we fly hard crews, who sit alert for multiple days at a time, so PEX isn't really needed for that. Different for a fighter squadron, though.

3. Haven't run across any so far. Having a backup server and database of PEX information is probably a good idea, but unless the folks running your PEX on base are complete morons, that's already being done.

[ACCBoomer]

Pex, GDSSII, and a huge whiteboard with lots of pucks

[C130GuardBum]

I think TACC uses some program to figure out what time the flight needs to be so that the crew shows too early for breakfast, flies through lunch and lands several time zones away after everything has closed so that you can go the entire day with the only thing to eat is a box nasty.

[WeMeantWell]

1) What method does the AF use to schedule it's flights (software,etc)?

2) Is the method standard for both training environments and real missions?

3) Are there any significant problems that arise with the way scheduling is currently done?

First you need to narrow this topic down considerably. I would do this by choosing a command to start with, AMC is completly different than ACC, versus AFSOC, AFMC, etc. Each has completely different taskings and time frames. Within each command, things still vary considerably. Within AMC you have wide range of missions. Some with months of advanced planning: Ice, Nuclear, state department, Presidential, some with planning limited to hours: locals(training), alert, diverts. Of course how we train and how we fight are completely different from those processes (read any PME book to find out how).

So I would choose one area within a command and focus on that. As Slacker said, the only people that know what is going on, are actually currently in that job; requirements/SOP/etc change daily, not to mention a whole new adminstration will shake things up a bit.

If you want to check the box for your degree, choose something simple like deployments (years in advance) or state department moves.

You asked for it, so here goes, amazingly simplified version:

1) Someone somewhere decides they need something. If they need something blowed up, they go to the CFACC (Combined Forces Air Component Commander) who then tasks a service to blow it up, or if that services equipment is CHOPed (Change of Operational Control) to the CFACC, then he just chooses that himself (has there ever been a female CFACC?). After hours of bickering and complaining (B&C) about who is getting credit and who is not getting used enough, The CFACC lackies publish an ATO (air tasking order) so those individual peeps know where to go and what to blow up. Luckly ACC usually runs this, so they got it down and usually works pretty good for the rest of us. Mostly classified/OPSEC stuff, forget it.

If someone needs something Moved, they contact TRANSCOM, who decides what is worthy of movement by air, then train and truck, then boat. If it goes by Air, AMC gets a call, and so on. After the user finds out that his tank is going to take 3 months to get somewhere on a boat, there is plenty of B&C and days before the user is supposed to be in place, they now have a high enough need that they become priority and they launch alert birds from somewhere to pick their bitchin a$$es up. Or someone gets called off of leave to fly bottled water to kuwait. From there, AMC can choose who flies what, when the user finds out they are going on a C-5 theres plenty of B&C and some C-17 mission that was going to S America, diverts to Pope and flies to Qatar with $hitter trucks. Or some guy with a tube of caulk to put in a windsheild is flown from seattle to baghdad, by himself, with no other cargo.

Once it is tasked to a wing, all the neat/fun/interesting trips are sucked up by the reserves who claim it is the only time their guys can fly. The next set of good deals go to FGOs that claim that trip happens to fall in their 'fly break', when the aforementioned mission to SA turns into a 3 week Afghan shuttle mission, their wife is suddenly in 'the window' and they have to stay home and make a kid. Finally the great line pilot that lives to fly and hasn't seen his family in 4 months, walks in the sqd after flying a 36 hr day and is told to go home and Crew rest for tomorrows mission, for he is "the only one". End result: a mission that was known about 3 months ago, is scheduled on a friday afternoon for a sat launch.

2) Luckily Training is completely different. However there is plenty of B&C to go around. Some civilian who has no idea what a rudder pedal is, is put in charge of scheduling training missions. This person has a kingdom more powerful than Titus himself. However, he has no concept that the sun sets later during a certain part of the year, so when you ask for some night AR, you end up wearing sunglasses. Every single base in the AF does this completely differently and their way is the correct way, the only common piece is a horse blanket; I have no idea what this is, but you sound important when you say it and somehow a tanker shows up when you need it.

