News:

The Forum Rules and Guidelines
Our forum has Rules and Guidelines. Please, be kind and read them ;).

Loading times

Started by jamespetts, August 22, 2011, 11:32:18 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

jamespetts

With Experimental 10.0 forthcoming, and its feature of being able to see vehicles' loading times in the same measure as journey times, I have noticed that the loading times of many vehicles are simply too long. For example, in Pak128.Britain-Ex 0.8, the BR Class 455 EMU has a loading time of 0:57 with the default distance scale of 250 meters/tile. However, when I timed one this morning, its stop to start loading time was closer to 0:20 (albeit at a relatively quiet station). Having excessive loading times is likely to distort the timings and all that depend on them somewhat; I do, therefore, need to correct these issues.

All those who know or have an idea of vehicles' dwell times (any sort of vehicle that features in Simutrans, from steam trains to jet planes, from trams to canal barges), I should be very grateful for any insight posted in this thread.
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.

ӔO

#1
some hints would be their intended use and the amount of doors they have.

express trains usually only have 1 or 2 doors at the edges to reduce cabin noise
commuter trains usually have 2 or more doors and are placed inboard, which allow easy loading and unloading, but they are louder as a result.

more recent developments have wider doors and all lengthwise seats that allow even faster loading and unloading.

Loading time also depends on how many people need to get on and off too so schedules are harder to keep during peak rush hour and there usually is more time allowed for it. Although, alternatively, the operator would choose an EMU or DMU that accelerates like a rocket.


my experience riding on TTC subway, which uses 23m, 4 door cars says anywhere between 10s and 30s, depending on congestion. JR commuter is anywhere between 10s and 60s and over, but 60s is when they are stuffing the train.

my experience with JR express trains say somewhere between 20s and 60s, but sometimes up to 4mins, as the train waits for an even faster express to go by. At the terminals, it's extra long, like upwards of 10mins, but no more than 20mins, as the janitors clean up and replace the head rest covers.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

Carl

Passenger planes seem to have a minimum loading time of around half-an-hour (perhaps a few minutes less). This seems to be how long quick-turnaround airlines like Ryanair and Easyjet allow between an arrival and a departure. Any extra time is either timetable-related or "padding".

As for railways: as AEO mentioned, loading times here seem to depend less on the kind of vehicle and more on how many people are leaving/joining the train. Here's an example. CrossCountry Voyagers can load in less than a minute at quiet stations, but at stations like Leeds where almost everyone gets off and hundreds of people get on, loading can take up to five minutes. (Obviously the kind of vehicle is also relevant for the reasons that AEO mentioned -- my point is that it's not clear it's the main factor in how long loading takes.)

I'm not sure whether this is already reflected in Simutrans loading behaviour. If it's not, then in the long term it would be good if volume of loading, as well as the kind of vehicle, was a factor in calculating loading times. Note that *unloading* is important here, too: even if nobody's joining a train, if lots of people are leaving this will take a couple of minutes.

The Hood

It should be relatively easy to categorise pax rail vehicles into several categories of loading times and these should be based on door configurations.  Most current UK stock is either end-door (long dwell time) or 1/3+2/3 double doors (shorter dwell times).  WIthin the second category, suburban metro trains (e.g. 376, 378) have simpler door mechanisms and wider vestibules which allow even shorter dwell times.  LUL (and some trams) have even more doors per train with even shorter dwell times.

jamespetts

Thank you all for your responses. There are already loading times encoded in Pak128.Britain-Ex, as can be seen in the depot window, and they all take account of the factors mentioned by AEO and The Hood. The relative loading times, in approximate terms, are already there, but they are out in absolute terms, so it would help me to know actual figures. (Carl - as to the five minute figure for the Cross Country Voyagers - is this really all loading time, or do you think that it might be timetable padding? I rather suspect that they could turn it around in no more than two minutes even at a crowded station if they really wanted).

Presently, Simutrans-Experimental does not change loading times based on the number of people boarding/alighting. If this were to be implemented, how best (1) to set the ranges for each individual vehicle; (2) to calculate the ratio between passengers/goods boarding/alighting/being loaded/being unloaded and where in the range that the loading time falls; and (3) to display this information to the player in the depot window?

Carl, as to aircraft: these are more complicated, as the boarding process is much slower. If the loading time was half an hour, then this would allow passengers/cargo arriving at the airport at 29:59 to board a second before departure. Should we take into account the fact that people have to arrive at the airport much earlier, and thus add a substantial amount to the waiting times irrespective of actual loading times? Does this figure apply for very small aircraft used, for example, in small islands? Thoughts on how to handle aircraft would be much appreciated.

Finally, does anyone have any idea of realistic loading times for lorries/mail vans/canal barges/ferries/cargo ships?
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.

Carl

#5
Quote from: jamespetts on August 23, 2011, 08:38:36 AM
(Carl - as to the five minute figure for the Cross Country Voyagers - is this really all loading time, or do you think that it might be timetable padding? I rather suspect that they could turn it around in no more than two minutes even at a crowded station if they really wanted).

My experience is that at busy times (which is most of the time on these routes!) loading can take up to five minutes. Even when these trains are running late it can take this long -- suggesting that it can't all be down to timetable padding. The same is true, incidentally, of Transpennine services from Leeds to Manchester. Of course, this is definitely at the upper end of the scale -- I wouldn't expect there to be any examples of loading taking longer than this. Next time I travel to Leeds I'll gather some more precise data.
 
Quote
How best (1) to set the ranges for each individual vehicle; (2) to calculate the ratio between passengers/goods boarding/alighting/being loaded/being unloaded and where in the range that the loading time falls; and (3) to display this information to the player in the depot window?

