Jump to content

Photo

Itineraries, Flights, and Routes (and Real Connecting Pax)

* * * * * 5 votes AE 4.0

  • Please log in to reply
155 replies to this topic

#81
Sheepy

Sheepy

    N/A

  • Member
  • 1,935 posts

User's Awards

        

If a 'one stop flight' changes flight number, it's no longer a one stop flight. It defeats the whole definition and advantage of such a flight.


Administrator of UnitedSkies alliance

and also a member of some other ones, but they're 2vip4u


#82
LLC

LLC

    AE Luver

  • Member
  • 461 posts

If a 'one stop flight' changes flight number, it's no longer a one stop flight. It defeats the whole definition and advantage of such a flight.


technically it is: if you are a passenger and you book on both flights (on the same aircraft) you would not change planes at the hub city

(unless required to due to immigration control - in that case you get off the aircraft , go through immigration control then get back on the aircraft

this happens when there are no immigration facilities at the destination [3rd] airport - A to B Intl, B to C Domestic)



#83
Sheepy

Sheepy

    N/A

  • Member
  • 1,935 posts

User's Awards

        

As a general rule, if an aircraft changes flight number, all passengers will be asked to disembark the aircraft.

A one, or more, stop flight is generally any set of routes that are operated by one plane under a single flight number.


Administrator of UnitedSkies alliance

and also a member of some other ones, but they're 2vip4u


#84
Andy-J

Andy-J

    probably been on here too many years now....oh well

  • Member
  • 158 posts

User's Awards

2   

I really dislike this entire concept.

For example, a couple of months ago my uncle received Virgin America (no coincidence of why I named my airline after it..) at the American-controlled D/FW airport. He really likes them, and hates paying majority prices, so on his trip to San Jose del Cabo, this option was unavailable. It makes me mad. Plus, wouldn't it be hard for huge airlines, ie. even just 100 routes or more to try to assemble connecting routes? If you have over 1000 routes, is it even logical to try to make yourself connect it all, when A-E is perfectly capable of doing it for you? One of the reasons I chose A-E was its simplicity and I plan on supporting keeping it that way.


-------Andy-J

#85
Sheepy

Sheepy

    N/A

  • Member
  • 1,935 posts

User's Awards

        

The many thousands of possible itineraries will indeed create an issue that'll have to be worked around somehow. :/


Administrator of UnitedSkies alliance

and also a member of some other ones, but they're 2vip4u


#86
Yuxi

Yuxi

    AE Developer

  • AE Developer
  • 4,362 posts


Perhaps pax offering the highest yield could fill the plane first, based on the amount they pay per mile?

Say pax A is competing with pax B for a seat on AAA-BBB.
Pax A is flying CCC-AAA-BBB, a total distance of 300 miles, and has paid $300 for the privilege. Thus, Pax A is paying $1 per mile.
Pax B is flying DDD-AAA-BBB, a total distance of 600 miles, and has also paid $300 for this. Thus, pax B is paying 50c per mile.
Pax A would have priority, as pax A is paying more per mile. PAX B would presumably have to find another airline.


Alternatively, to simulate the somewhat unpredictable nature of pax, priority could be completely randomized.

 

Now you're getting into airline yield management. Perhaps we can make it so the player's primary job is to maximize revenue over the entire network using different controls, policies, and product offerings. It's a complicated subject indeed... 

 

In reality, tickets sold on connecting routes have to be priced quite a bit below tickets that are non-stop in order to garner traffic. Would passengers be able to differentiate between a non-stop and connect carrier for a specific O&D and take price into consideration when deciding which carrier to fly with? For this to work, non-stop carriers would have to be allowed to take a price-premium against connect carriers and not lose all market share.

 

Source: This is what I do for a living.  :lol:

 

Yes, that would be one of the primary benefits of this model. For example, the LAX-JFK market would be open to much more competition from airlines offering 1-stop itineraries for cheaper than nonstop options. Otherwise, not differentiating O&D demand from flight traffic data would severely overstate total demand being simulated (like now).



#87
Yuxi

Yuxi

    AE Developer

  • AE Developer
  • 4,362 posts

The many thousands of possible itineraries will indeed create an issue that'll have to be worked around somehow. :/

 

The only way I can see this not overwhelming the user is organizing operations by markets (end-to-end city pairs), not individual itineraries. You would then drill down into a market and manage your inventory and pricing in that scope. For networks with a large number of connection combinations, this would help condense the structure to a more manageable form.



