International Route Scheduling Best Practices

[JJP]

Since starting AWS, I have exclusively concentrated on domestic hubs.  However, the last two games I played, I tried international hubs (Miami/Newark).  Although starting well, I bombed out of both games.  I have a suspicion it had to do with how I scheduled the aircraft.  I'd have to say that my aircraft utilization (in general) was quite poor.

So, I wanted to find out, from you international route veterans, your best scheduling tips.  How do you normally schedule flights?  How do you get more daily use from the aircraft?    What's the max range you should go for an international route?  What are the major do's and don'ts?

Your help would be greatly appreciated.

Thanks!

[slither360]

7 day scheduling ftw

[ucfknightryan]

Since starting AWS, I have exclusively concentrated on domestic hubs.  However, the last two games I played, I tried international hubs (Miami/Newark).  Although starting well, I bombed out of both games.  I have a suspicion it had to do with how I scheduled the aircraft.  I'd have to say that my aircraft utilization (in general) was quite poor.

So, I wanted to find out, from you international route veterans, your best scheduling tips.  How do you normally schedule flights?  How do you get more daily use from the aircraft?    What's the max range you should go for an international route?  What are the major do's and don'ts?

Your help would be greatly appreciated.

Thanks!

With international routes it's almost always worth your while to use 7-day scheduling (see attached screenshot).  It's not much more work for international aircraft (as opposed to domestic aircraft, where I don't fool with it any more except occasionally for coast to coast US flights). 

How far you should fly depends a lot on what aircraft you're using.  Generally, the larger the aircraft, the further you can get away with flying it.  For example, in my experience you don't fly the 762ER to it's max range because if fuel is anything other than cheap you will lose money.  But those same 6500-7000NM routes are profitable with something in the 777 size range

Were you one of my fellow competitors at MIA in the current Jet Age game?

-Ryan

[attachment expired]

[JJP]

With international routes it's almost always worth your while to use 7-day scheduling (see attached screenshot).  It's not much more work for international aircraft (as opposed to domestic aircraft, where I don't fool with it any more except occasionally for coast to coast US flights).  

By this, do you mean a separate route number for each day of the week?  Is this in order to get optimal spacing between flights in order to add extra routes to other destinations (I see you have squeezed in a couple shorter flights here and there)?

I had some international flights wherein I had quite a gap between flights, but it never seemed to be quite a large enough gap to place flights to other routes.  These were some 777s I had which didn't seem to turn much of a profit unless I put them on longer routes (4k - 6k nm).  Suggestions for this type of situation?

How far you should fly depends a lot on what aircraft you're using.  Generally, the larger the aircraft, the further you can get away with flying it.  For example, in my experience you don't fly the 762ER to it's max range because if fuel is anything other than cheap you will lose money.  But those same 6500-7000NM routes are profitable with something in the 777 size range

Just for fun and out of curiosity (and to say I did it!), I purchased (not leased) 2 A380s.  It was cool, but they didn't make jack for profit.  There was profit, but nothing special (no better than my leased 777s).  I was flying (IIRC) Newark to Beijing.  I thought I was going to rake in the dough becuase ticket prices were relatively high, and I had no lease cost.  I wonder what happened there?  Ideas?

Thanks for your help!

[ucfknightryan]

By this, do you mean a separate route number for each day of the week?  Is this in order to get optimal spacing between flights in order to add extra routes to other destinations (I see you have squeezed in a couple shorter flights here and there)?

I had some international flights wherein I had quite a gap between flights, but it never seemed to be quite a large enough gap to place flights to other routes.  These were some 777s I had which didn't seem to turn much of a profit unless I put them on longer routes (4k - 6k nm).  Suggestions for this type of situation?

If you schedule each day of a flight to a separate aircraft, you usually can have no more than the minimum necessary space between flights.  The one shorter flight on that schedule in the screenshot was just to fill the dead space after there was no room to schedule another long haul flight on the rotation.  Occasionally shorter flights will make it into the rotation because I can't find a destination where the arrival/departure times will be both during the travel day, but usually the only time I fly anything that isn't 3000NM+ on a long haul aircraft is to fill leftover space if I can't arrange the long-haul flights to avoid having said leftover space. 

With regards to the 777, I wouldn't fly it on anything shorter than 4000NM anyways unless I was at an airport with no slots, and therefore didn't have to worry as much about somebody sticking 2x as many 757s on the route and clobbering you.  How much and what kind of competition did you have on those shorter routes where you couldn't make money with the 777s?