3) Have you been paying attention? You have no idea... engineering management? This is Phd level work spread over a decade among dozens of Studs, and even then it will solve the problem of today, not for tomorrow. They have brought in experts from all over the world, nothing helps. Choose something simple, like how to air-refuel UAVs, how to safely move nukes, how to find bin laden.

[Guest bikerdood]

Not sure if I can add anything to what other folks have said, but I may be another data point for you.

Last base, was a flight scheduler for about 6 months (UPT/T-1s) before PEX got to AETC. TIMS is supposedly able to be used as a scheduling tool, but the AF-level TIMS program managers and base-level administrators are so far removed from the process and the program is so dependent on the base's network speed, that it was unusable as a scheduling tool. TIMS was only used by schedulers to put the schedule (built elsewhere) onto the network so that SARMS/Supervisors could access flight data and for the inputting of 781s.

For actual scheduling in the filght, we (flight schedulers -- 4 flights, about 12 schedulers) used a homemade Excel program, and more importantly, magnetic puck boards. Nothing beats a bunch of names on magnets to really sit back, digest and forecast two days in advance.

Once I left scheduling, I moved to squadron Programming (otherwise known as Mission scheduling, central scheduling or squadron Current Ops) for a couple years. There we used another, much more complicated, but extremely useful excel program to actually build the schedules. From pre-ORI meetings that we had with the other UPT bases, it seems that all T-1 programmers used the same product that was built about 8 or 9 years ago by a really smart FAIP. Great product.

We'd build in Excel, then manually export the final product into TIMS so the flight schedulers could put names against the jets. We'd post our excel schedule on the shared network drive and make sure all schedulers (academics/sims and flight/group schedulers) could access it, so this made the weekly meetings go a lot quicker. We'd have a couple of penciled-in changes and that's about it.

I could go into more detail about the entire scheduling process, but all I really know is UPT scheduling, along with the Sq & OG approval process and coordination with maintenance. I'm not familiar with scheduling in PEX or C2IPS. UPT may be going that way, but not when I left late last year.

My biggest pet peeve with the Air Force provided scheduling software is what others have said -- it's one-size-fits-all, it's got a ton of stuff that no one uses and it's wicked slow. The program managers and administrators don't necessarily see this because they don't always have to depend on the base's comm, as they're usually located in the same building as the server. Those programs cost large cash, when we could just lock a couple of smart FAIPS in an office for a week and get a more useful product (for scheduling, not tracking).

Hope this helps,

-BD

[Guest EverettP]

Thanks for all of the helpful information, I really appreciate it. As I get a better of idea of what I want to do and what I'm able to accomplish I'll be able to ask more in depth questions, but until then, I'm still in the big picture stage. I want to do something useful related to the Air Force, but I really just want to get on the same track that my advisor is so I can ensure that he has a vested interest in my project and that I know what to expect. This really helps me a lot so now when I go to meet with him again I can have a better idea of what direction that I want to go. Another general question...

Is the training scheduling based more on just fitting everyone in based on the planes available, managing priorities of who needs to get up, or both?

Thanks again for the help guys

[Guest C-21 Pilot]

...you should have gone through TUI University.

[WeMeantWell]

Is the training scheduling based more on just fitting everyone in based on the planes available, managing priorities of who needs to get up, or both?

Usually a multi-tiered approach, X number of pilots, WSOs, navigators, engineers, loadmasters, need X number of currency events. From historical analysis, approx # of currency events/sortie drives total # of sorties per quarter. Each base is alloted an amount of flying hours/year to fit this all in. From there, the needs of the Sqds are looked at, who is deploying, who needs more hours, who has a lot of upgrades or un-current people, etc. Then the Scheduler in each squadron takes an insane amount of input from everyone (availability, currency needs, wifes menstrual cycle, etc) and poops out a schedule. Some sqds are good, in AMC the best is about 4-5 days advance notice, but typically its 1-2 days notice. However, other communities have users that create a demand, everyone needs a tanker, so they have to fly extra if they are done with their currency, something to do with a horse blanket.

Many have tried advance planning, but is just not realistic in our ops tempo.