#88
Yuxi

Yuxi

    AE Developer

  • AE Developer
  • 4,362 posts

The many thousands of possible itineraries will indeed create an issue that'll have to be worked around somehow. :/

 

Using constraints to remove non-sensical connections based on timing and travel preferences (no connections longer than X hours, total travel distance cannot be longer than some multiple of the direct origin->destination distance, ...) may whittle down the number of itineraries to a more manageable number... :S

 

I'm open to ideas on this one :P



#89
LLC

LLC

    AE Luver

  • Member
  • 461 posts

Using constraints to remove non-sensical connections based on timing and travel preferences (no connections longer than X hours, total travel distance cannot be longer than some multiple of the direct origin->destination distance, ...) may whittle down the number of itineraries to a more manageable number... :S

I'm open to ideas on this one :P


I would prefer that the model be based on 2 and 3 stop flights more than connections (with connecting passengers at hubs working as it does now)

so if you want connecting passengers at your focus (non hub) city you would have to have all flights be at a minimum 1 stop flights:

for example:

Flt # 1: FLL - RDU - BWI
Flt # 3: FLL - RDU - PHL
Flt # 5: FLL - RDU - MDW

Flt # 7: MCO - RDU - PHL
Flt # 9: MCO - RDU - MDW
Flt # 11: MCO - RDU - BWI

Flt # 15: TPA - RDU - PHL
Flt # 17: TPA - RDU - BWI
Flt # 19: TPA - RDU - MDW

So in this example, in order to have 1 stop flights in each of  these city pairs:
(return flights not shown)

1x FLL - BWI  via RDU
1x FLL - PHL  via RDU
1x FLL - MDW  via RDU
 
1x MCO - PHL  via RDU
1x MCO - MDW  via RDU
1x MCO - BWI  via RDU
 
1x TPA - PHL  via RDU
1x TPA - BWI  via RDU
1x TPA - MDW  via RDU

would require 3 daily non-stops between each of these city pairs
(return flights not shown)

3x FLL - RDU
3x MCO - RDU
3x TPA - RDU
 
3x RDU - BWI
3x RDU - PHL
3x RDU - MDW

this is just a simplified example, it gets more complicated than this depending on how you route the aircraft



#90
Sheepy

Sheepy

    N/A

  • Member
  • 1,935 posts

User's Awards

        

So.

The lolhueg number of possible number of itineraries.

I think I've thought of a possible solution, but I doubt it'd be light on server load at all.

It combines not having to endlessly create thousands of itineraries. For example, a hub with 100 destinations gives 5050 possible connections. (100+99+98, etc.) 

Admittedly my maths could be wrong there, but it'd be a huge number.

So how do we make this manageable for the player?

No-one should be given an advantage because they can spend an hour a day adding 100 possible itineraries for each new destination they add. This will be true of any system where each itinerary must be manually added.

Therefore itineraries must be created at least semi-automatically.

Now how do you manage those itineraries?

Assuming the game creates every itinerary automatically, I'd suggest a 'mass itineraries with connections manage' screen.

Individual itinerary management on a separate page could create some issues.

First of all, if you open a new destination, it should auto-create the itinerary based on the prices of each non stop flight. (e.g. AAA-BBB-CCC would be charged at AAA-BBB and BBB-CCC).

This is far from the ideal revenue management system however. So on the mass itinerary page, you would have each itinerary with a connection on a new line, displaying PDEW on your flights, and total PDEW on that city pair. This could then be expressed as a percentage.

By pressing a button, more information about that itinerary could appear on the same page, such as connection time, etc.

It could also include an 'attractiveness score', ,which would be connection time/seat comfort/etc, being independent of price.

Now something important I think, should be that prices should be able to be changed from this main mass itinerary manage page. Probably as mass save.

 

Now this provides an advantage because prices are calculated automatically at first, so you could ignore the vast quantity of connections if you really wanted to.

 

Alternatively, you could use the attractiveness score to only focus your efforts on the itineraries likely to gain enough passengers to be worth your time.

