Page 1 of 2
| Ideas on multiple dwell times 06/09/2017 at 18:56 #101581 | |
GeoffM
6380 posts |
With ever more complex timetables, the issue of dwell times (red signal move-off, station stop, divide, crew change, etc) and how they apply is coming up more often. One change that will go ahead is how crew changes are handled. Currently this is a tick box per location, but in the future it will be handled as an activity (like Next, Divide, Share, etc). That way the TT author can decide whether it happens before or after other activities. Existing timetables will be supported by turning a "tick" into an activity, probably inserted last or prior to the first "next" activity (there should only be one). I would like a consensus on how dwell times should be handled when a train has multiple activities. I will list some obvious combinations but add more if appropriate. In each case, should the dwell times be additive, or the largest of the two numbers, or what? 1. Divide and a Next. Example: divide is 4 minutes; Next is 2 minutes; should the total dwell be (4+2)=6 or Max(4,2)=4, or what? 2. Divide+Divide+(others). Is the time taken to divide into three portions double that of two portions, the same, or somewhere in between? 3. Divide+Crew Change+Next 4. Join+Crew Change+Next 5. Crew Change+Next 6. Divide+Crew Change (for original train that continues) 7. Join+Crew Change 8. Divide+Join+Crew Change (does the 2nd crew have to wait for the 1st to finish running around, for example) Ignoring Divide+Join and Join+Join as that shouldn't be possible without something vacating the platform/siding and another train coming in, which is a completely separate delay. SimSig Boss Log in to reply |
| Ideas on multiple dwell times 07/09/2017 at 00:02 #101591 | |
JamesN
1611 posts |
If the ability to either combine or add the dwell times together is being looked into, why not add in "AND" and "THEN" activities. Any list of activities separated by "ANDs" takes all those activities; picks the longest dwell time from them and uses that (2,4,2=4) THEN acts as a separator to say do this, THEN once that dwell time has elapsed do that. DivideRear (5) THEN Crew Change (1) THEN Join (4) AND Next (2) would thus total 10 minutes minimum, and wait for replacement crew before doing join/next: 5 Minutes for DivideRear As long as replacement crew ready then 1 Minute for CC 4 Minutes for Join AND Next (4,2=4) Total 10 minutes. Perhaps AND should be default behaviour to not break existing behaviour (thus every list of activities has implied "AND"), but users could then make use of AND and THEN to force the dwell times to combine (or not) the way they want? Last edited: 07/09/2017 at 00:05 by JamesN Reason: Spelling and slight re-word Log in to reply |
| Ideas on multiple dwell times 07/09/2017 at 06:18 #101593 | |
kaiwhara
587 posts |
I'll review in more detail but it would be useful to also allow the variable of whether a different crew performs an activity (like a split) from the crew who actually brought the train in. HST's for example have different "reversal" values dependent on whether it is the same crew simply changing ends, or whether the return working is taken by someone else. Also, an out and back shunt could be a driver at both ends, which cuts the time to reverse direction down considerably depending what the stock actually is.
Sorry guys, I am in the business of making people wait! Log in to reply |
| Ideas on multiple dwell times 08/09/2017 at 21:28 #101630 | |
58050
2660 posts |
Depends Geoff on how you want to do it, either in the sense of fantasy or what actually happened? In BR days the crew change took place first irrespective of what else had to take place. Footplate crews back then wanted to finish & get off home(the biggest fiddle going was how you could get off duty asap within your day). Therefore if a train was required to carry out any station shunts etc, then crew going forward would be the ones doing it. Certainly not the crew who had worked the train in. This was pretty much the case on the freight side even since privatisation as the majority of the footplate crews were ex BR the attitude still stayed within the footplate grades. Therefore anything other than that for hertitage tts just wouldn't be real. Usually the forward crew would have booked on duty well before the train that they were releiving arrived. If the train was running early it would be held back somewhere prior to it coming forward for traincrew releif. This is what happened at Landor Street Jn & similarly at Bedford. Any train requiring releif at Bedford would be held at Bedford North Jn until the TCS contacted West Hampstead PSB to state that the forward crew had booked on duty & were ready. Even if there was some station shunting to be done the crew bringing the train in wouldn't do it. They would either be going home or would be off for there PNB & ready for there return working. Another reason in BR days that this was the case was due to the drivers not wanting to be stuck on the loco carrying out any shunting & the forward crew waiting for the loco to be back in place for the crew change to take place. The crew change ALWAYS came first irrespective of what needed doing.
