City based demand

[edidiot]

Hi Sami, I know this was mentioned quite a while back but are we any closer to city based demand? There are plenty of routes (to holiday places mostly) that have big demand from Gatwick but non from Heathrow. The stats are based on the real world experience but this is not because people would not be willing to fly from LHR to these destinations but rather that traditionally that has been that airlines have chosen Gatwick for leisure and Heathrow for business routes. I can see three ways of remedying this.

1) Create city based demand (the best way but I understand this would be a big task)

2) Allow an airline based in a city with multiple airports to base aircraft out of all of them

3) Allow short repositioning flights, at the moment I'm not allowed to fly an acf from LHR to LGW and start a route from there even though I accept that I wont get any revenue for the ferry flight, if this was allowed then 1 or 2 would not be necessary provided the positioning flight was not counted as a leg for turnaround basis other than the time taken to make the flight.

For example - I have a 747 that is based at my LHR hub that I want to send out of LGW. I make a positioning flight for the acf to LGW. Rather than a traditional leg the time taken to do this is just added to the turnaround time for the flight. Say if we needed 200 mins to get 0% delay possibility before from LGW this would be 260 mins. I think this would keep the balance of the game but stop the unrealistic impossibility of flying out of two airports in the same city.

[Sami]

Being planned for future versions but actual coding / planning is 0% at the moment, as that's for a thing for "AirwaySim v2" or something like that.

I acknowledge the problems in the current airport based model but that's the first version as it was faster and easier to make.

[edidiot]

thats cool, thanks Sami

[Sami]

I'm slowly gathering data on this and planning the system, however this is NOT expected anytime soon. If expected at all, I cannot tell yet.

But the first step has been to get list of cities worldwide and also their coordinates, and population... Here's a few pics where I've compared the results from a single country (Finland) to what Google is giving on their maps function. As you can see the city dots (location) is very close to what they should be on the map. The city location is of course needed as I must tie each airport to the cities and each airport would have a certain 'catchment' area... The system cannot be fully accurate but plan is to find data on cities with population of over 1000-10 000 people which would bring the countrywide accuracy of population to 90-95%.

But this is all just very early testing, the decision to start building the city based system instead of airport based is not even made or planned.

(the maps won't be ever visible to users, there's so much data to display .. but just showing here what kind of data the system is built on)

[attachment deleted by admin]

[Sigma]

Very, Very cool Sami.  This is definitely something I'd like to see in the future.

[bukatino2000]

I'm slowly gathering data on this and planning the system, however this is NOT expected anytime soon. If expected at all, I cannot tell yet.

But the first step has been to get list of cities worldwide and also their coordinates, and population... Here's a few pics where I've compared the results from a single country (Finland) to what Google is giving on their maps function. As you can see the city dots (location) is very close to what they should be on the map. The city location is of course needed as I must tie each airport to the cities and each airport would have a certain 'catchment' area... The system cannot be fully accurate but plan is to find data on cities with population of over 1000-10 000 people which would bring the countrywide accuracy of population to 90-95%.

But this is all just very early testing, the decision to start building the city based system instead of airport based is not even made or planned.

(the maps won't be ever visible to users, there's so much data to display .. but just showing here what kind of data the system is built on)

very good sami but please do not forget that statistical/economic indicators have to be available to players otherwise I see no chance/logic to take decisions or drive the company. For heaven's sake is not to earn money nowadays!

[Sami]

the economics of a country is a completely different module and that of course is needed for the city demand too.

[Filippo]

Very cool!

But, what about airports which serve cities but that also have a huge catchmen area?

Like Malpensa (Milan - Italy). That airport serves basically all the north of Italy, not just Milan.

