Author Topic: Airport control towers  (Read 951 times)

0 Members and 1 Guest are viewing this topic.

Offline jamespetts

  • Simitrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 15857
  • Total likes: 407
  • Helpful: 177
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Airport control towers
« on: May 26, 2017, 01:16:47 AM »
I should be interested in people's views on a thought that I have had recently concerning airport control towers. Currently, these are either compulsory for all airports, or not compulsory for any airports. In the latter case, they are merely decorative.

However, it has occurred to me that there is a straightforward way of making them compulsory for only larger airports by defining larger airports as those with more than a certain number of lines/lineless convoys of the air type registered to call at them. A sensible default may be 3. One could then have a check to see whether any airport without a control tower has the requisite number of scheduled services calling at it, and refuse to allow any other schedules to be altered so as to add it to them unless either a control tower be built or the number drop below the maximum.

However, there is a potential problem with this, which is that it is possible to build multiple airports connected to one runway. This can be useful in game for building multiple termini, but this could be exploited by players with small airports wanting to avoid the expense of a control tower, and finding a way to deal with this exploit might add considerably to the complexity of the code for what was intended to be a simple and straightforward feature.

If anyone has any views on a very simple and robust way to code a means of preventing this exploit, I should be very grateful.

Incidentally, apologies that I have not had time to look into bug reports a great deal of late: I have been spending an enormous amount of time trying to narrow down the desync issues with network/online gaming, and have made considerable progress in this regard, but it is an extremely time consuming sort of issue to deal with (but does leave testing down-time during which I am not able to work on the code).
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Offline Ves

Re: Airport control towers
« Reply #1 on: May 26, 2017, 12:35:38 PM »
I think it is the wrong approach to limit the number of lines that a station equiped airport can handle. Besides the exploit you generate, you will also eliminate the possibility for a small aistrip handling a large number of lines with only one plane in each line, which I think it should be possible to do.

I think one could look what a control tower does in the real life:
It provides ATC (Air traffic control) which basically is a separation service between airplanes. Without any ATC (like class F and G airspaces) the airplanes are required to maintain separation by them self following certain rules like minimum 1000 feet below clouds etc. When landing in uncontrolled airspace, it is the pilots responsibility to check if the airstrip is free and then to announce everything he/she does over the unicom radio. It works just fine, but its quite ineffective.
In controlled airspace with ATC (and there are different grades of that also) the ATC can see through the clouds, predicting all other airplanes movement in the area, therefore having a much better overview, allowing more planes in narrower space.

Now that is quite advanced stuff, and we dont need any class airspaces, only a control tower when we want big airports.

What about something like this:
Untie the control tower from the stations and make them a standalone building, much like signalboxes.
The control tower will have a control radius of, say 3 km, within it is considered controlled airspace.
If an airplane tries to take off from an airfield, it will see if there is any control towers nearby. if there is, it will take off immediately if the runway is clear. If there is no control tower in the vicinity, it will have a 30 second to a minute of "looking around, checking for other airplanes" before it takes off.
The same thing goes on for landings: If a plane wants to land at a runway within a control tower sphere, it will reserve and land like the current behavior. If there is no control tower, it will have to circle around for 2-3 minutes while "checking for other planes" before it can actually land.
If the runway gets booked by another plane while the plane is waiting its 30 seconds to 2 minutes checking the surroundings, the timer resets, thus becoming very ineffiecient when many planes operate in close vicinity.

This would have the benefits:
Higly promote the use of control tower at airports with a bit more traffic.
You can have multiple separate "stations" within the airport served by the same control tower, and also multiple runways.
It would be hard to exploit, as the tower radius would set a limit to its usage.
Its a more realistic use of the tower and you can place the tower more strategically in the middle of the airport and not directly at the gate.

Some decissions to make:
If one part of the runway is within the control tower sphere and the other is not, either the entire runway is then considered within or not.

Now, I apologize, this might not be a straight forward solution, but one which I have been thinking about recently.

Offline jamespetts

  • Simitrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 15857
  • Total likes: 407
  • Helpful: 177
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Airport control towers
« Reply #2 on: May 26, 2017, 05:14:39 PM »
Thank you for your feedback. This sort of solution is likely to involve considerably more complex coding than what I had proposed. Do you have any thoughts as to how this might be coded robustly and efficiently? I am always troubled by having to code anything that involves interaction with vehicle movements (for example, the signalling code proved to be extremely difficult, and there are still bugs being found in that even now, although work was started on this in January 2015).

Also, does anyone else have any thoughts on these issues, how in practice airports without ATC work and whether either or both of these suggestions would realistically simulate the economics of such airports? I do wonder how busy that the busiest airports without ATC are in reality and whether the maximum realistic frequency of aircraft from an uncontrolled airport would be realistically simulated by either of the measures proposed here.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Online Vladki

Re: Airport control towers
« Reply #3 on: May 26, 2017, 09:22:07 PM »
I don't know how airports without atc work. But I think that the big airplanes have to report to some ATC all the time. Not only for takeoff/landing. But that may have not been the case in early years.