[Guest EverettP]

First you need to narrow this topic down considerably. I would do this by choosing a command to start with, AMC is completly different than ACC, versus AFSOC, AFMC, etc. Each has completely different taskings and time frames. Within each command, things still vary considerably. Within AMC you have wide range of missions. Some with months of advanced planning: Ice, Nuclear, state department, Presidential, some with planning limited to hours: locals(training), alert, diverts. Of course how we train and how we fight are completely different from those processes (read any PME book to find out how).

So I would choose one area within a command and focus on that. As Slacker said, the only people that know what is going on, are actually currently in that job; requirements/SOP/etc change daily, not to mention a whole new adminstration will shake things up a bit.

If you want to check the box for your degree, choose something simple like deployments (years in advance) or state department moves.

You asked for it, so here goes, amazingly simplified version:

1) Someone somewhere decides they need something. If they need something blowed up, they go to the CFACC (Combined Forces Air Component Commander) who then tasks a service to blow it up, or if that services equipment is CHOPed (Change of Operational Control) to the CFACC, then he just chooses that himself (has there ever been a female CFACC?). After hours of bickering and complaining (B&C) about who is getting credit and who is not getting used enough, The CFACC lackies publish an ATO (air tasking order) so those individual peeps know where to go and what to blow up. Luckly ACC usually runs this, so they got it down and usually works pretty good for the rest of us. Mostly classified/OPSEC stuff, forget it.

If someone needs something Moved, they contact TRANSCOM, who decides what is worthy of movement by air, then train and truck, then boat. If it goes by Air, AMC gets a call, and so on. After the user finds out that his tank is going to take 3 months to get somewhere on a boat, there is plenty of B&C and days before the user is supposed to be in place, they now have a high enough need that they become priority and they launch alert birds from somewhere to pick their bitchin a$$es up. Or someone gets called off of leave to fly bottled water to kuwait. From there, AMC can choose who flies what, when the user finds out they are going on a C-5 theres plenty of B&C and some C-17 mission that was going to S America, diverts to Pope and flies to Qatar with $hitter trucks. Or some guy with a tube of caulk to put in a windsheild is flown from seattle to baghdad, by himself, with no other cargo.

Once it is tasked to a wing, all the neat/fun/interesting trips are sucked up by the reserves who claim it is the only time their guys can fly. The next set of good deals go to FGOs that claim that trip happens to fall in their 'fly break', when the aforementioned mission to SA turns into a 3 week Afghan shuttle mission, their wife is suddenly in 'the window' and they have to stay home and make a kid. Finally the great line pilot that lives to fly and hasn't seen his family in 4 months, walks in the sqd after flying a 36 hr day and is told to go home and Crew rest for tomorrows mission, for he is "the only one". End result: a mission that was known about 3 months ago, is scheduled on a friday afternoon for a sat launch.

2) Luckily Training is completely different. However there is plenty of B&C to go around. Some civilian who has no idea what a rudder pedal is, is put in charge of scheduling training missions. This person has a kingdom more powerful than Titus himself. However, he has no concept that the sun sets later during a certain part of the year, so when you ask for some night AR, you end up wearing sunglasses. Every single base in the AF does this completely differently and their way is the correct way, the only common piece is a horse blanket; I have no idea what this is, but you sound important when you say it and somehow a tanker shows up when you need it.

3) Have you been paying attention? You have no idea... engineering management? This is Phd level work spread over a decade among dozens of Studs, and even then it will solve the problem of today, not for tomorrow. They have brought in experts from all over the world, nothing helps. Choose something simple, like how to air-refuel UAVs, how to safely move nukes, how to find bin laden.

How to find Bin Laden sounds like a good topic :) Like I said before, the whole point right now is to find a topic and if it's too complex (which this sounds like it may be) I'll focus somewhere else. Right now I'm just trying to do some fact finding and adjust from there. I'm not trying to get in over my head at all, but I would like to do something of value.

Many have tried advance planning, but is just not realistic in our ops tempo.

