Signal Berth Locations

[GB]

New to Train Data but most interesting.  Open Train Times has a great schematic of Signal Berths. The corresponding TD API seems to work very well and lines up. 

Question is there any information out there on the geographic location of  each signal/berth  lat/long or OSGR   

Hope someone can help 

Thanks 

George 

[Peter Hicks]

Hi George

A berth doesn't cover a particular point on the railway, it covers an area.  That area then depends on where the train has come from, as a step from WM510 to platform 1 st Euston will cover a different area to a step from WM306 to platform 1.  Not vastly different, but different enough.

There may be data out there (not open) with the latitude and longitude of individual signals, but mapping those to berths is a not insignificant exercise (see https://www.rssb.co.uk/research-catalogue/CatalogueItem/T1246 which I wrote a couple of years ago).

You'd be better off using GPS data from trains, which will correctly show the train's last reported position, speed and heading.  Trying to interpolate things, as signalbox.io does, is fine for a rough position, but the more you zoom in, the less accurate it can get.

 

Peter

--
You received this message because you are subscribed to the Google Groups "A gathering place for the Open Rail Data community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openraildata-t...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/openraildata-talk/8c83426c-2b76-4b58-bfac-70415c76dc90n%40googlegroups.com.

[GB]

Thanks, Peter 

The centre point of the berth areas would be good enough to provide a rough visualisation,  at a push in some areas it might be possible to roughly 

work something out from the open train times schematics,  I was just hoping there was a simple set of lat, longs somewhere. 

My understanding is that there is no GPS data available from trains (is that correct)   in an ideal world they would be tracked and the head and tail of each train would be visible   

When I first started looking at this it seemed odd they weren't given all buses are tracked in the UK. 

I am talking to Signalbox and trying to get access to their data anyway ( not open,  commercial product) but looking at it, it does seem to have issues and is missing freight traffic. 

George 

[Peter Hicks (Poggs)]

Hi George

> On 18 Nov 2024, at 11:31, GB <gbk...@gmail.com> wrote:
>
> The centre point of the berth areas would be good enough to provide a rough visualisation, at a push in some areas it might be possible to roughly
> work something out from the open train times schematics, I was just hoping there was a simple set of lat, longs somewhere.

That’s my point - there is no ‘centre’. You would be better off taking the position of the signal head itself and snapping a train in a berth to that, and being open and honest that you’ll have discrepancies.

If the latitude and longitude of berths were available, you can guarantee that I’d be showing it on OpenTrainTimes and pointing people to the data source :-)

> My understanding is that there is no GPS data available from trains (is that correct) in an ideal world they would be tracked and the head and tail of each train would be visible
> When I first started looking at this it seemed odd they weren't given all buses are tracked in the UK.

There’s a lot of GPS data sent out by trains - a lot. However it’s not openly available for a number of reasons, nor does it show where the ‘head’ and ’tail’ of each train is. GPS antennas could be in the middle of the middle carriage of a unit, they could be at the ‘A’ cab end, they could be at the ‘B’ cab end, or they could be a third of the way down the carriage behind the ‘A’ end. Units may be coupled in multiple, and you might have both ‘A’ ends next to each other so get two GPS fixes that are about 5m from each other, or the ‘B’ ends may be coupled together and you could have two GPS fixes a fair distance from each other. Unless you know that unit 1 is coupled to unit 2, and which way around they are (see trains may have this data available, but it may not be accessible outside the train), I don’t think you can even begin to work out the extent of track occupied by a train.

I think it’s probably important to say that train describer berths reflect the occupation of a section of track, and say nothing about where a train is within that section, or how fast it’s travelling, or whether it’s stopped.

A bit of a crude analogy, but think of a berth like a toilet cubicle - you don’t care if the person inside has just started, is just finishing, or is doing Sudoku and will be quite a long time - you just care whether the cubicle is free.


Peter

[James Uppington]

I did an amount of mapping of some of the lines I am interested in while on the GWR “Flying Banana" HST farewell tour a few years ago.

I wrote a noddy app which takes the GPS position, knowledge of the train it’s currently on and (crudely) matches the berth step time to roughly work out the lat/long where a train steps out of a given berth.

It gives me a reasonable location of the next signal a train will pass and have written some basic dynamic KML to draw trains onto Google Earth. As others have said though, it’s not particularly accurate and is limited to the areas I’ve been out and surveyed, mostly on said rail tour. I keep meaning to go out and extend my coverage, but it’s not that high on my priority list!

If anyone wants it and is interested, https://trackit.uppyjc.co.uk/TrackIT/KML.aspx can be added to Google Earth Pro as a "Network Link” to evaluate the vague possibilities.

From James Uppington

upp...@me.com

--
You received this message because you are subscribed to the Google Groups "A gathering place for the Open Rail Data community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openraildata-t...@googlegroups.com.

To view this discussion, visit https://groups.google.com/d/msgid/openraildata-talk/1710DD93-473D-4A93-A507-23E2204C56B9%40poggs.co.uk.

[Seb Dazeley]

Something to consider as well is that on plain line there is quite often one track section covering both the approach to a signal and the overlap. So as there is no track circuit start/end (or axel counter) at the signal itself, the berth step occurs at the start of the new track circuit, which would be at the end of the overlap. 

In practical terms this means the step occurs around 200 yards after passing the signal. Or immediately if there is a seperate track circuit for the overlap. Good luck finding that data! 

To view this discussion, visit https://groups.google.com/d/msgid/openraildata-talk/DCA25F1C-A2C1-492F-8B32-EAC458B71166%40me.com.

