AirwaySim
Online Airline Management Simulation
Login
Username
Password
 
or login using:
 
My Account
Username:
E-mail:
Edit account
» Achievements
» Logout
Game Credits
Credit balance: 0 Cr
Buy credits
» Credit history
» Credits FAQ

Author Topic: Connecting flights/transiting passenger  (Read 5705 times)

Online Sami

  • Administrator
  • Members
  • Posts: 14535
    • AirwaySim - Are you the next Richard Branson?
Re: Connecting flights/transiting passenger
« Reply #20 on: November 12, 2009, 09:48:34 PM »
Having truly connecting flights for all possibilities would likely require more processing power than all users can supply during a 30 min gameday.

That is one of the issues. With a 24h game day for example (= full real time) that wouldn't be a problem at all. But try to count all that data in just 30 mins.

(And some guys even wish to play something like 15 min days (= no way). :P)

Offline ali5541

  • Members
  • Posts: 486
Re: Connecting flights/transiting passenger
« Reply #21 on: November 13, 2009, 10:40:50 PM »
So whats the decision on this idea?
Top Airway Sim Fan, Member since July-2007

pruteanucristian

  • Former member
Re: Connecting flights/transiting passenger
« Reply #22 on: November 29, 2009, 12:54:47 PM »
I wouldn't mind if the game progressed a little slower (even 2-3 hours/day), if I were to get a system closer to reality. And connecting flights is something very important imo.

Offline lastchancer

  • Members
  • Posts: 142
Re: Connecting flights/transiting passenger
« Reply #23 on: November 29, 2009, 03:30:53 PM »
The ideal would be a system like the software every travel agency uses: it is feeded with all flight data and on every request it is showing a long list of possibilities. The client choses his flight out of the list acording to his own preferences like price, date, number of transfers airline etc.
In this game, the computer would have to generate millions of travellers ("Sims") with different preferences taking their choice. This means billions of calculations for the software!

Iīm not deep in programming, but Iīm sure this not something made in a week...

So, i think it would be the only way to have a fast solution to create connecting flights semi-manually. Maybe the system could help:
whenever you created a new route, the programme automatically shows you a list with flight starting in a timeframe after the arrival time of the new route.
It could be a popup, saying something like" Do you want to create connecting flight to xyz" and a check box. (A connecting flight from New York to Boston via Los angeles wouldnīt make sense!). If you click yes, the route is automatically created.
This functionality is close to the  technical stop function, so it should be easy to integrate?!
The connecting routes might also been shown as direct line on the map, maybe on a second layer.

Edit:
Ups, just found an answer somewhere else:
The number of route pairs goes up more than exponentially by this so before it can be even thought of the calculation system must be tested and partly re-optimized. Not a simple process.  (ie. if we'd have 110 000 route pairs now, with connections automatically this could go to millions, which is way beyond the design limits)
« Last Edit: November 29, 2009, 04:01:00 PM by lastchancer »

Offline TommyC81

  • Members
  • Posts: 62
Re: Connecting flights/transiting passenger
« Reply #24 on: January 17, 2010, 06:26:18 PM »
As I know the planning is in progress for the next version (1.2), I'd just like to make one more push for this feature.

jneil121

  • Former member
Re: Connecting flights/transiting passenger
« Reply #25 on: January 18, 2010, 12:03:40 PM »
Sorry I only just noticed this thread,  :P

Ok heres how I see the system working, much like the real world system with airlines.

Passengers booking online punch in their departure dates, destinations etc, normally direct flights are listed first, then connecting flights with the same airline or their partner airlines.

So maybe when inserting a new route you could include an option to choose which flights this new route will connect with. If they click Yes to 'Connect flight with;' they could select connecting flights from either a drop down menu or popup screen of available flights from that airport after the previous flights turnaround time is completed. (IE a 737 lands, the system will show all flights available 40 minutes after arriving, 747 it will be 120 minutes etc.) Maybe we could even make it more standard to just 90 minutes ore something round.

I was thinking maybe too connect with alliance members, could the database search for flights departing from airport with the alliance logo attached to the route? (IE in route planning alliance logo is attached to airline name, could database search for that?)