The heuristic is designed to be a dynamic model that is easily adjusted to a changing situation

Thanks again for the thorough responses

Edited January 21, 2009 by EverettP

[WeMeantWell]

The heuristic is designed to be a dynamic model that is easily adjusted to a changing situation

Unfortunately, the changing situation is so complex, that once a program is designed to address 'everything', everyone loses flexibility. I bet things worked great before computers. Pucks on a board is great way to schedule, because the program is a 'human'. Every other program is a compatability or interface program between humans. Yes, there are tools that help us, but none great enough to handle the dynamics of the military. A program like yours is used everywhere, UPS, Fedex, for Logistics, etc, the problem is, one day FedEx is going to decide to move everything via motorcycle, because that would be financialy ridiculous... the military doesn't care, if that is an effective way to move crap, the cost be damned.

I think you really have to limit the scope of this to make it work. (just check the box and come see how it really is)

[Dupe]

I've now seen TIMS, PEX, and COOL. COOL is the best, but I've ultimately seen each software product 'bypassed' with magnetic pucks and dry-erase markers. I'm entirely convinced that 'big board' sheduling is the best way to see the schedule. Maybe someday, large touch screens will be cheap enough for this purpose.

[Guest Safe&Clear]

What you should really research is how many MILLIONS of $$$$$ the military-- every branch!! Real world and training commands!!-- have spent on scheduling software, and how many years (10-plus in the case of PEX!) it takes the developers (Lockheed in the case of TIMS) to get it in use!

And of course when they do come online, they work, well... so so. Lots of potential in this enterprise, but in the end we always end up returning (although not "officially") to magnetic puck-boards and Excel spreadsheets to actually get the job done.

The military never holds a contractor's feet to the fire, despite what you may think. So we spend a hundred times the money to get half the product that a corporation (JetBlue, Southwest, etc.) could practically buy off the shelf...

[Guest bikerdood]

I've now seen TIMS, PEX, and COOL. COOL is the best, but I've ultimately seen each software product 'bypassed' with magnetic pucks and dry-erase markers. I'm entirely convinced that 'big board' sheduling is the best way to see the schedule. Maybe someday, large touch screens will be cheap enough for this purpose.

Reminds me of a funny story one of my DOs told me.

At base X, the OG spent all this money on big plasmas because someone decided that they need them for scheduling boards. So, when they arrived, he mandated that they be used by all schedulers.

A scheduler, who loved his magnetic puck board, wasn't happy, but he replaced his puck board with the plasma. He moved the puck board into an adjacent office and set up a web cam hooked up directly to the plasma.

Looking at the plasma, you'd think it was just some sort of fancy scheduling program until the scheduler (back in the office) decided to make a change and you'd see an arm reach up on the plasma and move the pucks around.

Puck boards really are the best in some situations.

-BD

[slacker]

Once it is tasked to a wing, all the neat/fun/interesting trips are sucked up by the reserves who claim it is the only time their guys can fly.

Don't hate the player...hate the game.

[Toro]

A scheduler, who loved his magnetic puck board, wasn't happy, but he replaced his puck board with the plasma. He moved the puck board into an adjacent office and set up a web cam hooked up directly to the plasma.

Genius.

[Guest EverettP]

Just another clarification, I'm not making some type of computer program to do this for you, I would just be applying a heuristic (rule of thumb) to scheduling based on dynamic priorities versus a deadline based schedule which most are.

I think I have enough info to get this going in a better direction now. Once I get more clarification and learn what the adviser expects I'll touch base back here.

A scheduler, who loved his magnetic puck board, wasn't happy, but he replaced his puck board with the plasma. He moved the puck board into an adjacent office and set up a web cam hooked up directly to the plasma.

Edited January 21, 2009 by EverettP

[WeMeantWell]

Don't hate the player...hate the game.

Oh, I don't Hate, I love it and can't wait for my chance...

[Container STS]

As a scheduler and an IE in management, I can tell you that there is no process to anything in scheduling. It is human hand-phucked every 2 seconds by an ADO who thinks he knows scheduling, the OG who has no idea how line squadrons schedule, and a wing commander that wants some left-field procedure from TAC/SAC days implemented because he thinks it can help the line squadrons out. But since the OG doesn't know, we get ass-raped in scheduling.