[CX717]

My very ideal imagine: 
probably they will pick the closest airport(AIRPORT A) that serve the flight to AIRPORT C
if the closest airport are too far away from the city,they will pick another closest airport(AIRPORT B)with connection to AIRPORT A.
that generate AIRPORT A to AIRPORT B demands,and also boost the demand from AIRPORT A to AIRPORT C(connecting passenger).

we can have different kinds of passenger AI,"wealth passenger" AI,who aim for luxury."business traveller"AI,who aim for flight timing,fastest flight,late departure and morning arrival.
"budget travller"AI,who aim for lowest fares,they don't care about comfort."normal"AI,they aim for decent service,and normal fares.etc.......

[samomuransky]

I know that city based demand is planned. But there's another problem - what about close airports which are in different cities (sometimes even countries), but passengers are popular for people from both cities.

For example, Bratislava and Vienna. That airport are about 59km from each other and people from both cities use both airports... Vienna is large airport and provide many destinations which are not available in Bratislava (that's why many people from Bratislava use Vienna) and on other hand, Bratislava airport is cheaper and is more attractive for low costs like Ryanair and therefore, passengers from Vienna which are looking for low fares often use Bratislava...

So, any ideas how situations like this could be implented?

[Sami]

Yes that's a problem with it still ...

Each airport will have a radius of influence and all cities under it will be served by that airport. But as you pointed out there are very close-by airports, even at the same city .. So dealing this may be complicated and haven't yet found a proper solution. Calculating demand individually for each city & village isn't really an option as the city database has over 100 000 locations.

[Sigma]

Close airports and the changes you could have there over the current system, are one of the best potential features of having the city-based demand.  So I wouldnt call it a "problem", I'd call it almost the entire reason why we need the feature in the first place.

On of the best parts of having the feature is that you could either Bratislava or Vienna, or DFW or Love, or JFK or LaGuardia or Newark (to pick 3 in close proximity), to become to dominate airport in the region in our world, regardless of what reality has dictated.  Airport size and runway length would still be 'realistic' and would create a limiting factor, but one of the defining 'needs' for the feature, are that you should have to worry about a start-up coming to the airport across town and flying to your destinations and stealing your pax.  As it is now, an airport 10 miles away can't steal one iota of your demand to a destination city, whereas in real-life airlines do this constantly.   Southwest Airlines' entire business model for example is based off the idea.

If a solution can't be found to the problem, it takes away a good 90% of the value of the feature in my mind.  It still would lend itself to regional demand variations, which would be a plus, but it would still be a significant hindrance on the 'realism' of the game, as it would still leave us in an environment that is simply a mirror of the reality we know with little ability to change it (which, I suppose, mirroring reality is "realistic" in its own right, but not "realistic" as a business sim of sorts).

[Sami]

Yes, I know the benefits and I'd like to have the city system too, I just described the biggest problem what I am having with the planning right now .. All clever ideas welcome 

[Sigma]

Well, my post was in regards to Samo's calling it a "problem" of city-based demand -- in my opinion it's not the problem, it's the reason.  The problem is trying to make it work.

I would have to think it would be fairly easy (and I use the term loosely) to have airports within Xkm be, essentially, the 'same' airport as far as the pax demand was concerned.  You would see the same demand graphs (and associated flights so you'd need a header for airport at the bottom so it was clear where the competition was flying from) from either airport.

However, that would be a very big "band-aid".  Why, for example, would it be 100km and not 200km or 50km?  And what if you're 2km based the 'border'?  You don't have to worry at all anymore.

In a perfect world, the further an airport got from another, the less demand it would 'share'.  But that becomes exceedingly difficult to both program as well as display on a demand chart.  If you're 100km away and half the demand for your airport is willing to drive to a further one to grab a flight, how do you put that on a graph?  You can't show the flight capacity like we do now for sure.   And how do you show the passengers that aren't as willing to travel.  And then you have to put a travel function in the preference model -- at what price are they willing to drive Xkm to a different airport?

It certainly gets difficult.

[samomuransky]

Problem is that it doesn't exits now and it will not exist if it will compare just airports in one city

But, yes, same problem is that we don't have any solution... I was thinking lot about it and I didn't found nothing...