 

It allows both laziness and customisation.

 

But yeah, it'd probably kill the server.

 

Note that what I've said above is only for itineraries with connections, I think the current system would be better maintained for non stop or possibly even direct flights.


Administrator of UnitedSkies alliance

and also a member of some other ones, but they're 2vip4u


#91
txaggie

txaggie

    AE Know It All

  • Member
  • 157 posts

I chimed in at the beginning.

 

I'll chime in again-

 

Ive always wanted to operate one of thsoe tiny regional 7 seater airlines...an island hopper in Alaska kind of thing, or the australian outback or perhaps just a 3 legged flight between two small american cities before hitting a major hub. .This requires two layers, flights and itineraries.

 

I fly from A to B to C to D to E to F to G (which perhaps is a hub) which also flies to H, I, J, K, L, M, and various other combinations such as A to D to E to G, etc..., etc... You fill the seats on the different segments via 100 different itineraries.

 

As it is now, you have to set up separate flights filling with their own demand between A and B, and B and C, and C and D, etc... Demand between two small towns is usually too low to sustain anything, and making them all hubs to enable the generation of connecting passengers also unsustainable.

 

That might make no sense at all, I wrote that in about 5 seconds.



#92
lakhbir2012

lakhbir2012

    AE Addict To-Be

  • Member
  • 12 posts

Love the idea..when is it going to be implemented into the game?



#93
Amadeus Asia CEO

Amadeus Asia CEO

    AE Player

  • Member
  • 27 posts

We should create our own timetables, in conjuction with connecting flights at our hub!


Kevin Soun

Amadeus Asia CEO,O2


#94
Yuxi

Yuxi

    AE Developer

  • AE Developer
  • 4,362 posts

We should create our own timetables, in conjuction with connecting flights at our hub!


That's the general idea, but the problem at hand is how to enable that level of control without overwhelming the player with micromanagement.

#95
LLC

LLC

    AE Luver

  • Member
  • 461 posts

Using constraints to remove non-sensical connections based on timing and travel preferences (no connections longer than X hours, total travel distance cannot be longer than some multiple of the direct origin->destination distance, ...) may whittle down the number of itineraries to a more manageable number... :S

 

I'm open to ideas on this one :P

 

Yuxi, you should base this on real life airline examples, so for example if there are 8 hour connections (between flights) in real life, then they should exist in AE4



#96
Northern

Northern

    Data Collector

  • Data Collector
  • 1,623 posts

User's Awards

2    2    4   
AE4

banner_signature_northern.png


#97
davedave

davedave

    AE Know It All

  • Member
  • 140 posts

User's Awards

      2    2   

That's the general idea, but the problem at hand is how to enable that level of control without overwhelming the player with micromanagement.

 

Can't you generalise it a bit? Look at the overall effects, rather than modelling in detail. So, for this example, players could have a checkbox for 'connect with other flights'. If it's checked, then there's a bit of a time penalty per-plane, to reflect the increased difficulty of scheduling within the constraints - that could depend on things like how many planes are expected to work to matching schedules, perhaps - and also a boost to connecting passengers. 



#98
ar157

ar157

    Resident Australian Arnimal

  • Member
  • 1,476 posts

User's Awards

     

Can't you generalise it a bit? Look at the overall effects, rather than modelling in detail. So, for this example, players could have a checkbox for 'connect with other flights'. If it's checked, then there's a bit of a time penalty per-plane, to reflect the increased difficulty of scheduling within the constraints - that could depend on things like how many planes are expected to work to matching schedules, perhaps - and also a boost to connecting passengers. 

Let's all make our own timetables ^_^ that'll put the big airlines off. hue hue hue.



#99
Sheepy

Sheepy

    N/A

  • Member
  • 1,935 posts

User's Awards

        

Making one's own timetables would be good.

And yeah, actually that could hit the lolspansion things hard.

So an 'alternative' thing you could do would be to reschedule for moar efficiency.


Administrator of UnitedSkies alliance

and also a member of some other ones, but they're 2vip4u


#100
ar157

ar157

    Resident Australian Arnimal

  • Member
  • 1,476 posts

User's Awards

     

Me thinks the CEO of Flunjace, however would not approve.







Also tagged with one or more of these keywords: AE 4.0

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users