Just an idea, I dont know much about this sort of stuff, but its my opinion anyways.  :laugh:
« Last Edit: January 18, 2010, 12:08:23 PM by jneil121 »

Offline JumboShrimp

  • Members
  • Posts: 5992
Re: Connecting flights/transiting passenger
« Reply #26 on: June 21, 2010, 03:07:48 AM »
Any news on this subject?

This would be my #1 request for the game, in order to add realism.

With this addition, the game could be a lot more dynamic.  Instead huge 3000 static demand between large airports, the inherent demand of, say ATL - DFW should be a lot more modest, say 1000 or so, corresponding to local demand.  It would only grow to 3,000+ if passengers from connecting flights create this demand.

Ideally, the system should handle as much, or everything in the background.  If there is a route A-B and A-C, the system should check the quality of possible connection B-C based on timing of connecting flights.  Direct, non-stop B-C flight(s) would be perfect, and 100% of inherent B-C demand could be realized and turned into ticket purchases.  Less perfect connections, such as B-A-C would receive some scoring based on timing etc, and would receive less that 100% score, meaning less than 100% of the inherent B-C demand would turn into ticket purchases.

Now the system could do that all in the background, no changes in interface would be required for the bare bone functionality.

An are for enhancements would be ability to set prices (discounts) for connecting passengers and maybe limits of how may of the discounted transferring passengers the player is willing to take.

Another area to make the world more dynamic (if desired) would be a way for airports to grow based on demand.  If slots are sold out, airport could decide to expand and create more slots.  But this would be more of a departure from the Sim trying to mimic the real world.

OTOH, passengers transfers from connecting flights would make the Sim closer to the real world - and frankly, a lot more fun.

Offline Laama

  • Members
  • Posts: 143
Re: Connecting flights/transiting passenger
« Reply #27 on: February 21, 2012, 12:37:12 PM »
Bumping this because this would be really great to get, but to be able to fully use this feature it would also need the code-share agreements.

To be able to get the connecting passengers it could need some restrictions on the time spent on the airport by the connecting passenger. For example

Small airport  30min to 1 hour
Middle-sized airport 1 hour to 2 hours
Significant airport 1,5 hours to 3 hours
Large airport 2,5 to 5 hours

This is just so that the passengers won't be making turnarounds which would be impossible in real life or turnarounds which nobody (expect people who love airplanes) would be really making like staying 12 hours on an airport.

And why do I support the connecting passenger idea is, prolly the same as for everyone else. You may be flying from B to C with a 150 seater, but the demand is 120/day and the demand from A to C is 20/day so you could get more people on board if the schedule is reasonable.

radu.poenaru

  • Former member
Re: Connecting flights/transiting passenger
« Reply #28 on: February 21, 2012, 02:55:56 PM »
Just dropping my 2c for an easy way to model connecting flights.
Generally speaking, a traveler flying from a small-airport to another far-away-small-airport will take 3 flights: local-airport-to-close-hub, hub-to-remote-hub and remote-hub-to-remote-small-airport. Of course, there are variation, but that's the general gist of it.

A simpler way to put this : A large number of SERVED domestic flights at an airport will increase the demand of short international and large international routes at that airport.
Also, a larger number of SERVED international routes can increase the demand on domestic routes.

For my case, there are 2 airports located in Bucharest: LROP and LRBS. Having a large number of domestic flights from most local cities to LROP might create an increase in international routes, so that more passengers will fly to, say, EDDH from LROP than LRBS, because of the additional connecting passegers.

While not exactly easy this should be much simpler to implement than some sort of connectivity-path-finding, and it will model connecting passengers close enough: eg. larger airlines operating at large airports now have an incentive to partner up with smaller domestic airlines that would bring their passengers for international routes.

Offline JumboShrimp

  • Members
  • Posts: 5992
Re: Connecting flights/transiting passenger
« Reply #29 on: February 24, 2012, 11:07:19 AM »
Just dropping my 2c for an easy way to model connecting flights.
Generally speaking, a traveler flying from a small-airport to another far-away-small-airport will take 3 flights: local-airport-to-close-hub, hub-to-remote-hub and remote-hub-to-remote-small-airport. Of course, there are variation, but that's the general gist of it.

A simpler way to put this : A large number of SERVED domestic flights at an airport will increase the demand of short international and large international routes at that airport.
Also, a larger number of SERVED international routes can increase the demand on domestic routes.