Just for fun and out of curiosity (and to say I did it!), I purchased (not leased) 2 A380s.  It was cool, but they didn't make jack for profit.  There was profit, but nothing special (no better than my leased 777s).  I was flying (IIRC) Newark to Beijing.  I thought I was going to rake in the dough becuase ticket prices were relatively high, and I had no lease cost.  I wonder what happened there?  Ideas?

Thanks for your help!

I've never operated the A380 so I can't speak about that from experience, but I would suspect the competition again.  Frequency is king in this game, so smaller aircraft usually have the advantage, until the flight becomes long enough that they don't carry enough passengers to pay the overhead.  I don't think I'd even look at the A380 unless I was slot constrained or was going to fly a bunch of extremely long routes.

[JJP]

If you schedule each day of a flight to a separate aircraft, you usually can have no more than the minimum necessary space between flights. 

But, isn't there only a certain "window of opportunity" you have to schedule flights (depending on the range) in order to not make it a flight that lands or takes off during "red eye" times?  For example, I seem to remember my flights to Europe had to take off from the East Coast some time in the afternoon to arrive at a decent hour in Europe (and to arrive back in the States at a decent hour).  This would leave me to believe that you can only shift your flight times so much.  Maybe I'm making more out of this than it is. 

Also, I don't like to spend hours on end setting up routes.  It seems to me that setting up these optimal routes is possible but takes a lot of real life time to implement. 

With regards to the 777, I wouldn't fly it on anything shorter than 4000NM anyways unless I was at an airport with no slots, and therefore didn't have to worry as much about somebody sticking 2x as many 757s on the route and clobbering you.  How much and what kind of competition did you have on those shorter routes where you couldn't make money with the 777s?

I started to get a lot of competition from the one main guy at my hub.  He filled the routes to capacity, and so did I.  Yikes!

I've never operated the A380 so I can't speak about that from experience, but I would suspect the competition again.  Frequency is king in this game, so smaller aircraft usually have the advantage, until the flight becomes long enough that they don't carry enough passengers to pay the overhead.  I don't think I'd even look at the A380 unless I was slot constrained or was going to fly a bunch of extremely long routes.

At first, I had no competition.  IIRC, I made less than $1 million profit/wk each on these birds which only compared modestly to what I was making on my leased birds to Europe!   With competition, it was a losing cause.

[ucfknightryan]

But, isn't there only a certain "window of opportunity" you have to schedule flights (depending on the range) in order to not make it a flight that lands or takes off during "red eye" times?  For example, I seem to remember my flights to Europe had to take off from the East Coast some time in the afternoon to arrive at a decent hour in Europe (and to arrive back in the States at a decent hour).  This would leave me to believe that you can only shift your flight times so much.  Maybe I'm making more out of this than it is. 

Also, I don't like to spend hours on end setting up routes.  It seems to me that setting up these optimal routes is possible but takes a lot of real life time to implement. 

usually you can start at one side of that window and the flights will slide across it, and you can also alternate flights going east and west, as times that usually don't work to Europe (from NA) usually work going to Asia.  Usually it doesn't take too long to find destinations that work.

[LemonButt]

7-day scheduling isn't just for long haul routes--it's for all routes if you are looking to utilize your fleet as much as possible.  I fly 100% domestic and 100% BAC 1-11 and use 7-day scheduling on all my planes.  Yes, it can be tedious, but you only have to do it once and after you've done it a few times you get really quick at it (I go to an already scheduled aircraft and open new tabs for every route on a schedule and then add new route like this and then change the day of the week to create new routes).  As a result, I have one of the top fleet utilizations in DOTM2 with 16.6 hours--and this is only flying routes under 1500nm!  To answer your question about flight times and hitting desired windows, you can see from the screenshot that I schedule longer flights for overnight travel.  I can leave Chicago and fly west at 2000 and take off from the west coast by 2300 for a redeye back to Chicago where it lands at 400 or 500 local time and I can then schedule my next flight early in the morning during rush hour.  This way, I never miss a beat.  Worst case scenario I can fly to the east coast at 2300, land around 300, take a 2 hour turnaround (versus 1 hour) and takeoff from the east coast during morning rush hour.

