Novice question - what is pathing time?

Peregrine 4903 · 19 Sep 2021

jfollows said:
An interesting discussion going on here: could you please explain this particular point though because I'm baffled? Thank you!

Someone like @The Planner might be able to explain it better but I'll give it a shot.

Basically you have sectional running times (SRT's) and Theoretical Running Times (TRT's). Section Running Times are the time it takes for a Model Train to travel between two timing points. So for example model train of a class 375 has a 1 minute Sectional Running Times for a Stop to Pass between East Croydon and Windmill Bridge Junction. You get SRT's for Stop - Stop, meaning the train is stopping at both of the timing points, pass to pass meaning the train is passing through both of the timing points, pass - stop and stop - pass which means the train is stopping at one of the timing points bot not stopping at the other. SRT's have been got in theory by doing timing runs on the network travelling on that specific class of train, e.g. a class 375 and then collected and inputted into a system. Once they are in that system it allow Network Rail to publish the schedules into the Downstream systems such as TRUST, and basically allows the public to see them on booking wbebsite, departure boards, realtime trains and so forth.

The issue have is that for various timing loads particuallry freight ones, there often is no Sectional Running Time between two timing points as the train may be doing a move that that model of train doesn't usually do and means that no one has gone out onto the network and got an official SRT for it. This means that the train cannot be published by Network Rail into the downstream systems. This means if you want to publish the train, you have basically two option. You can either request an SRT for that particular move, but that can sometime take a long time to do and also if its for a freight train/engineering haulage train doing a one off move at a junction say that it is unlikely to do again anytime soon, your not going to be the top priority to get an SRT request.

The other option is to use Theoretical Running Times. This is when you use the Model Train SRT's of similar class of train that does say have an SRT for that move between the two timing points. That will get rid of the Data error and allow you to publish the train downstream. The issue is of course, is that it may take the class of train you are trying to run between the two timing points longer than the model train that is running your specific train, which can cause delays.

There is lots more to it, and I probably have't explained it very well, but that is sort of the basic explanation behind it.

RailUK Forums

Wilts Wanderer · 19 Sep 2021

Deleted as Peregrine has explained it better.

Edit: although I take exception at the term ‘stop-stop’ as it’s misleading. Should be ‘start-stop’, whatever TPS might call it! :lol:

Peregrine 4903 · 19 Sep 2021

Wilts Wanderer said:
It is avoidable, in that at specific locations ARS has to be designed to only attempt to route a train a maximum of 2 mins before it’s booked time, as an alternative to the ‘muddle through and provide 2 greens for everything’ of standard ARS routing. However individual signals can’t have mixed settings, so it’s all or nothing, plus it has to be specced at the design stage, which timetabling functions don’t normally get involved with (although various people are pressing for this to be improved.) Also it needs a good advance understanding of what the post-remodelling timetable structure will be, which isn’t always possible particularly given NR’s recent record of deferring major timetable change due to internal resource constraints.

Personally, without traffic management (such as Luminate on the Western, which is excellent from all accounts), I think the wider application of ARS is foolish and a result of the badly though through ROC strategy which requires it. Someone - a signaller - once said that having ARS is like working with the worst trainee signaller ever - it just makes the same preventable mistakes again and again, day after day.

I do agree that traffic management systems really do hopefully need to be adopted. ARS just adds so many unessecary constraints not only to signaller, but also from the timetable planning perspective. I've heard similar anecdotes about ARS.

I didn't know ARS though could be designed like that though, so thankyou for letting me know.

The Planner · 19 Sep 2021

Shows i need to retire as the issue of missing SRTs never used to take so long to sort in years gone by. You don't need excessive modelling and real world data to have a good idea of what a one off move would be. Back in the days of TrainPlan (yes, ok...) you could get a SRT in the system to get your schedule sorted and as long as it was put in A/B plan shortly after to sort the upload then you were good to go.

In terms of the ARS issue, that is for NR and train planning to solve. There is absolutely no reason why it shouldnt be involved from development stage.

jfollows · 19 Sep 2021

Peregrine 4903 said:
There is lots more to it, and I probably have't explained it very well, but that is sort of the basic explanation behind it.