For my case, there are 2 airports located in Bucharest: LROP and LRBS. Having a large number of domestic flights from most local cities to LROP might create an increase in international routes, so that more passengers will fly to, say, EDDH from LROP than LRBS, because of the additional connecting passegers.

While not exactly easy this should be much simpler to implement than some sort of connectivity-path-finding, and it will model connecting passengers close enough: eg. larger airlines operating at large airports now have an incentive to partner up with smaller domestic airlines that would bring their passengers for international routes.

That would be something along the lines of Fake Passenger Connection thread I started a while ago:
http://www.airwaysim.com/forum/index.php/topic,25004.msg126089.html#msg126089

It would be a diversion.  Since the real thing is far away, I thought it would be a good stop gap measure.  But spending time on this (which would bring us no closer to the real thing) would just postpone the real thing even more..

radu.poenaru

  • Former member
Re: Connecting flights/transiting passenger
« Reply #30 on: February 24, 2012, 11:43:55 AM »
When you are saying "the real thing", you mean actually computing connecting passengers with route-finding ala "I want to fly to Rome and the cheapest way is via a 2h stop in Budapest"? Cause' I don't think that can actually be modeled in a 35 min day.

What my (and your) proposal will do is allow airlines to "create" these airport-hubs (or destroy them by going bankrupt) thus adding a little variance in the demand of the routes.. not the real thing for sure, but still nice.

Offline JumboShrimp

  • Members
  • Posts: 5992
Re: Connecting flights/transiting passenger
« Reply #31 on: February 24, 2012, 03:04:25 PM »
When you are saying "the real thing", you mean actually computing connecting passengers with route-finding ala "I want to fly to Rome and the cheapest way is via a 2h stop in Budapest"? Cause' I don't think that can actually be modeled in a 35 min day.

A lot of corners will have to be cut to get it off the ground, such as at first only connecting within 1 airline (most have only 1 hub), probably no time of day etc...

What my (and your) proposal will do is allow airlines to "create" these airport-hubs (or destroy them by going bankrupt) thus adding a little variance in the demand of the routes.. not the real thing for sure, but still nice.

Definitely, every game world would be different.

irro

  • Former member
Re: Connecting flights/transiting passenger
« Reply #32 on: June 07, 2012, 05:12:38 AM »
Would it be possible to charge more on smaller legs to make serving smaller cities viable?  In real life United charges little on the IAD-ORD route, then make up the money on the regional flight to LAN, ect. 

Offline Tiberius

  • Members
  • Posts: 235
Re: Connecting flights/transiting passenger
« Reply #33 on: October 26, 2012, 05:38:04 AM »
Bump.

Connecting passengers is a must.  I have a friend who won't play again (purchase credits) until there are connecting passengers (and he wants mergers too).

I've read most of the post on this subject, and would like to voice my support for some sort of dynamic and potentially significant bump to demand when considering connecting passengers.

I realize the difficulty of this task because of the real-world information influencing airport and city pair information, but think there should be some way to build a hub in a smaller city (and by build I mean right-size to bigger aircraft as I grow).

Offline Andre

  • Members
  • Posts: 1091
Re: Connecting flights/transiting passenger
« Reply #34 on: October 26, 2012, 05:05:25 PM »
Why make it so complicated? The data is already there! Let me give you an example.

You have a hub (base airport) at A.
Two of your destination cities, B and C have 50 pax demand between them, but no flights offered.
These pax will have to fly to A, and then hop on a transfer flight from there.

Pax should prefer direct non-stop flights, so the formula for tech stops can be used. If another airline (or yourself) open a direct route between B and C, the pax will prefer to go on that route. This is how it works in real life.

One could set the "patience" of the pax to for example 4 hours (Maybe more on long haul routes?):

Mr. Pax lives in C, but wants to go to B. Unfortunately for him there are no direct routes, so he flies from C to A at 6:00 and arrives at 7:00.
This means the flight from A to B has to depart within 11:00.
If there are no flights scheduled within that time frame, Mr. Pax simply won't fly from C to A in the first place.