Offline Ves

Re: Airport control towers
« Reply #4 on: May 26, 2017, 09:46:26 PM »
I don't know how airports without atc work. But I think that the big airplanes have to report to some ATC all the time. Not only for takeoff/landing. But that may have not been the case in early years.
The big airplanes (ie passenger jets) dont require ATC. However, the kind of airport those planes usually fly to are generally equipped with a control tower, and therefore ATC. But nothing would prevent a 747 to land at an airport with no active control tower. The procedure would simplyfied be that the plane is handed over to some unicom frequency by the last center control, and is then expected to keep its own navigating, telling over the radio to other planes in the area what its intentions are. Probably the plane would fly a traffic pattern over the runway to check if its clear and then land.

some information from wikipedia:
https://en.wikipedia.org/wiki/Non-towered_airport
https://en.wikipedia.org/wiki/Controlled_airspace
and another site:
http://www.angelflightwest.org/pilot-page/a-culture-of-professionalism/safety-articles/non-tower-airport-operations/

Offline jamespetts

  • Simitrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 15857
  • Total likes: 407
  • Helpful: 177
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Airport control towers
« Reply #5 on: May 26, 2017, 10:55:08 PM »
The Wikipedia article states,

"When the traffic volume at an airport gets too high for safe and efficient operations, or when the mix of aircraft types and speeds becomes too large, an airport may be considered for a tower," which does suggest that there is a maximum level of usage possible for a non-towered airport.

Also, from what I understand, it is not generally necessary for aircraft to line up on the runway and wait at non-towered airports: they can just keep going from the taxiway, as they will have been checking their environs visually and by radio during taxiing.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Offline DrSuperGood

Re: Airport control towers
« Reply #6 on: May 27, 2017, 07:26:26 AM »
Airport control towers do not just control the sky over the airport, they also control the sky for quite a large area around the airport. Instead of being part of a stop, they are instead an entirely separate building which applies to all airport stops in a large area.

One might be limited to only 2-4 simultaneously in transit flights to airports within 50-100km of each other. When a plane tries to depart a stop it will refuse to do so if there are too many in transit flights within the 50-100km area of its intended destination unless there is also a airport control tower within that area. This is a bottlenecking limit unrelated to line configuration or number of convoys.

The actual limit should change with time. Back in the 1960s plane safety was nowhere near as good as now and neither was the control tower technology so one could expect more risks such as more planes heading to an airport etc. In 2017 control towers are pretty much mandatory for any airport that receives all but the smallest of traffic volumes as they aim for as good as no collision chance.

The control area of a control tower should increase with technology and size. A 1960 one might only be able to control a small area due to mostly relying on vision or basic radar. In 2017 advanced tracking keeps watch on aircraft 100s of km away with the control towers only needing to be positioned right near runways for visual security (in case all the tech fails they can still look out the window).

As such one could have control towers with a few properties.
1. Line of Sight range. Some aircraft require being in this range to take off, especially modern large scale passenger aircraft.
2. Radio range. Some aircraft require their destination to be in this range before departing, especially modern large scale passenger aircraft.
3. Control range. Aircraft require that their destination be in this range before departing if more than a time period relevant number of aircraft are in transit within a time period relevant range of their destination.

Practically all, not light aircraft, modern airports have line of sight and radio range. It would be too dangerous to land otherwise and put passengers at too much risk. However control is only needed for busy areas to handle more than a couple of flights at the same time. As time progresses, control becomes more important as more strict regulations mean that fewer aircraft can safely land in close proximity.

Additionally one could add services to airports, such as fire response. Some aircraft types require that both source and destination airport have such services before departing, eg modern large scale passenger aircraft. This would be a different type of building and have a service range (5-10 km odd, usually 1 is enough to cover entire airport complex).

Offline jamespetts

  • Simitrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 15857
  • Total likes: 407
  • Helpful: 177
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Airport control towers
« Reply #7 on: May 27, 2017, 11:47:34 AM »
Thank you for the interesting ideas. I am not sure how workable that some of these things would be, however. For example:

One might be limited to only 2-4 simultaneously in transit flights to airports within 50-100km of each other. When a plane tries to depart a stop it will refuse to do so if there are too many in transit flights within the 50-100km area of its intended destination unless there is also a airport control tower within that area. This is a bottlenecking limit unrelated to line configuration or number of convoys.