No, that's fine, thank you. I was happy with SRT and indeed I remember reading a book with E455 SRTs for the WCML a long time ago. I just hadn't come across "TRT" and didn't know what the first "T" stood for - now I do and you've given me more background. I'd probably refer to them as "best guess running times" myself!

Wilts Wanderer · 19 Sep 2021

The Planner said:
Shows i need to retire as the issue of missing SRTs never used to take so long to sort in years gone by. You don't need excessive modelling and real world data to have a good idea of what a one off move would be. Back in the days of TrainPlan (yes, ok...) you could get a SRT in the system to get your schedule sorted and as long as it was put in A/B plan shortly after to sort the upload then you were good to go.

In terms of the ARS issue, that is for NR and train planning to solve. There is absolutely no reason why it shouldnt be involved from development stage.

There is a short-cut process to consult and agree missing SRTs which takes about a week, and is used quite widely. As long as something can be justified, for example “it’s the same as this other traction” then no one ever really objects. I think the days are gone when a planner could just guess a value off the top of their head though. I think the problem is that too many erroneous values have been entered to BPlan without end dates and further down the line they trip something up.

Peregrine 4903 · 19 Sep 2021

The Planner said:
Shows i need to retire as the issue of missing SRTs never used to take so long to sort in years gone by. You don't need excessive modelling and real world data to have a good idea of what a one off move would be. Back in the days of TrainPlan (yes, ok...) you could get a SRT in the system to get your schedule sorted and as long as it was put in A/B plan shortly after to sort the upload then you were good to go.

In terms of the ARS issue, that is for NR and train planning to solve. There is absolutely no reason why it shouldnt be involved from development stage.

To be fair, getting most SRT's into B Plan can be pretty quick, often it can take a week, like @Wilts Wanderer says. Often they just use a an SRT for a similar model of train that does have an SRT for that move and just input that. Particularly when its moves into and out of sidings such as say Sevenoaks Siding where a 700 and 465 model train are going to have basically the same SRT in practice.

Spartacus · 19 Sep 2021

Wilts Wanderer said:
There is a short-cut process to consult and agree missing SRTs which takes about a week, and is used quite widely. As long as something can be justified, for example “it’s the same as this other traction” then no one ever really objects. I think the days are gone when a planner could just guess a value off the top of their head though. I think the problem is that too many erroneous values have been entered to BPlan without end dates and further down the line they trip something up.

About a week as a “short cut” is about a week longer than it used to take to put it into TrainPlan, and is just as much too long for much STP planning. True, there were dodgy ones in, but there were plenty of dodgy ones put in during the early TPS days, including a few negative ones, mostly down to human error.

My general rule of thumb was if the SRT was 10 minutes or less, put in up to half the SRT in pathing time, otherwise put in a stop. If for some reason this wasn’t viable, make sure suitable fiddle/adjustment time is added afterward.

ARSe really did add difficulties, with it not being able to ‘read’ pathing, or any other allowances, so it’s prone to sending trains, especially freights, out early. I know one planner who used to put in practically everything as adjustment, in part to fool ARS, and in part to fool the delay attribution software which will generate a delay if a train’s say 3 minutes late into and out of a section with 3 minutes of some kind of ‘recovery’ in the middle. Putting it all in as adjustment avoids that as the software doesn’t see that.

Bald Rick · 19 Sep 2021

Many, many years ago I was taught by the then head of train planning in Birmingham that pathing time was there to prevent:

Operationally
Substandard
Headways
In
Timetables

306024 · 19 Sep 2021

Bald Rick said:
Many, many years ago I was taught by the then head of train planning in Birmingham that pathing time was there to prevent:

Operationally
Substandard
Headways
In
Timetables

I was advised of the ’6p’s
Proper Planning Prevents P*** Poor Performance.
No doubt there are others.

Meanwhile a week is a long time in Train Planning.

Bald Rick · 19 Sep 2021

306024 said:
Meanwhile a week is a long time in Train Planning.

Certainly is when it’s 14 weeks from Christmas and there’s an informed traveller deadline to meet!

The Planner · 19 Sep 2021

Bald Rick said:
Many, many years ago I was taught by the then head of train planning in Birmingham that pathing time was there to prevent:

Operationally
Substandard
Headways
In
Timetables

How many years ago? Wasn't Mr W was it?