Then you have MX who doesn't want to buy the late takeoff/CNX because that affects his production numbers, likewise with the OG, since the wing commander is going to drop a ton of shit on whoever is the worst that week. Then MUNS does not want to CNX because they just uploaded a ton of shit.

You can't evaluate a process with the amount of human finger-screwing that goes on in scheduling.

Edited January 22, 2009 by Container STS

[disgruntledemployee]

While you're at it, might as well take on similar epic problems like world hunger and a non-biased news media.

BTW, we love using the pencil/clipboard and the puck board. It's a reminder that we can do our job without computers.

And another thing, why is it that so many commanders think there is an increase in efficiency/effectiveness if we have lots of plasma screens to show our crap. All that means is double the work.

And lastly EverettP, I imagine some dude has probably already done a thesis on such a topic. At least check out the Air University pages for similar papers/theses.

Out...and good luck.

[Guest EverettP]

While you're at it, might as well take on similar epic problems like world hunger and a non-biased news media.

BTW, we love using the pencil/clipboard and the puck board. It's a reminder that we can do our job without computers.

And another thing, why is it that so many commanders think there is an increase in efficiency/effectiveness if we have lots of plasma screens to show our crap. All that means is double the work.

And lastly EverettP, I imagine some dude has probably already done a thesis on such a topic. At least check out the Air University pages for similar papers/theses.

Out...and good luck.

I don't intend to solve the problem necessarily. I want to compare current methods to a somewhat related scheduling heuristic and basically see if it works. I may have to distill the problem down and make some assumptions to make it a little easier to digest, however, since both methods would have similar assumptions and "distillation" the end result in the effectiveness should relate closely to whether or not it could apply to the real world. If it doesn't work... that's fine, it'll just answer a question for a bigger problem and determine whether it should be looked into a bit deeper or not.

Thanks again everyone for the input.

[ol-IEWO]

Flight scheduling is dynamic. It is affected by events over which you have no control and may as well be considered random. Once you get everything scheduled, a flight lead is DNIF for a head cold; the wx goes down and it's below the FNG's minimums; the tanker aborts for some unknown reason. As you move people around to cover these holes, you are affecting flights occurring later in the day. A thunderstorm over the base stops aircraft refueling so take-offs are delayed and a pilot in the flight can't make a sim in the afternoon. Tomorrow's schedule is affected by the events of yesterday and today. In my years of flying fighters, I never knew more than 30 hours in advance what I would be doing, generally it was about 13 to 24 hours. In most squadron's the schedule was approved and published by 16:30 for the next day.

As Biaspoint said, pick a specific mission and pick it carefully. Good luck.

[Guest EverettP]

Flight scheduling is dynamic. It is affected by events over which you have no control and may as well be considered random. Once you get everything scheduled, a flight lead is DNIF for a head cold; the wx goes down and it's below the FNG's minimums; the tanker aborts for some unknown reason. As you move people around to cover these holes, you are affecting flights occurring later in the day. A thunderstorm over the base stops aircraft refueling so take-offs are delayed and a pilot in the flight can't make a sim in the afternoon. Tomorrow's schedule is affected by the events of yesterday and today. In my years of flying fighters, I never knew more than 30 hours in advance what I would be doing, generally it was about 13 to 24 hours. In most squadron's the schedule was approved and published by 16:30 for the next day.

As Biaspoint said, pick a specific mission and pick it carefully. Good luck.

Everything you just said give's me hope that this is a decent direction. The whole idea of this method is to schedule and be able to adapt to a dynamic environment, just like you mentioned. Like I said before, I'm not trying to solve a problem per se, I'm just trying to see if this scheduling method has the possibility of being applicable to this situation and if so, would eventually require further research into the topic.

[Yossarian]

Howdy gents.  Getting ready to go back to old school scheduling w/Puck Boards.  What are the actual magnetic pucks that were roughly 3/4" x 2.5" originally called?  I'm trying to order the magnetic "????s" and have no clue what to type into the Google.  For those born after 1990 these are the pucks that have a slot/lip (sts) on the top/bottom for laminated names to slide in/out of (sts).