Log in to reply |
| Ideas on multiple dwell times 08/09/2017 at 21:33 #101631 | |
postal
5269 posts |
Isn't this all complicating a relatively simple concept? Where is the actual problem if the dwell time is a single entity not related to anything else which defines the minimum time the train may stop at a location? If the TT writer wants to overload the stop with actions so the total time of the actions exceeds the minimum dwell time specified then so what? The dwell time specification has been met and anything beyond that is down to whatever the TT writer has specified. Looking at the converse view, is the assumption that the dwell time should be related mathematically to the times of the various actions? If that is the case, as a TT writer can I ask for an additional "minimum platform time" variable that the TT writer can specify in order to avoid rule after rule of the type "Train XYZ must not depart from A until N minutes after Train XYZ arrives at A", please. “In life, there is always someone out there, who won’t like you, for whatever reason, don’t let the insecurities in their lives affect yours.” – Rashida Rowe Last edited: 08/09/2017 at 21:45 by postal Reason: None given Log in to reply |
| Ideas on multiple dwell times 08/09/2017 at 21:42 #101635 | |
|
58050
2660 posts |
Further to the above, crew changes weren't henerally long affairs. The WTT states 2 minutes, but usually this was much quicker. The forward crew would meet the crew that had just brought the train in would meet on the platform or beside the track & the usual exchange in words between the drivers went something like this:- driver off the incoming train would say "You've got 36 on & she's(referring to the loco) a good one(or she's low in power) or something simialr if the inwards driver thought the loco wasn't performing. Forward driver would say "OK thanks mate" & the forward driver would give the driver who came off the train a master key as he would have left the master key in the driving desk on the loco. The only time I witnessed something different to that was when I was secondman on the Northfleet coal & the driver I was booked with was a scab who had worked in the 1982 strike. The in-coming crew sabotaged the loco by pressing down the emergency brake plunger. No handover was given as they wouldn't speak to scab drivers. The driver who was booked to work the train forwarded then had to waste 5 minutes or so tryig to figure out why he couldn't get the train brake to release. Fortunately for me on that trip I rode in the back cab with the guard as there was a road learner booked on the train & he sat in the front. The guard knew what had happened & he didn't say anything as he was the NUR rep & I'd been warned by other drivers at the depot not to speak to him or else they wouldn't speak to me. Fortunately I spend a week in the back cab with the guard playing cards on our way to Northfleet in the rear cab of the Cl.58. I've never forgot that week as half the dpeot told me not to give the road learner my seat in the front & the other half of the depot said if they saw me talking to either of them they wouldn't speak to me anymore. I took the easy option & sat in back out of the way. Needless to say the rest of the week they(the forward driver & road learner) obviously knew what to look out for after that. But that kind of thing also happened all over the country. When I worked at Bescot there were some drivers there with issues with other driver dating back to the 1955 strike. There was no time for cit chat in BR days. You could chit chat all you liked when you saw other footplatecrews in the messroom, but once your train arrived it was a case of sooner you get on & go the sooner you'd get home. Log in to reply |
| Ideas on multiple dwell times 08/09/2017 at 21:48 #101636 | |
|
GeoffM
6380 posts |
Unfortunately, Pascal, that is not how the railway works these days. I've had several drivers/traincrew tell me it entirely depends on the service as to whether the dismounting crew do any (un)coupling or the new crew. That is why I am giving you the flexibility of deciding where you want to have it in the order of activities. For backward support, that crew change "tick" has to turn into a list item of activities, so does it go as the first entry (per your description) or last (per these days)? No right answer but it's got to go *somewhere* so a default position at least doesn't break old timetables. The TT author can come back and change the order at their leisure. Crew changes also are a lot slower these days. More cab equipment, more paperwork. But again you're in control, so what worked for you in the 1970s won't work for a 2017 TT and vice versa. Complicating? Not overly so. Flexibility, which is what people keep asking for. You can't have flexibility without complicating things a bit. If the location dwell time should just override the whole lot then perhaps that is an idea for discussion with all the issues that potentially raises. SimSig Boss Log in to reply |
| Ideas on multiple dwell times 08/09/2017 at 22:01 #101639 | |
|
postal
5269 posts |
In which case my bid is for whatever you wish to do with the concept of dwell times to make a coherent system but can the TT writers have a separate independent "minimum platform time" variable, please.