Clearly each vehicle should have:
(a) an upper limit on loading time -- how long it would take to empty the whole convoy and fill it again
(b) a lower limit -- how long a convoy would stop if nobody was either alighting or boarding

Between these ranges there should presumably be a sliding scale. Say that a convoy can carry 100 passengers. The extreme case is 100 passengers alighting and 100 loading -- and this corresponds to the upper limit of loading time. If only 50 passengers alight and 50 load, we would expect the loading time to be at the exact midpoint between the upper limit and the lower limit. If 50 load and none alight, the loading time will be one-quarter of the way between the lower limit and the upper limit. Etc.

I'm aware that this doesn't directly answer your question about how these limits/scales should be set...  :)

As for information displayed in the depot: I would have thought it sufficient to display just the lower limit and the upper limit. But you could also display a third figure -- e.g. how long it would take for, say, 20 passengers to alight and 20 passengers to board.

Quote
Carl, as to aircraft: these are more complicated, as the boarding process is much slower. If the loading time was half an hour, then this would allow passengers/cargo arriving at the airport at 29:59 to board a second before departure. Should we take into account the fact that people have to arrive at the airport much earlier, and thus add a substantial amount to the waiting times irrespective of actual loading times? Does this figure apply for very small aircraft used, for example, in small islands? Thoughts on how to handle aircraft would be much appreciated.

Two options I can think of immediately. The first corresponds with your suggestion: have a minimum waiting time for aircraft which reflects the fact that one cannot simply turn up to an airport immediately before departure. The second is to allow (say) 15 minutes for aircraft loading before putting the plane into a further state, between loading and departure, for 15 minutes. This would mean that passengers who turn up in this 15 minute period won't be able to board the plane -- as in real life. However, it would also presumably add 15 minutes to the journey time of the plane, unless this time could be counted as waiting time rather than travelling time.

I'm afraid I've no idea how this all works with small aircraft.

ӔO

for smaller aircraft, like a cesna, there's a long pre-flight check list, refuelling and what not, which results in about 15 to 30mins before it is allowed to take off.

For larger aircraft, there can be a few places where time is added. They will often park at the apron to do preflight checks and refuel. Then they will head over to the terminal gates, where they are loaded up. Loading from the ground takes about the same, maybe a bit longer, compared to loading from the building arm. Depending on facilities, it's possible to load and unload from 2 doors instead of one.

One very important thing to consider, is that a lot of express train services died out, because of competition with air. The trains only started to make a come back when their faster speeds could compete with boarding times and costs of air travel. A lot of local lines also had trouble keeping competition with express buses.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

jamespetts

Thank you all for your input. I did some timings to-day on my ordinary journeys, and found that a two doored double decker 'bus seems had a stop-to-start dwell time of between 20 and 40 seconds in conditions ranging from light to quite busy (the latter being a standing room only 'bus where perhaps about 1/5th - 1 quarter of passengers boarded/alighted, the former being a 'bus with plenty of spare seated capacity in which only one or two people boarded/alighted), and that deep level Underground ("Tube") trains dwelt between 15 and 30 seconds depending on how busy that they were (again, ranging from a train with seats to spare at a station in which only a few people boarded/alighted to a train where a number of people were standing in each carriage and quite a few people were getting on). I have not had a chance to test these modes of transport in very crowded conditions, however.

As to air transport, it seems to me that it is somewhat unique in being impossible to board the vehicle very shortly before it is due to leave, irrespective of the turnaround time: it is always necessary to spend at least half an hour in the airport checking one's baggage in and going through security checks before boarding. I do wonder whether allowance should be made for adding a user-specifiable amount to the waiting times of passengers/cargoes boarding aircraft to simulate this? Otherwise, even if we give aircraft a very, very long start-to-stop dwell time, some passengers/goods would inevitably arrive just before it was due to leave, thus reducing the waiting time.
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.

Carl

One idea I had for gathering data on this without leaving the armchair is to scour YouTube for videos of vehicles loading. For instance, this video of a tube at Tottenham Court Road: http://www.youtube.com/watch?v=h3QISN1Zs3c. The train seems fairly busy here. Doors open at 0:13, and close at 0:40; a total of 27 seconds loading.

So far I haven't managed to find any videos of mainline trains loading. There are lots of videos of trains, of course, but most people seem only to film them either arriving or leaving a station, and not during their dwell time. I'll keep looking, because this could be a useful source of data.

Junna

Quote from: carlbaker on August 24, 2011, 08:07:49 AM
So far I haven't managed to find any videos of mainline trains loading. There are lots of videos of trains, of course, but most people seem only to film them either arriving or leaving a station, and not during their dwell time. I'll keep looking, because this could be a useful source of data.


I have a lot of those full-length or nearly-so cab-view videos, and judging from the London Euston to Liverpool service hauled by a Class 86 in 1986, dwell time at Watford Junction at 33 seconds, Milton Keynes at 55 seconds and a time-tabled dwell time at Crewe for 1 minute 30 seconds.

Class 223 express from Kyoto to Kobe at heavy traffic has the following dwell times along the route: Takatsuki 50s, Shin-Osaka 41s, Osaka Central 2m 25s, Amagasaki 1m 12s, Sannomiya 43s.

Carl

I travelled on a Class 323 in south Manchester today, and at quiet stations (only a few people joining/alighting), the doors were only open for around 15 seconds. I suspect these 323s are a good model for commuter trains more generally -- I suspect that the lower bound of the loading times for such trains should be set at around 15 seconds.

jamespetts

Thank you very much both of you for your useful input! I travelled on a class 442 to-day from Croydon to Victoria, and its dwell at Croydon was over a minute, although I stopped measuring precisely at about a minute, as the dwell was artificially lengthened by somebody trying to get on when the doors were closing. Incidentally, I keep meaning to get around to drawing a class 323.
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.