Demand&Supply System/Scheduling

#1

Hello everybody!

I know ideas regarding the scheduling system already came up at different places in the forum. I want to collect some of the ideas and expand on them. Of course, a demand&supply system and scheduling are two big things for itself, however - as they are highly interconnected and the first is necessary for the second - I discuss them in one topic

Demand&Supply System

In order to have realistic routings and frequencies we need a (maybe basic?) demand&supply system. The question of course is on which variables such a system is based. I’ve checked the airport database and saw that information about locations of airports as well as a very basic characterization (small, medium, large) of airports is already available. So first two variables defining demand between two airports could be the distance between them and its size. For the player’s airport characterization one could take into account the size of the airport that is near to the player’s airport in real life.
However, for this approach, it might be helpul if there are more than three categories for the size of an airport. Due to the large amount of airports ingame I’m not sure how to get data (=passenger data) of every individual airport. Any ideas on that?

Scheduling

Based on possible demand, airlines start to make flight offers at player’s airport (after having signed a master contract, I like that idea ;))
To have something like a “slot-system”, I suggest 4 different slots over the day. Of course this number could be changed. You can find an example of that in the pic.

Two examples how offers could look like

Airline A: AA1234 3/7 Monday, Wednesday, Thursday PMI-XXX-PMI in a morning-morning slot

  • Schedule with flight number AA1234 should operate three weekly (3/7) with flights on Monday, Wednesday, Thursday
  • Routing is PMI-XXX-PMI (XXX=Players airport)
  • Turnaround should take place in a morning-morning slot, meaning the arrival as well as the departure should be in the morning

Airline B: BB5678 2/7 Saturday, Wednesday LPA-XXX-LPA (afternoon-evening)

  • Schedule with flight number BB5678 should operate two weekly (2/7) with flight on Wednesday and Saturday
  • Routing is LPA-XXX-LPA
  • Turnaround should take place in a afternoon-evening slot, meaning the arrival has to be in the afternoon, the departure in the evening slot.

These are recurrent schedules, meaning they are repeated every week unless cancelled by the airline or airport.

Further possibilites of expanding that idea:

  • If you cannot provide the time-slot that is requested by the airline you can try to allocate a different slot for one or more flights included in the offered schedule. For example you could ask Airline A if the flight on Monday could also operate in a morning-afternoon slot. Of course, chances that this will be accepted by the airline increase a bit if you offer a discount for that flight. That’s a possibility for negotiation.
    If - after some time - you are able to reallocate it to the requested schedule the airline will pay the usual fee determinded in the master contract. Of course an overview about flights that do not operate according to requested schedule would be helpful.
    (See post below for an improvement of that idea)

  • Depending on the success of a certain destination, airlines can decide to expand a schedule, e.g. Airline A could add a fourth weekly rotation on Friday. However, they can also decide to cancel rotations or the entire route if demand is low.

  • It could even be possible that airlines ask for an upgrade or downgrade of the used aircraft, assumed suitable aircrafts are available. Especially when an airline would like to upgrade, the player has to make sure that suitable infrastructure is in place or otherwise has to reject the offer of an upgrade which could reduce airline satisfaction.

  • When having a supply&demand system another nice thing would be an overview about the performance of schedules. Of course airports in real life do not necessarily have information about the financial performance of a route for the airline, but they have information about passenger capacity usage. If you see e.g. that Airline B only has a usage of 50% on its LPA-XXX-LPA schedule above you could decide to financially support marketing activites of the airline a bit in order to increase usage and decrease the likelihood that the schedule is cancelled.

  • If you see that a route is perfoming very well but the airline has not offered more flights one could maybe even activly ask for more flights on that route.

  • Kind of airline competition: The supply&demand algorithm gives a certain demand value for a specific routes. If more than one airline serve that specific route, demand might increase a bit beyond that initial value, but at some point there is on oversupply and airlines start to reduce or cancel services.

Of course, these are just ideas - maybe they are good, maybe not - what are your opinions on that topic? :slightly_smiling_face:

2 Likes
#2

Sorry, for the double post, but I already have an improvement for my idea regarding the “rescheduling discount”, when you cannot assign a flight according to the request. (first bullet point unter “further possibilites” in the post above)

It might be a little bit tiring to negotiate that “rescheduling discount” for every single schedule an airline offers, when you are not able to comply with the request regarding the slot.

A better solution might be to include that “rescheduling discount” already in the master contract with an airline. Then you are basically free to assign a flight wherever you want, but get a lower payment when it does not satisfy the requested time slot.