“In life, there is always someone out there, who won’t like you, for whatever reason, don’t let the insecurities in their lives affect yours.” – Rashida Rowe Log in to reply |
| Ideas on multiple dwell times 08/09/2017 at 22:15 #101640 | |
|
58050
2660 posts |
Geoff if your saying the tt writer decides when adding in the activity or setting the dwell times for a crew change. Personally I'm more interested in getting it right for the BR period, I'm not bothered about todays railway & no doubt others would say the opposite to me. What I don't really want to see is something similar to when the Join & Detach facility was first put in where everything took approx. 4 mins to complete & as a result everything in the Carlisle 1979-1980 timetable departed a few minutes late. I'm not bothered which way it's done providing it can suit both periods. For me in BR days the crew change happened first before anything else & that was certainly the case on the freight side(EWS) when I was trainsmaster at Bescot. Drivers would be very reluctant to do any shunting whether it was at Bescot or Peterborough or Toton when they arrived there with there train. The main reason for this was that they could end up in the wrong end of the yard to get to there return train or engine. Yes I accept what you say about there being more cab equiptment todeal with these days & as for the passenger side I can't really speak for that as I never dealt with that side of the railway in privatisation, but no doubt takes longer as there's more beaurocracy involved today. In the timetable I'm currently writing for Noel when a loco arrives at a point where the crew changes ends I've actually cut that down from 2mins to 1min as the secondman would be in one end driving & the driver would be in the other end driving, so as soon as one driving desk is closed & the other opened the loco is ready to roll & that could be done within 30 seconds. Obviously not on a Cl.66 though as there is only one key per loco & not a master key where every footplate crew memeber had one or 2 or more of them.
Log in to reply |
| Ideas on multiple dwell times 08/09/2017 at 22:48 #101642 | |
|
GeoffM
6380 posts |
postal in post 101639 said:In which case my bid is for whatever you wish to do with the concept of dwell times to make a coherent system but can the TT writers have a separate independent "minimum platform time" variable, please.Can you explain what you mean as it seems you're not talking about the custom dwell time. 58050 in post 101640 said: Geoff if your saying the tt writer decides when adding in the activity or setting the dwell times for a crew change. [...] What I don't really want to see is something similar to when the Join & Detach facility was first put in where everything took approx. 4 mins to complete & as a result everything in the Carlisle 1979-1980 timetable departed a few minutes late.Are you not using the custom dwell times? They override the default value which ranges from 2 to 4 minutes. SimSig Boss Last edited: 08/09/2017 at 22:54 by GeoffM Reason: None given Log in to reply |
| Ideas on multiple dwell times 08/09/2017 at 22:55 #101643 | |
Soton_Speed
285 posts |
58050 in post 101635 said:When I worked at Bescot there were some drivers there with issues with other driver dating back to the 1955 strike.Across the cab the driver sits and he's staring straight ahead. He's not spoken to me for eighteen months since I started at the shed. He's drawn a line across the cab for each of us to stand Since he found I'd joined the NUR while he was a ASLEF man. - EIGHT FREIGHT BLUES - Dave Goulder In Zone 6, no one can hear you scream... Log in to reply |
| Ideas on multiple dwell times 09/09/2017 at 00:04 #101645 | |
|
postal
5269 posts |
GeoffM in post 101642 said:Can you explain what you mean as it seems you're not talking about the custom dwell time.I'm am talking about a custom dwell time but not in terms of how you are defining that option. I am talking about a custom dwell time such as was written into the Royal Mail/EWS contract in 1996. In that instance every station stop had a defined platform time (called "dwell time" in the contract) which was the minimum time the train had to be available at the platform for Royal Mail purposes. Whatever the railway had in their TT was irrelevant beyond that "dwell time". That obviously does not align with what you are viewing as dwell time and is probably the reason for all of this discussion. If SimSig does not take a "dwell time" as the time a train must be stopped at a location, then there is a need for a a rule for every stop which has a an operationally required minimum platform time or different entity (maybe "minimum platform time"). This would be applicable in any era as even in eras after Pascal's preferred time zone, a TOC or FOC may require a defined stand time outwith any other operating considerations. Immaterial in the overall scale of things, but the option would be nice rather than adding to the multitude of rules that already exist in complicated TTs. “In life, there is always someone out there, who won’t like you, for whatever reason, don’t let the insecurities in their lives affect yours.” – Rashida Rowe Log in to reply |