As for prices, pax will pay the price a direct flight would cost, minus the penalty for a stop (Like tech-stop pricing).
This means that the transfer pax flying from C to B will create less than half the revenue as the direct pax on the same flight (who's destination is A). The transfer pax will not be as lucrative as the direct pax, but still a welcome supplement.

C to B Direct Route price: $100
C to B Transfer Pax price: $80

$80 / 2 flights = $40 on each flight (C to A and A to B)

« Last Edit: October 26, 2012, 05:23:16 PM by Andre B »

exchlbg

  • Former member
Re: Connecting flights/transiting passenger
« Reply #35 on: October 26, 2012, 07:12:58 PM »
Sounds really simple. But it isnīt in any way. You say, pax pay some price minus something and take two of your flights to get anywhere.Who determines, which price they are paying ? The system? Pricing is your task. What statistics tell you about which passenger is flying where at what price? The whole route/airplane documentation would have to look different, that one additional column in your example doesnīt tell you anything.

Since the game and itīs basic data refer to real-world numbers, connection demand is worked into the game already. Otherwise the demand for real-world hubs would be much,much smaller. That makes it essential to have city-based demand introduced first, or traffic/demand will grow to unnaturally high figures.
I donīt have insight of the technical possibilities of the gameīs equipment, but computing connection demand/traffic might take much longer,so game would have to change to real-world days to handle all that stuff.

Next problem would be building up personal hubs. Should we stick to real world infrastructure or should we allow popping up world hubs in the middle of nowhere implying politics would simply build a big hub for you?

So there are many reasons why it`s not incorporated yet.
« Last Edit: October 26, 2012, 07:23:34 PM by exchlbg »

Offline Andre

  • Members
  • Posts: 1091
Re: Connecting flights/transiting passenger
« Reply #36 on: October 26, 2012, 11:39:06 PM »
Sounds really simple. But it isnīt in any way. You say, pax pay some price minus something and take two of your flights to get anywhere.Who determines, which price they are paying ? The system? Pricing is your task. What statistics tell you about which passenger is flying where at what price? The whole route/airplane documentation would have to look different, that one additional column in your example doesnīt tell you anything.

Since the game and itīs basic data refer to real-world numbers, connection demand is worked into the game already. Otherwise the demand for real-world hubs would be much,much smaller. That makes it essential to have city-based demand introduced first, or traffic/demand will grow to unnaturally high figures.
I donīt have insight of the technical possibilities of the gameīs equipment, but computing connection demand/traffic might take much longer,so game would have to change to real-world days to handle all that stuff.

Next problem would be building up personal hubs. Should we stick to real world infrastructure or should we allow popping up world hubs in the middle of nowhere implying politics would simply build a big hub for you?

So there are many reasons why it`s not incorporated yet.


Sounds really simple. But it isnīt in any way. You say, pax pay some price minus something and take two of your flights to get anywhere.Who determines, which price they are paying ? The system? Pricing is your task. What statistics tell you about which passenger is flying where at what price? The whole route/airplane documentation would have to look different, that one additional column in your example doesnīt tell you anything.

- Who determines what price they're paying? You of course. Thats how the system is now. It calculates a price fitting for the route, but you're free to change it however you want. What's the problem? Yes, of course the route pages would have to be updated and changed. There would have to be a route page for transfer pax. Let's say you open a route to a new destination. On the same page, you can get info on transfer pax willing to go there from other destinations too. It's not simple, but it's not a huge problem. I made the additional column to show you how it could look.

Since the game and itīs basic data refer to real-world numbers, connection demand is worked into the game already. Otherwise the demand for real-world hubs would be much,much smaller. That makes it essential to have city-based demand introduced first, or traffic/demand will grow to unnaturally high figures.
I donīt have insight of the technical possibilities of the gameīs equipment, but computing connection demand/traffic might take much longer,so game would have to change to real-world days to handle all that stuff.


- That really isn't a problem. The pax who wants to go from C -> B just goes through A instead. If an airline establishes a route directly from B -> C, the demand for C -> A -> B decreases depending on the seats offered from C -> B. Also, decreasing demand on some direct routes would be a very small price to pay if we got transfer pax into the system, wouldn't it?

Next problem would be building up personal hubs. Should we stick to real world infrastructure or should we allow popping up world hubs in the middle of nowhere implying politics would simply build a big hub for you?