Imagine two airports 500km from one another, airport A and airport B. Airport B is also 200km-1,000km away from, say, 10 other airports to which it has direct connexions. From each of those airports, an aircraft departs to airport A at 1 hour intervals and fly at different speeds. In such circumstances it would be more or less random whether any aircraft from airport A could take off for airport B at any given time. Remember, I am aiming for large map sizes of thousands of tiles (at 125m/tile) in each direction, and thus potentially over 1,000k in at least one dimension. How would players be able to deal with what they can actually hope to send to a non-towered airport in such circumstances?

Also, are airport control towers not quite distinct things from airspace control towers?

Quote
The actual limit should change with time. Back in the 1960s plane safety was nowhere near as good as now and neither was the control tower technology so one could expect more risks such as more planes heading to an airport etc. In 2017 control towers are pretty much mandatory for any airport that receives all but the smallest of traffic volumes as they aim for as good as no collision chance.

Has the proportion of non-towered airports declined over time? It would be interesting to see data on this. Also, is this idea of an actual traffic limit on airports without towers not similar to my original suggestion? Also, would variation with time not break players' routes in ways that is incompatible with the idea of players being able to leave a multi-player game unattended for several in-game months at a time without significant adverse consequences?

Quote
The control area of a control tower should increase with technology and size. A 1960 one might only be able to control a small area due to mostly relying on vision or basic radar. In 2017 advanced tracking keeps watch on aircraft 100s of km away with the control towers only needing to be positioned right near runways for visual security (in case all the tech fails they can still look out the window).

As such one could have control towers with a few properties.
1. Line of Sight range. Some aircraft require being in this range to take off, especially modern large scale passenger aircraft.
2. Radio range. Some aircraft require their destination to be in this range before departing, especially modern large scale passenger aircraft.
3. Control range. Aircraft require that their destination be in this range before departing if more than a time period relevant number of aircraft are in transit within a time period relevant range of their destination.

Practically all, not light aircraft, modern airports have line of sight and radio range. It would be too dangerous to land otherwise and put passengers at too much risk. However control is only needed for busy areas to handle more than a couple of flights at the same time. As time progresses, control becomes more important as more strict regulations mean that fewer aircraft can safely land in close proximity.

Additionally one could add services to airports, such as fire response. Some aircraft types require that both source and destination airport have such services before departing, eg modern large scale passenger aircraft. This would be a different type of building and have a service range (5-10 km odd, usually 1 is enough to cover entire airport complex).


Would this not involve a full simulation of airspace control centres as well as airport control towers? It is hard to see how this could work with a reasonably limited coding effort (given that this is not a priority feature, so is only likely to be implemented before at least 3-4 years into the future if it can be done very easily without adding another entire simulation layer).
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Offline Ves

Re: Airport control towers
« Reply #8 on: May 27, 2017, 01:40:37 PM »
Quote
Also, from what I understand, it is not generally necessary for aircraft to line up on the runway and wait at non-towered airports: they can just keep going from the taxiway, as they will have been checking their environs visually and by radio during taxiing.
Yes, I guess they might be able to do that. I was, however, using that as a point that it is the planes that have to check for themself, and so waiting at the runway entrance would not seem to be an artificial thing to do because of that.

I would also think there should not be modelled specific airspace control etc. In my vision, there is whatever airspace there is, and then the airport is controlled or uncontrolled. As stated somewhere (either in this thread or in the articles), the airport is within a buble inside whatever airspace is around.

Quote
Has the proportion of non-towered airports declined over time? It would be interesting to see data on this. Also, is this idea of an actual traffic limit on airports without towers not similar to my original suggestion?
I think there is a big difference between number of lines serving the airport, and number of vehicles at the same time that serves the airport. While the limitations by vehicles seems reasonable, the limitations of lines feels kind of artificial. An airport with only one line can certainly be more buisy than an airport with 10 lines! Just fill that one line with alot of vehicles and you will have a buisy airport.

Offline jamespetts

  • Simitrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 15857
  • Total likes: 407
  • Helpful: 177
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Airport control towers
« Reply #9 on: May 27, 2017, 01:51:33 PM »
I think there is a big difference between number of lines serving the airport, and number of vehicles at the same time that serves the airport. While the limitations by vehicles seems reasonable, the limitations of lines feels kind of artificial. An airport with only one line can certainly be more buisy than an airport with 10 lines! Just fill that one line with alot of vehicles and you will have a buisy airport.

The difficulty is that there is also a very big difference between how easy that it is to code for the number of lines/lineless convoys serving an airport and the number of total aircraft serving an airport. Just think of all the different places in the code where one would have to check for something that would amount to adding a new aircraft to an existing line and find a way of stopping it that makes sense to the player and cannot be worked around/exploited. Also, it is very rare for airports to have traffic on one route so frequent that only one or two routes (and I was thinking of a maximum of 2 or 3 lines for non-towered airports) would make the aircraft frequency so high as to require a tower.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.