Spartacus said:
ARSe really did add difficulties, with it not being able to ‘read’ pathing, or any other allowances, so it’s prone to sending trains, especially freights, out early. I know one planner who used to put in practically everything as adjustment, in part to fool ARS, and in part to fool the delay attribution software which will generate a delay if a train’s say 3 minutes late into and out of a section with 3 minutes of some kind of ‘recovery’ in the middle. Putting it all in as adjustment avoids that as the software doesn’t see that.

That soon spread very quickly.

Bald Rick · 19 Sep 2021

The Planner said:
How many years ago? Wasn't Mr W was it?

25. MD.

The Planner · 19 Sep 2021

Bald Rick said:
25. MD.

Makes me feel younger

Class 170101 · 21 Sep 2021

306024 said:
I was advised of the ’6p’s
Proper Planning Prevents P*** Poor Performance.
No doubt there are others.

Yes indeed

306024 said:
Meanwhile a week is a long time in Train Planning.

Yes indeed, though time seems to pass at speed

Bald Rick said:
Certainly is when it’s 14 weeks from Christmas and there’s an informed traveller deadline to meet!

Far too early to be talking about Christmas.

Tomnick said:
Something that I've noticed seems to have crept in recently, in the context of diverted long(er)-distance passenger trains, is the insertion of stops (sometimes unadvertised, sometimes shown as public stops) at implausible intermediate stations on the diversionary route rather than using pathing time approaching a conflict. Any views from the experts?

Realtime Trains - 1R50 0437 Nottingham to Liverpool Lime Street

Unfortunately I have seen stops added like 1R50 above by NR but without the unadvertising of the stops cue arguments about whether the train should stop on the day.

Wilts Wanderer · 21 Sep 2021

Class 170101 said:
Yes indeed

Yes indeed, though time seems to pass at speed

Far too early to be talking about Christmas.

Unfortunately I have seen stops added like 1R50 above by NR but without the unadvertising of the stops cue arguments about whether the train should stop on the day.

If it doesn’t stop on the day (unadvertised or not) then the ‘value’ of putting in those stops is negated anyway. The actual need is to stop trains presenting early at ARS controlled junctions. All the NR planner has managed to do is to hide the issue, which is probably worse tbh.

Bald Rick · 21 Sep 2021

Class 170101 said:
Far too early to be talking about Christmas.

not for the train planning teams! Christmas is nearly boxed up and ready to go.

Peregrine 4903 · 21 Sep 2021

Bald Rick said:
not for the train planning teams! Christmas is nearly boxed up and ready to go.

Is it? Maybe at the old timescales, but not the current ones unless your talking exclusively about tocs.

Bald Rick · 21 Sep 2021

Peregrine 4903 said:
Is it? Maybe at the old timescales, but not the current ones unless your talking exclusively about tocs.

It’s published in a few weeks, hence the word ‘nearly’!

Peregrine 4903 · 21 Sep 2021

Bald Rick said:
It’s published in a few weeks, hence the word ‘nearly’!

Still a long way to go though as its not been validated yet.

Eloise · 21 Sep 2021

Peregrine 4903 said:
Someone like @The Planner might be able to explain it better but I'll give it a shot.

Basically you have sectional running times (SRT's) and Theoretical Running Times (TRT's). Section Running Times are the time it takes for a Model Train to travel between two timing points. So for example model train of a class 375 has a 1 minute Sectional Running Times for a Stop to Pass between East Croydon and Windmill Bridge Junction. You get SRT's for Stop - Stop, meaning the train is stopping at both of the timing points, pass to pass meaning the train is passing through both of the timing points, pass - stop and stop - pass which means the train is stopping at one of the timing points bot not stopping at the other. SRT's have been got in theory by doing timing runs on the network travelling on that specific class of train, e.g. a class 375 and then collected and inputted into a system. Once they are in that system it allow Network Rail to publish the schedules into the Downstream systems such as TRUST, and basically allows the public to see them on booking wbebsite, departure boards, realtime trains and so forth.