With this solution micromanagement would be reduced, but those tycoons who really like to optimize everything, can still try to assign everything according to the reqeust and get a higher payment for that.

1 Like
#3

I really like your ideas @Pax2017 . Here are some additional things I want to add.

Due to the large amount of airports ingame I’m not sure how to get data (=passenger data) of every individual airport. Any ideas on that?

Do you mean something like this?

Back in 2017, as a part of modding development me and couple of people managed to create a list of annual passenger and cargo data of all 14.144 airports in game files.

Back in days, my and other players’ idea was this;

  • Small airplanes will be used in between player airport and small sized destination airports
  • Medium airplanes will be used in between player airport and medium sized destination airports
  • Large airplanes will be used in between player airport and large sized destination airports
  • Every airplane will have max range and minimum operation range so you wouldn’t see A380 in between London and Birmingham as same as Cessna Caravan in between London and Hong Kong.
  • Modded airplanes will have a home country so there wouldn’t be Aeroflot flight from USA or Japan Airlines flight from France to player airport.

Data is still with me and it has been shared with developers back in 2017 :wink:

Airline A: AA1234 3/7 Monday, Wednesday, Thursday PMI-XXX-PMI in a morning-morning slot

Airline B: BB5678 2/7 Saturday, Wednesday LPA-XXX-LPA (afternoon-evening)

  • It is how it suppose to be. Currently we can plan any flights as how we want. I prefer to have some limitations from airlines.

  • I like time slots but moreover, I think ground time should also be decided by airlines. Some airlines will request 2 hours ground slot with less services but while some will request 6 hours with full service

  • Moreover, some airlines may request jetway stand but meanwhile some may request remote stand. IRL, jetway stands are more expensive than remote stands so some airlines may prefer remote. Also their aircraft type can also have an effect on it. Currently you can park ATR to jetway stand with unrealistic jetway connection.

  • There could also be (evening-morning) flights so that means airplane will stay overnight and departure at the next morning.

  • I agree with upgrade and downgrade of airlines, airplanes of frequency but I think “negotiating” would make micromanagement level to “boring” and “hard”. I mean simply, player can deny that flight and airline can suggest another flight for player.

Another thing I want to add is “fleet” information for airlines. Currently airlines can offer tons of flights which makes things boring. So here is how it should be;

  • Every airline should have a fleet. For example: (10) Cessnas, (5) A320, (2) A330 and (1) 787. There shouldn’t be 11 Cessnas on the player’s airport at a time. There shouldn’t be 11th Cessna flight while other 10 Cessnas are either arriving,on ground or departing. Same for 6th A320, 3rd A330 and 2nd 787.

  • In game default airlines can have infinite fleet in order to not breakdown gameplay but it would be very useful for mods and people who want to add “spice” to their games.

4 Likes
#4

Wow, I did not know that there is already a database with pax (and even cargo) figures for all airports. I think this can be very useful in setting up an algorithm calculating demand.

Modded airplanes will have a home country so there wouldn’t be Aeroflot flight from USA or Japan Airlines flight from France to player airport.

One problem could be that e.g. a Ryanair does not have only one “home country”. Of course it is registered in Ireland, but operates bases in a lot of countries in Europe. Maybe instead of home countries we should use “base countries” or even “base airports”.

I think ground time should also be decided by airlines. Some airlines will request 2 hours ground slot with less services but while some will request 6 hours with full service

This would be great. too. And of course it would also work with the suggested slot system.

Every airline should have a fleet. For example: (10) Cessnas, (5) A320, (2) A330 and (1) 787. There shouldn’t be 11 Cessnas on the player’s airport at a time. There shouldn’t be 11th Cessna flight while other 10 Cessnas are either arriving,on ground or departing. Same for 6th A320, 3rd A330 and 2nd 787.

That is also a cool idea. One could also expand that, e.g. that an aircraft with a special livery only can appear once at an airport, because also in reality this special livery only exist on one plane.
However, this is only relevant for mod-airlines, so it will probably not be top priority for the devs.

1 Like
#5

Skip this. There are people like me, who want to build an Airport near their home location and they don’t want to be limited to small airfields only just because they aren’t living in an metropolitan area.

If there may be versatile difficult options some time, supply and demand could be a part of that to be individually selected.