| Ideas on multiple dwell times 09/09/2017 at 00:12 #101646 | |
|
GeoffM
6380 posts |
postal in post 101645 said:GeoffM in post 101642 said:In that case I'm not sure we're talking about different things at all. The SimSig dwell time is the minimum time that a train will spend at a station. Taking the simplest case of no activities, no rules in place, and a simple passenger train that leaves stations on time, and has a custom dwell time of 10 minutes, and a scheduled arrival of 09:00 and departure of 09:02. If the train arrives at 09:00 then it will depart at 09:10. If the train arrives at 08:45 then the train will depart at 09:02. Arrival at 09:15 would mean departure at 09:25.Can you explain what you mean as it seems you're not talking about the custom dwell time.I'm am talking about a custom dwell time but not in terms of how you are defining that option. I am talking about a custom dwell time such as was written into the Royal Mail/EWS contract in 1996. In that instance every station stop had a defined platform time (called "dwell time" in the contract) which was the minimum time the train had to be available at the platform for Royal Mail purposes. Whatever the railway had in their TT was irrelevant beyond that "dwell time". That obviously does not align with what you are viewing as dwell time and is probably the reason for all of this discussion. If the above is not what you want then I'm still confused. SimSig Boss Log in to reply |
| Ideas on multiple dwell times 09/09/2017 at 00:17 #101647 | |
Steamer
3997 posts |
postal in post 101645 said:GeoffM in post 101642 said:That's already covered by the Dwell Time box in the 'Add/Edit Location' window, isn't it? You can enter any value there to define the minimum time the train is stopped at that station. It overrides the specified Station Forward/Station Reverse values.Can you explain what you mean as it seems you're not talking about the custom dwell time.I'm am talking about a custom dwell time but not in terms of how you are defining that option. I am talking about a custom dwell time such as was written into the Royal Mail/EWS contract in 1996. In that instance every station stop had a defined platform time (called "dwell time" in the contract) which was the minimum time the train had to be available at the platform for Royal Mail purposes. Whatever the railway had in their TT was irrelevant beyond that "dwell time". Returning to the main topic- all I ask is that whatever method is chosen is fully documented, as the present interaction (or intended interaction) between different dwell times has never been written up. This has added substantially to the confusion. "Don't stress/ relax/ let life roll off your backs./ Except for death and paying taxes/ everything in life.../ is only for now." (Avenue Q) Log in to reply |
| Ideas on multiple dwell times 09/09/2017 at 02:01 #101648 | |
|
postal
5269 posts |
GeoffM in post 101646 said:In that case I'm not sure we're talking about different things at all. The SimSig dwell time is the minimum time that a train will spend at a station. Taking the simplest case of no activities, no rules in place, and a simple passenger train that leaves stations on time, and has a custom dwell time of 10 minutes, and a scheduled arrival of 09:00 and departure of 09:02. If the train arrives at 09:00 then it will depart at 09:10. If the train arrives at 08:45 then the train will depart at 09:02. Arrival at 09:15 would mean departure at 09:25.But we are most definitely talking about different things because things don't work the same once you move beyond your simplest case. If you refer to Mantis you will see #9713 which I posted on 28/10/13 which Karl (Headshot119) duplicated as #17837 on 05/09/17. Both bugs refer to the fact that the dwell time clock does not start ticking until any actions at the location have been completed. You specifically closed my bug by telling me that is the way that dwell times were meant to work. Therefore I, Karl and anyone else who has worked in an operational capacity has a different view of dwell time to the definition you are applying. In summary, the definition of the SimSig dwell time as the minimum time that a train will spend at a station has drawn 2 reports on Mantis because it is the minimum time the train will spend at a station after any activities have been completed. This is emphatically not the minimum time the train will spend at a station if the time spent on activities is added in. Whichever way this debate pans out, there is still the need for a TT writer to be able to specify a minimum station time inclusive of any activities which is not the way it currently works. “In life, there is always someone out there, who won’t like you, for whatever reason, don’t let the insecurities in their lives affect yours.” – Rashida Rowe Last edited: 09/09/2017 at 10:24 by postal Reason: None given Log in to reply |