The end result is I can fly ~1-5 extra flights/week that I otherwise not be able to schedule and the gross margins basically go straight to my bottom line.  All of my planes have 20-25 routes each:

[attachment expired]

[ucfknightryan]

7-day scheduling isn't just for long haul routes--it's for all routes if you are looking to utilize your fleet as much as possible.  I fly 100% domestic and 100% BAC 1-11 and use 7-day scheduling on all my planes.  Yes, it can be tedious, but you only have to do it once and after you've done it a few times you get really quick at it (I go to an already scheduled aircraft and open new tabs for every route on a schedule and then add new route like this and then change the day of the week to create new routes). 

Heh, you've obviously never upgraded a fleet of airplanes with 7-day scheduling for domestic flights.  I've done it, twice.  First BAe 146s to 737-500s and then A320s to CSeries.  It makes the tedium of the initial scheduling pale by comparison.  After those experiences I have a hard time convincing myself that it's worth the effort.

[JJP]

Holy Toledo, Lemon.  That's nuts!  I think I'd agree with UcfKnight on the change over.  I would go insane!

[LemonButt]

Holy Toledo, Lemon.  That's nuts!  I think I'd agree with UcfKnight on the change over.  I would go insane!

I leased the planes new and plan on buying them eventually and flying them the entire gameworld   There are several ideas for 7 day schedule in feature requests to make it more painless, but until then it will remain tedious.  I guess you could say I'm obsessed with efficiency.

[ucfknightryan]

I leased the planes new and plan on buying them eventually and flying them the entire gameworld   There are several ideas for 7 day schedule in feature requests to make it more painless, but until then it will remain tedious.  I guess you could say I'm obsessed with efficiency.

I planned on flying the BAe 146s the entire scenario as well, then they closed the damn production line on me while I had no money to shell out to keep it open.  Then I had to choose between operating a split fleet and replacing all the damn things.  Since this was under v1.11 I could actually get the 735s very quickly, and eliminating a fleet was appealing since I'd been operating on paper thin margins for quite a number of years.  In retrospect I should have passed and just concentrated on buying all the 146s.

[knobbygb]

JJP,

One or two steps down from 7-day scheduling - and a good introduction I suppose are two other methods I use:

For flights that take around 21hrs to 23hrs (including both turnarounds), you can 'slip' the times by an hour or so each day so that all those 1 or 2hrs gaps combine to let you fly a daily service. So, for example, depart on an eastbound transatlantic at 23:00 on day 1, 22:00 on day 2 etc.  To be slightly more sophisticated, you can do this in two 'blocks' (sorry, this is difficult to describe and screenshots wouldn't help).  Here's one of my schedules for TXL-YUL:


721/722   26   TXL-YUL   3239 nm   18:30 - 13:25   20hr50
719/720   37   TXL-YUL   3239 nm   15:30 - 10:25   20hr50
717/718   41   TXL-YUL   3239 nm   12:30 - 07:25   20hr50
715/716   5   TXL-YUL   3239 nm   11:05 - 06:00   20hr50
723/724   6   TXL-BCN   815 nm   08:00 - 15:15   9hr10
(Maint. Tue. 09:00 - 14:00)
Boeing 757-200, turn times 1hr55 (1%)

As you see, it only involves four 'flights' being set up, not seven on the YUL run and there is time for both maintenance and and extra leg (to BCN) once a week.  This is on a flight where there isn't time to send the a/c daily at the same time.

Another idea is to combine just two (or three) destinations (instead of five or six) on your 7-day rotating schedule like this:

1099/1100   1   ITM-KBL   3204 nm   15:20 - 09:30   19hr50
1127/1128   2   ITM-KOJ   297 nm   11:20 - 15:50   6hr10
1095/1096   2   ITM-KBL   3204 nm   18:00 - 12:10   19hr50
1097/1098   3   ITM-KBL   3204 nm   13:50 - 08:00   19hr50
1107/1108   4   ITM-AUH   4150 nm   09:45 - 08:20   24hr15
1101/1102   5   ITM-AUH   4150 nm   10:15 - 08:50   24hr15
1103/1104   6   ITM-AUH   4150 nm   10:30 - 09:05   24hr15
1105/1106   7   ITM-AUH   4150 nm   10:45 - 09:20   24hr15
(maint. Tue. 10:15 - 15:15)
Vickers Super-VC10, turn times 1hr40 (1%)

1117/1118   1   ITM-AUH   4150 nm   10:15 - 08:50   24hr15
1119/1120   2   ITM-AUH   4150 nm   10:35 - 09:10   24hr15
1121/1122   3   ITM-AUH   4150 nm   10:55 - 09:30   24hr15
1109/1110   4   ITM-KBL   3204 nm   14:55 - 09:05   19hr50
1125/1126   5   ITM-KOJ   297 nm   11:00 - 15:30   6hr10
1111/1112   5   ITM-KBL   3204 nm   18:05 - 12:15   19hr50
1113/1114   6   ITM-KBL   3204 nm   13:55 - 08:05   19hr50
1123/1124   7   ITM-KBL   3204 nm   14:15 - 08:25   19hr50
(maint. Sun. 09:10 - 14:10)
Vickers Super-VC10, turn times 1hr40 (1%)

So I basically have two aircraft covering seven days a week to two destinations, where there would only be time for six to AUH and seven to KBL (using the first method above), with maintenance time and an extra leg thrown in.  