#6

Sure, this is also an option. Maybe an even better one :slightly_smiling_face:

So, in the beginning you can determine the general demand at your airport - eg. “low”, “medium”, “high”, “very high”.
Then the offered routes may only depend on the destinations airport size, but not on the player’s airport, or more precisly demand of player airport is predetermined by the user. With this it would be possible to build a “mega-hub” everywhere. :slightly_smiling_face:

#7

One problem could be that e.g. a Ryanair does not have only one “home country”. Of course it is registered in Ireland, but operates bases in a lot of countries in Europe. Maybe instead of home countries we should use “base countries” or even “base airports”.

There could be multiple Ryanair airlines based on other countries. Like how it is for TUI;

TUI UK
TUIfly Belgium
TUIfly Deutschland
TUIfly Netherlands
TUIfly Nordic

That is also a cool idea. One could also expand that, e.g. that an aircraft with a special livery only can appear once at an airport, because also in reality this special livery only exist on one plane.

Could be :wink:

However, this is only relevant for mod-airlines, so it will probably not be top priority for the devs.

It is also true. Currently with modding, all you can do is adding new businesses and airlines to the game. Developers are working on game updates and new features so I don’t expect anything new for modding.

But when the time comes I really want to be able to create gameplay changing mods like how Cities Skylines or Prison Architect have.

#8

This makes sense from a “simulating” point of view where I am recreating an existing airport. But from a chronological gameplay perspective where you build up an airport, it would be weird to have a very popular airport (with many airline offerings) that still operates on grass stands.

There could possibly be an option where the amount of flights offered should grow steadily of 10% per year/period. Realistically an airport won’t grow much faster than that under normal circumstances. In my opinion there should be a balance between ‘realistic’ operations and a player’s freedom to build whatever he wants.

This would probably mean that an airport’s popularity should be unlimited in size, even regardless of services. If a player wants to make an airport with 50 stands and quick turnarounds without baggage handling, why should he be limited? If a player wants to build an airport in the Sahara desert, where aircraft always require de-icing, should he be limited just because it isn’t realistic?

#9

I like all your ideas on supply/demand and how it can make flight scheduling more realistic, to build on your ideas:

  1. Make flight scheduling route based rather than flight based. This would mean that airlines launch routes from your airport to another airport and then determine the number of flights on that route they’d like to offer - initially, say, 1 flight / day. Then based on how successful their flight is (a function of airline and customer satisfaction with the airport) they can offer additional flights / day, which you schedule. Airlines wouldn’t want flights to overlap so could request certain time blocks through the day. From a UI perspective, ACEO can then view their airports route network, as well as filter the slot usage by route to see how well their airport serves that particular route. Over time the route network will build out and so the overall attractiveness of the airport will grow.

  2. Make certain times of day more attractive through higher passenger numbers. This feature would mean that airlines would be willing to pay more to operate at certain times of day and also larger / smaller aircraft at different times. From a UI perspective, showing the route and demand data in a simplified way (total demand each day, morning, day-time, evening) would give more insight in to what infrastructure to construct.

Adding the two features above would mean that, as the airline grows and adds more frequencies and more routes (and also as new, competing airlines launch) the demand for flight schedule slots grows. Nice impacts of this could be:

(a) different types of routes, eg business routes where morning and evening are important vs leisure routes where middle of the day are more attractive (and so more demand).

(b) dynamic pricing of the landing slot based off time of day (simulating airlines competing for the most profitable times for them).

© Driving the infrastructure of the airport development (ACEO needs to achieve the right balance of gate size, facilities etc) to meet the demand coming from airlines.

This would add a lot of depth to the revenue side of the game, more on how to you optimise a/b/c to drive the maximum revenue for your airport.

If the demand for routes is based on real life passenger demand, say from 2018, then this would make building an airport in any city a totally different experience and challenge - since demand for air travel is never the same in two cities (think building at London Heathrow vs Ibiza).

Demand Data:

On the demand data side, one important feature is whether at some point ACEO will include connecting passengers who can transit through the airport from one flight to another. If they include that then we need to know, on any plane how many passengers are terminating at the airport vs connecting to another flight.

All this data exists from IATA (DDS is what it’s called). I’m not sure if there’s a free source… This data has “route” based data showing passenger demand by routes and also “O&D” based data showing passenger demand between any airport pairs, taking any itinerary. Most simply, using airport pair data for routes as a starter would be OK. Building a model that thinks about connecting passengers and simulating where they are connecting (as so how much of that connecting demand your airport would get) is a much bigger problem to solve! I’m sure it can be found somewhere or a summarised version that has the data needed!