| Ideas on multiple dwell times 09/09/2017 at 08:58 #101653 | |
|
postal
5269 posts |
GeoffM in post 101581 said:...Join+Join as that shouldn't be possible without something vacating the platform/siding and another train coming in, which is a completely separate delay.What about the situation which sometimes arises at the end of a day of multiple EMU or DMU sets terminating one after the other in the same platform then forming a single working to the depot. Off the top of my head I think that happens in the Edinburgh 2006 TT and may happen in the original Lime St TT as well. Edit: And Divide + Join; through platform - train splits with original train remaining as front section and rear portion as new train. Further train then arrives at front and joins to original train. “In life, there is always someone out there, who won’t like you, for whatever reason, don’t let the insecurities in their lives affect yours.” – Rashida Rowe Last edited: 09/09/2017 at 09:59 by postal Reason: None given Log in to reply |
| Ideas on multiple dwell times 09/09/2017 at 09:33 #101654 | |
|
58050
2660 posts |
That's certainly the case at Birmingham New Street BR era timetables with DMUs heading back to Tyseley T&RSMD at the end of the day in several cases all running as 5H50.
Log in to reply |
| Ideas on multiple dwell times 09/09/2017 at 17:10 #101668 | |
|
GeoffM
6380 posts |
postal in post 101648 said:But we are most definitely talking about different things because things don't work the same once you move beyond your simplest case.No, we're not. I don't agree with myself for that original resolution, hence my confusion. Thanks for pointing it out and sorry for the confusion. I will turn it into a minimum overall dwell time, not one that excludes activity time. SimSig Boss Log in to reply The following user said thank you: postal |
| Ideas on multiple dwell times 09/09/2017 at 17:12 #101669 | |
|
GeoffM
6380 posts |
postal in post 101653 said:GeoffM in post 101581 said:What I mean is there is an enforced delay by the signalling system before the subsequent trains enter the platform for the join so the times for the joins are separated by an external factor. Ignore the vacating part....Join+Join as that shouldn't be possible without something vacating the platform/siding and another train coming in, which is a completely separate delay.What about the situation which sometimes arises at the end of a day of multiple EMU or DMU sets terminating one after the other in the same platform then forming a single working to the depot. Off the top of my head I think that happens in the Edinburgh 2006 TT and may happen in the original Lime St TT as well. postal in post 101653 said: Edit: And Divide + Join; through platform - train splits with original train remaining as front section and rear portion as new train. Further train then arrives at front and joins to original train.Same, later train will have to wait for the call-on/shunt aspect. SimSig Boss Log in to reply The following user said thank you: postal |
| Ideas on multiple dwell times 09/09/2017 at 18:36 #101673 | |
|
postal
5269 posts |
Going back to where we started, I've been thinking about Crew Changes as an action. Is there any point in putting the crew change anywhere in the list except as the last action? Presumably if the crew change is higher up the list and the new crew is not yet available, all actions below the crew change will be held until the crew change has taken place. However, if the crew change is the final item on the list and the new crew is available before the list of actions has worked through to the final item there should be no impact (in the same way as two section due to join could be buffered up to each other but do not join until the actions in the list above the join have been completed). Or have I missed something? “In life, there is always someone out there, who won’t like you, for whatever reason, don’t let the insecurities in their lives affect yours.” – Rashida Rowe Log in to reply |
| Ideas on multiple dwell times 09/09/2017 at 18:38 #101674 | |
|
postal
5269 posts |
GeoffM in post 101668 said:postal in post 101648 said:Apologies if I was a little intemperate in trying to get my point across earlier in the topic.But we are most definitely talking about different things because things don't work the same once you move beyond your simplest case.No, we're not. I don't agree with myself for that original resolution, hence my confusion. Thanks for pointing it out and sorry for the confusion. “In life, there is always someone out there, who won’t like you, for whatever reason, don’t let the insecurities in their lives affect yours.” – Rashida Rowe Log in to reply |