Either of these methods might be a good introduction to 7-day scheduling without the risk and work of having to obtain seven large aircraft to practice on.

Hope this helps.

[JJP]

Wow, that's cool, Knobby.  That's the kind of info I was looking for.  The only thing: doesn't it take a LOT of time to figure out just the right combo to fill up all the plane's time? 

[LemonButt]

Wow, that's cool, Knobby.  That's the kind of info I was looking for.  The only thing: doesn't it take a LOT of time to figure out just the right combo to fill up all the plane's time? 

No.  You can basically figure out which routes would max out a schedules utilization by it's mileage.  If you are looking for a route of a specific length, simply take a route of x nm and schedule it for 0:00 and you can easily see how many hours (plus turnaround time at the end) it takes to fly the route round trip.  If it's too long or short, adjust the mileage accordingly until you find a viable route.

[knobbygb]

You can basically figure out which routes would max out a schedules utilization by it's mileage

Almost but not always exactly. Routes into large airports tend to take longer due to taxi time etc.  This isn't a big issue with long haul but with short haul, say, a 200nm flight into ATL or ORD may take the same time as a 300nm flight to Nowheresville.  I don't have a specific example to hand but it CAN make a difference. Also, on long haul the winds can make a difference. I've noticed east/west return trips can take longer than those that are the same distance but are predominantly north/south (or is it the other way round?).

Don't worry TOO much about maxing out the schedule. the NUMBER of trips is most important.  If you fly, say, two round trips per day then it doesn't really matter if they take 20hr + 4hrs = 24hr or 19hr+3hrs = 22hrs - you'll probably make the same profit on either.  Of course - fitting that elusive third route into the schedule makes a HUGE difference.

And yes, doing this scheduling CAN take a long time.  You get quicker with practice and it becomes almost instinctive (which is what I think LemonButt sorta means). I like to think of it as getting good value for my credits on a Tuesday night when there's nothing on TV and I've no beer money. Honestly, I promise that I absolutely NEVER get up at 5am on a Sunday because I have an aircraft due and I need to get it's rotational schedule sorted while slots are available. Wouldn't ever even consider it. Honest.  

Give it a try and let us know how you go on.  I promise it's MUCH more satisfying to "beat the system" and fly a route it at first appears you can't.

[LemonButt]

Almost but not always exactly. Routes into large airports tend to take longer due to taxi time etc.  This isn't a big issue with long haul but with short haul, say, a 200nm flight into ATL or ORD may take the same time as a 300nm flight to Nowheresville.  I don't have a specific example to hand but it CAN make a difference. Also, on long haul the winds can make a difference. I've noticed east/west return trips can take longer than those that are the same distance but are predominantly north/south (or is it the other way round?).

Don't worry TOO much about maxing out the schedule. the NUMBER of trips is most important.  If you fly, say, two round trips per day then it doesn't really matter if they take 20hr + 4hrs = 24hr or 19hr+3hrs = 22hrs - you'll probably make the same profit on either.  Of course - fitting that elusive third route into the schedule makes a HUGE difference.

And yes, doing this scheduling CAN take a long time.  You get quicker with practice and it becomes almost instinctive (which is what I think LemonButt sorta means). I like to think of it as getting good value for my credits on a Tuesday night when there's nothing on TV and I've no beer money. Honestly, I promise that I absolutely NEVER get up at 5am on a Sunday because I have an aircraft due and I need to get it's rotational schedule sorted while slots are available. Wouldn't ever even consider it. Honest.  

Give it a try and let us know how you go on.  I promise it's MUCH more satisfying to "beat the system" and fly a route it at first appears you can't.