- I really don't understand what you're saying here. What are you talking about? Your base airports are your hubs. The system with airport bases would be just the way it is now. Only difference is that transfer pax could fly through your base airport to get from B -> C.


brique

  • Former member
Re: Connecting flights/transiting passenger
« Reply #37 on: October 27, 2012, 12:44:14 AM »
I think you are missing the nature of the calculations involved for each and every flight : currently, the system calculates the load for a given flight, after considering all the variables: price, demand, supply, CI, RI, seats, frequency, etc , and the same for any competition on the route to determine who gets what share: for the above connecting system to work, each flight would also require additional calculations made for every other airport you serve. first to determine demand : how many pax wish to connect through your destination airport to that destination; and then the additional load: how many are actually going to fly that day and not forgetting : that the flights involved meet any connection time limitations : and is there competition, and how will that affect matters.

It will not just be pax wanting to go from B to C via A, after all : there will also be pax wanting to go B to D, E, F, G and so on.

So if you serve 50 airports, every flight will need its main calculation : A to B; then an extra 48 sets of calculations to determine connecting pax : that's an awful lot of calculating going on : for every flight : and every time you open a new route, you increase that set of calculations and increase the processor load again.

Just looking at one airline in MT#7 : it has over 17000 weekly flights to over 350 airports : we are going to need a bigger server.
« Last Edit: October 27, 2012, 12:48:35 AM by brique »

exchlbg

  • Former member
Re: Connecting flights/transiting passenger
« Reply #38 on: October 27, 2012, 10:10:53 PM »
You really donīt understand anything Iīm saying. Concerning demand figures: I said, that connecting passengers already are worked into the system, otherwise Atlanta for example would have the demand of any other middle-sized city of the US.
So the numbers of your transit passengers would just add on top of already worked in numbers, leaving demand/traffic hugely oversized.And Iīm not talking just lowering the figures a bit, it needs a complete data overhaul. (City based demand)

Concerning hubs: With connecting passenger system you can build a hub out of any base airport you choose.If we use city-based demand   itīs up to the players where they evolve their hub for connections. Real-world infrastructure data have evolved over ages with traffic/politics as given by history, but if we develop a new system of hubs, infrastructure would have to grow after our needs, turning airports like maybe Kansas City into a major american hub,leaving Atlanta a mediocre airport for example. Itīs going to be a completely new world for every game, not just re-playing history over and over.

In your example you just talk about one additional connection via hub. But whatīs about all other connections that are possible within 4 (why?) hours transit, each and every one forming a route you would have to establish, make prices for and check regularly.And itīs not just concerning you, but every player in game, the data to be handled would not just add up ,they would explode.
What about the time needed for every player, to harmonize all that additional demand/traffic? With big businesses, one additional flight from hub to a new destination might open up numerous of possible new connection routes, each and every one to be managed.How many staff you are going to hire in RL to help you manage that monster? How are the single passengers documented per aircraft, because on every single plane there will be direct passengers and passengers for/from up to hundreds of other destinations, not to mention the possibilities of connecting flights via more than one hub, connections to other alliance members, to other airlines, codesharing?
How should planning/documentation of flights/aircrafts look, when every flight is also used as a part of another one and every passenger has another final destination? How do you set prices, when splitting complete price evenly over used single legs canīt be the answer considering SH/LH connections? How do you work out single flight/aircraft results, if in worst case every passenger is paying a different fare for his single leg seat? What about combinations of different cabin classes?
You will have to think just one or two steps further to see all those additional problems to be solved besides your nice shiny new column.So itīs fully understandable why it is not implemented yet.
« Last Edit: October 28, 2012, 07:41:05 PM by exchlbg »

Offline Captim

  • Members
  • Posts: 1323
Re: Connecting flights/transiting passenger
« Reply #39 on: October 31, 2012, 08:03:41 PM »
+1 exchlbg

This would create much more dynamic GW's. Any hub with a big enough runway could host an alliance.

Perhaps alliance funds could be channelled into building the terminal infrastructure etc.?

Why not trial this is a small scenario first?

The game is great, but to have a GW that truly moves with the players decisions and not just following RL events would be a special challenge...
 

 

WARNING! This website is not compatible with the old version of Internet Explorer you are using.

If you are using the latest version please turn OFF the compatibility mode.