The issue have is that for various timing loads particuallry freight ones, there often is no Sectional Running Time between two timing points as the train may be doing a move that that model of train doesn't usually do and means that no one has gone out onto the network and got an official SRT for it. This means that the train cannot be published by Network Rail into the downstream systems. This means if you want to publish the train, you have basically two option. You can either request an SRT for that particular move, but that can sometime take a long time to do and also if its for a freight train/engineering haulage train doing a one off move at a junction say that it is unlikely to do again anytime soon, your not going to be the top priority to get an SRT request.

The other option is to use Theoretical Running Times. This is when you use the Model Train SRT's of similar class of train that does say have an SRT for that move between the two timing points. That will get rid of the Data error and allow you to publish the train downstream. The issue is of course, is that it may take the class of train you are trying to run between the two timing points longer than the model train that is running your specific train, which can cause delays.

There is lots more to it, and I probably have't explained it very well, but that is sort of the basic explanation behind it.

This isn’t my understanding of how TRTs work. TRTs are calculated by TPS using its behind the scenes infrastructure model. Based on the timing loads acceleration and braking curves and the coded in speeds and gradients of the line in question. Only a limited number of Planners and Specialists have this ability on their logon and most, but not all, are in the freight teams.

Your definition of TRT is where you do a change-en-route (CeR). Or a fudge. Certainly caught TOCs doing this morphing trains from DMU to HST and back, and from EMU to DMU and back to get around missing SRTs. Tut tut. They were spoken to quite severely.

You also have IRT (Indicative Running Times) which are proposals from advanced and modelling work ahead of consultation.

(I’ll do a post-post edit and say what I put above is what TRT has meant since TPS was introduced in c2010.)

Peregrine 4903 · 21 Sep 2021

Eloise said:
This isn’t my understanding of how TRTs work. TRTs are calculated by TPS using its behind the scenes infrastructure model. Based on the timing loads acceleration and braking curves and the coded in speeds and gradients of the line in question. Only a limited number of Planners and Specialists have this ability on their logon and most, but not all, are in the freight teams.

Your definition of TRT is where you do a change-en-route (CeR). Or a fudge. Certainly caught TOCs doing this morphing trains from DMU to HST and back, and from EMU to DMU and back to get around missing SRTs. Tut tut. They were spoken to quite severely.

You also have IRT (Indicative Running Times) which are proposals from advanced and modelling work ahead of consultation.

(I’ll do a post-post edit and say what I put above is what TRT has meant since TPS was introduced in c2010.)

Ah yes, sorry that's my mistake, I always get the two mixed up.

Class 170101 · 22 Sep 2021

Bald Rick said:
not for the train planning teams! Christmas is nearly boxed up and ready to go.

Far too much time spent talking about Christmas in that job.

Horizon22 · 22 Sep 2021

Wilts Wanderer said:
It is avoidable, in that at specific locations ARS has to be designed to only attempt to route a train a maximum of 2 mins before it’s booked time, as an alternative to the ‘muddle through and provide 2 greens for everything’ of standard ARS routing. However individual signals can’t have mixed settings, so it’s all or nothing, plus it has to be specced at the design stage, which timetabling functions don’t normally get involved with (although various people are pressing for this to be improved.) Also it needs a good advance understanding of what the post-remodelling timetable structure will be, which isn’t always possible particularly given NR’s recent record of deferring major timetable change due to internal resource constraints.

Personally, without traffic management (such as Luminate on the Western, which is excellent from all accounts), I think the wider application of ARS is foolish and a result of the badly though through ROC strategy which requires it. Someone - a signaller - once said that having ARS is like working with the worst trainee signaller ever - it just makes the same preventable mistakes again and again, day after day.

ARS is good if used correctly with traffic management as you say. As someone who sees it used on the other side (Control) you can see where it's been left in ARS incorrectly/lazily by the signaller through either ad-hoc moves left standing or trains just blasting into terminals 4-5E (ignoring the pathing time) and delaying a RT departure

Novice question - what is pathing time?

Established Member

RailUK Forums

Established Member

Established Member

Veteran Member

Established Member

Established Member

Established Member

Established Member

Veteran Member

Established Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Established Member

Established Member

Veteran Member

Established Member

Veteran Member

Established Member

Member

Established Member

Established Member

Established Member