East/West are the jet stream/tradewinds, which are stronger towards the equator (created by the earth's rotation).  Assuming you aren't flying into a different latitude zone (I think there's only 4 in each hemisphere in real life, but not sure how it's modeled in AWS) where the winds change, it's a non-factor   An ideal strategy if you fill up the major routes at a base is to determine how many miles you can fly in a given day assuming x number of routes per day.  For example, if you have a regional jet and can fly 4000nm/day (2000nm each way) across 3 routes (6 turnarounds), you can simply add up the routes you want to schedule to see if it will fit.  You might have 500/500/1000 or 250/250/1500.  Either way the numbers should be accurate enough to figure out which routes are possible and which ones aren't.  You may also want to build in an extra 5-10 minute turnaround time so if you need to shave a few minutes off a route, you can.

[knobbygb]

The way I actually do it for a typical medium/short haul a/c like the 737 is...

First decide how many routes per day each aircraft should fly - for a 737 it's typically three.
List in Excel ALL the destinations I'd like to fly from my base and sort them in to distance order (remembering to list routes multiple times if I need to fly them more than once)
For each aircraft...
Take the one from the top of the list (longest)
Take the one from the bottom of the list (shortest)
Take one from the middle of the list that then fills the schedule out as much as possible.
Cross those three of the list and repeat...
As the final one or two aircraft are delivered, swap a few of the routes about to make sure everything fits and everything is covered as much as possible.

In this way, with a lot of planning, it's possible to open up at a new base, with the appropriate aircraft about to arrive and cover everything efficiently within just a few game days.  The planning really pays off.

[LemonButt]

The way I actually do it for a typical medium/short haul a/c like the 737 is...

First decide how many routes per day each aircraft should fly - for a 737 it's typically three.
List in Excel ALL the destinations I'd like to fly from my base and sort them in to distance order (remembering to list routes multiple times if I need to fly them more than once)
For each aircraft...
Take the one from the top of the list (longest)
Take the one from the bottom of the list (shortest)
Take one from the middle of the list that then fills the schedule out as much as possible.
Cross those three of the list and repeat...
As the final one or two aircraft are delivered, swap a few of the routes about to make sure everything fits and everything is covered as much as possible.

In this way, with a lot of planning, it's possible to open up at a new base, with the appropriate aircraft about to arrive and cover everything efficiently within just a few game days.  The planning really pays off.

I do this to an extent as well, but I don't list the distance of the flights--I list the total route time.  I also note how many flights per day it would take to satisfy the demand.  If I have a flight with ~500 pax/daily that is at least 5 hours, I can successfully schedule five 100 seaters flying that route once/day while incorporating A-checks on -234-67 so that I have extra supply on Monday/Friday for the heavy routes.  For routes that have 2 or 3 flights/day you can simply put A-checks on -----67.  By doing this, you can ensure a destination with only 1 flight/day doesn't get bumped from 7 days/week to 6 days/week due to maintenance.

[forsberc]

(I think there's only 4 in each hemisphere in real life, but not sure how it's modeled in AWS)

Just to clarify the weather aspects. I apologize for this simple tangent.

There are three atmosphere cells in each hemisphere. The cells are (Equator to Pole) Hadley, Ferrell, and Polar. Between these cells, various jet streams occur- i.e. between the Polar and Ferrell cells, the polar jet exists. There is sinking air above both poles, in essence creating a boundary, but because the Polar cells surround this sinking region, it's not really counted as a boundary.

My very simple illustration depicting where the cells exist:

\     Polar (~60˚-90˚)
  \   Ferrell (~30˚-60˚)
   |  Hadley (~0˚-30˚)
   |  Hadley
  /   Ferrell
/     Polar

The boundary between the various cells are called:
Polar & Ferrell: Polar Jet -> Generally associated with rising air and rain. One of the main reasons for forests and lush lands around 60˚N/S across the world

Ferrell & Hadley: Sub Tropical Jet -> This is typically associated with sinking air, just doesn't develop a lot of rain. One of the main reasons why deserts exist around 30˚ N/S across the world

Between both Hadleys: Inter Tropical Convergence Zone (ITCZ) -> Associated with VERY strong updrafts and almost constant thunderstorm (and rain) activity. Aka: Rain forests and 'tropical' weather.



Tip: I remember the cell names because the actor, Will Ferrell, lives in the Ferrell Cell (not named for him ). The polar cell makes sense, leaving the Hadley cell at the equator.

Hope that it all made some sense.

Cameron

PS: In order to regain the thread's course, I generally use a 7 day rotation when doing LH Intl' Operations. Though, when I can fit a simple daily schedule on a LH route, I will probably do it, merely out of laziness.