[cmschroeder]

.  I'm going to argue that even if airports are close together, if their is real world demand then they are in different regions.  As posted above there are places where people fly 40km.  However, there are places where people don't fly 40Km.  Manchester-Boston Regional to Boston (about 40km)  has a free bus that runs between the two, no one needs to fly it because even though they are in Diffrent US States they are in the same region.  It would be a ton of work, but Probably the best way to handle it is to manually assign regions to airports, or manually tweak it once you do it automatically.

I wonder, as I'm typing this, if it makes sense not to have one regional catigories, but several.  People will travel farther to travel farther.  I'm not going to drive two hours for a 45 minute flight, but I will drive (or rail or what ever) that long for an 8 hour flight.  What I'm thinking is You have a Metropolitan region, that draws from a very small area around the airport, The pax from this region don't want to fly very far at all.  Then you have a 'local' region (that needs a better name,) People come from a larger region to travel farther,  I would say this would be larger than states, but not the whole country for the US, and European countries for the EU.  The final region would be the International Region, these are the PAX that are looking to travel long haul.   Now to help with the connecting flights Idea, you  take a percentage of the pax from the smaller region and add it the demand of the next region.  IE  some pax will be flying regionally to fly internationally.  Now your DHC-6's help bring regional travel to your Hub.   I'm not sure I'm explaining this as well.   Does it make sense?

Christopher

[Maarten Otto]

The radius can also be a problem as for example EHAM (Amsterdam) is a main airport for the entire nation. That's why train services from all corners of the country go to the airport, and not Amsterdam.

[Seattle]

-First of all, I'd like to say, Im only typing with on hand so, there are bound to be alot of mistakes -

As I sent in a Pm to you, Sami, its pretty easy to do cities with with 1-2 airports, If the second one is very small.

Examples: for Seattle you have SEA and BFI, the main is Seattle and a very small amount pax go through BFI [Less than 20,000].
Another: for Portland you have three, PDX, Troutdale, and then Hillsboro. The last two are like BFI, with very few passengers. To get more accurate demand, these airports can - A: Be left alone and have indiviual route data submitted, B: Demand is generated by only the local communities. The main airports obviously have demand from their whole cities population.
----

For Airports like DFW/DAL and IAH/HOU, where we have a clear main airport.... but there is still a secondary airport that has noticable service. So, as before the major airport has demand from the whole area, but this time this larger secondary airport takes demand from a larger chunk of the local area. In both cases, HOu/DAL are located closer to the city's center. So they might take demand from the county/cities/region/prefrecture/provence/etc that they are in.
----
Alright, Im going to split this post into two as.... The forum does funky thing after making it longer than this, which makes it hard to type.

[Sami]

There will be no "submitted routes" as the idea is not to mimic what real airlines do and where they fly, the demand must be fully generated..

[Seattle]

So, then you get to the messy cities with more than three major airports: Los Angeles, New York, London, Berlin, Bay Area, Stockholm, London, London, and London again   .

What I proposed to Sami in my pm was that [Im using LA for an example] the main airport, in this case LAX caters to the whole of LA's population and then the smaller airports: SNA, BUR, LGB, ONT all have demand from there local areas. For example, SNA is in Orange County, therefore it serves Orange County, pop. 3 million. For BUR, you can use the local cities population and you kind of guestimate that some of LA's [the municipal city] 3.8 million people will use it [500-600,000?] along with Burbank, Pasidena, Glendale, and San Fernando.

It's compliacted, but I dont really see another way to do it for these huge cities with 5 airports.
----

For cities like New York, where you have to main airports and a third major airport, you could potentially have demand for JFK/EWR come from the whole NY area and for LGA you could have just New York's municipal population. For Whiteplanes/Newburgh/Long Island, demand comes from the local communities.

Of course, the way I see it is pretty complicated but, I dont really think there is anything else outhere.