| Ideas on multiple dwell times 09/09/2017 at 19:35 #101676 | |
|
GeoffM
6380 posts |
postal in post 101673 said:Going back to where we started, I've been thinking about Crew Changes as an action. Is there any point in putting the crew change anywhere in the list except as the last action?It can't be after any Next activities though. As to *when* it typically takes place, well, there is no "typical". Sometimes it's before a join/divide, sometimes it's after. My references to which position earlier relate to the conversion of a tick box into an indexed activity - the author will still be able to move it around as they please. With parallelism I'm wondering if a "per activity" override is going to be needed too. For example, if the author thinks adding a crew change won't affect the total time for joins/divides, then the override could be set to zero for that activity - but it would still wait for the crew to arrive, if applicable. SimSig Boss Log in to reply The following user said thank you: Meld |
| Ideas on multiple dwell times 09/09/2017 at 19:49 #101677 | |
|
postal
5269 posts |
I was thinking more in terms of making a sim run to time without input from the user in terms of moving things up and down the actions list. I suspect there are quite a few users who are not comfortable with editing a TT; for those users pending any "per activity" override the only sensible option would be to make the crew change the last item (subject to any N action of course). Looking at the parallelism concept, do you envisage that would allow a detach at both ends at the same time (loco change at the front, detach of a portion at the rear being able to take place simultaneously rather then sequentially)? That could be helpful. “In life, there is always someone out there, who won’t like you, for whatever reason, don’t let the insecurities in their lives affect yours.” – Rashida Rowe Log in to reply |
| Ideas on multiple dwell times 09/09/2017 at 20:19 #101678 | |
|
Steamer
3997 posts |
I've outlined a proposal below. This proposal assumes that the current method of working out times is thrown out completely, and starts from a clean sheet of paper. Train Type dwell times are specified as now. The 'Edit Location' box contains a 'Custom Dwell Time' box as now. Crew changes are an 'Activity', as proposed by Geoff. Each activity may have a custom 'activity dwell time' specified, which overrides the normal value for the train type. Divides, Joins and Crew Changes (with their associated dwell times) are executed sequentially: Crew changes with late crew will delay any further activities listed. These times will sum to give a value, which I'll call 'A'. If 'A' is less than the specified dwell time for Station Forward, Station Reverse, Terminate Forward or Terminate Reverse (depending on the train's future activities), the dwell times for Station Forward, Station Reverse etc. must elapse before the train can depart. These are timed from when the train 'arrives' at the station. Station Forward/Station Reverse values can be overridden using the 'Dwell Time' box in the 'Edit Location' window. Terminate Forward/Terminate Reverse can be overridden using the 'activity dwell time' box. This should cover everything- if you want to simulate two simultaneous divides, use the 'activity dwell time' box to half the normal divide time. If you don't care where the crew change happens, dump it as the last thing before the Next activity (if there is one), and the train will depart as soon as the crew arrives and changes over, subject to the other dwell times having elapsed. Want to simulate passengers not being able to board until the divides/joins have completed? Specify an 'activity dwell time' value for Station Forward etc. that's however many minutes you want longer than the total divide/join time. For the convenience of users, and to avoid breaking timetables, existing timetables would have 'crew change' moved to the last thing before a Next activity, if there is one; TT writers can move it if they wish. Discuss! "Don't stress/ relax/ let life roll off your backs./ Except for death and paying taxes/ everything in life.../ is only for now." (Avenue Q) Last edited: 09/09/2017 at 20:20 by Steamer Reason: None given Log in to reply The following user said thank you: TimTamToe |
| Ideas on multiple dwell times 09/09/2017 at 21:36 #101680 | |
|
postal
5269 posts |
Could the "crew change" be used when there is a train in sim that has a long layover (e.g. a unit that sits at the buffers for a couple of hours before departing on its next service). In that sort of instance, there would be no physical crew change handover, but the train would not be able to depart until the new crew had arrived.
“In life, there is always someone out there, who won’t like you, for whatever reason, don’t let the insecurities in their lives affect yours.” – Rashida Rowe Log in to reply |