[GB]

Hi Peter 

All makes sense and actually as expected really,  the aspiration was just to translate the berth transition to something geographical, however rough,  to drop into google maps at some level alongside buses and ferries 

Didn't know there was GPS data,  perhaps eventually the DfT will find a way to make it available,  the way they did for Buses  (BODS) 

And great website by the way 

James and Seb,  thanks for the insights.    

James, great idea and if need be I could easily send someone out with one of my tracking / web logging apps on trains to gather info 

George  

[Matthew Burdett]

I think CIGP is the gps platform via National Rail it just isn't public (yet)

To view this discussion, visit https://groups.google.com/d/msgid/openraildata-talk/83fb4d08-8e92-4042-8293-933bb99ed2a6n%40googlegroups.com.

[GB]

Hi Mathew 

My application is not strictly public,  its for cities and councils (transport authorities) 

Do you know who this is made available to at the moment 

Thanks 

George 

[Matthew Burdett]

Hi

As per Peter's Wikipedia site athttps://wiki.openraildata.com/index.php?title=National_Rail_Data_Portal

The customer information gps portal powers the Darwin reports for train movements and I'm guessing this is what Peter spoke about earlier about the GPS updates from trains.

It's not live to the public, because of operator concerns, but it is a highly demanded Want on the rail data marketplace and could become available later in time.

To view this discussion, visit https://groups.google.com/d/msgid/openraildata-talk/3c1387d3-26e9-4d3f-9c92-1f0ba45e4d89n%40googlegroups.com.

[Bigbigcheese]

Hello all,

Sorry to bring up a long deceased thread, but is the general consensus that there isn't a (public facing) way to link all of the signal "names" that you'd see on the post e.g. SN113 (which I used cab ride videos to place here: https://www.openstreetmap.org/node/12588327641) with their corresponding C code berth? i.e. "D3 0113". (Obviously the 113 is the same but that's it).

And that secondly, there isn't a public GPS feed to find the positions of (any part of) the trains?

Thanks

[Peter Hicks]

On Monday, 19 May 2025 at 20:42, Bigbigcheese <bigbigc...@gmail.com> wrote:

Sorry to bring up a long deceased thread, but is the general consensus that there isn't a (public facing) way to link all of the signal "names" that you'd see on the post e.g. SN113 (which I used cab ride videos to place here: https://www.openstreetmap.org/node/12588327641) with their corresponding C code berth? i.e. "D3 0113". (Obviously the 113 is the same but that's it).

That is still the case - there is not a 1:1 mapping between a berth and a signal, as some signals can have more than one berth, some berths (dual berths) can have more than one signal, some berths can have no signals, and some signals can have no berths.

However, that's just given me an idea, and I'll have a conversation with some people over the next few days to see if we can get some 'helpful' data released.

And that secondly, there isn't a public GPS feed to find the positions of (any part of) the trains?

Correct - there are industry feeds, but they are very high throughput and not available to the public for a variety of reasons.

Peter

[Evelyn Snow]

Hi Big Big Cheese,

Afraid to say that you're correct on both counts. There's not any straightforward/universal way
to match berths to signal identification plates. The RDG reneged on releasing CIGP and we've not
heard anything more about that since then.

Evelyn

[Thomas Wood]

Generally for an area's XX123 signal, it can be assumed that it will have berth 0123 associated. However mapping the signal prefix letters (or signal number ranges where the prefix scheme is route wide) is a harder problem without access to the source information other than by observation of the feed.

 Generally on any one signalling workstation, any 4 digit signal number will be unique regardless of prefix letters, to avoid chance of confusion.

Another common exception to the rule is that occasionally a signal box's berths will be relayed to the data feeds via an adjacent signal box's link to smart. In this case, you will usually get berths of the form X123, where X is a suitable letter for the relayed signal box. Eg, D1 (Reading area) and FH (Feltham) used to report all of Wokingham's signal berths with a W prefix letter.

Be aware that there are always exceptions to the general conventions. 

Regards,

Thomas Wood 

To view this discussion, visit https://groups.google.com/d/msgid/openraildata-talk/3609cc82-8a6a-4aae-83db-fa391433924bn%40googlegroups.com.

[Bigbigcheese]

Thanks for the quick responses all.

I'm intruiged by what the idea was Peter, and what 'helpful' data might come of it.

With what Thomas said about exceptions to every rule in mind - Will each C coded message area (i.e. D3) always be within one signal box area (i.e. SN)? i.e. is the mapping one way, will none of the signals in D3 ever be in the T area? (It's an example, I'm using D3/SN because I know it well enough but I'm sure A1/A2/A3 into X, B1-B5 into Y and C1 into Z makes the same amount of sense).

Kind regards,

Joe

[Juhani Pirttilahti]

Hi all,

I do actively maintain a dataset of my own on platforms and signals and their associated berth(s) with their geographic coordinates. I've built my thing over the NWR Track Model, snapping the signals to the tracks. This dataset is behind of a realtime map application I've been developing over the years. My sources for signals so far have been the great ITPS model (which isn't always correct), Bing/Google satellite imagery, a FOI request of a list of signals with their locations on the network (ELRs and mileages, not coordinates as they won't release that) and the scheme plans from the TD mailing list I was sadly kicked out. I've been wondering if I can publish this data through an API (and possibly the locations of trains, therefore). But, there is always a but, Network Rail isn't keen to release information where their signals are exactly and there is this FOI security exemption in place regarding the coordinates. I don't like to breach security even though it is more like security through obscurity. Also, it has taken me hundreds of hours to compile this dataset. But I've still been thinking can I put this dataset on RDM for example, maybe for a small cost?

Speaking of GPS data I've seen that the DfT has the Bus Open Data Service with an API endpoint for GTFS-RT Vehicle Positions messages. Why not have something similar for trains then?

Just my two cents.

[Evelyn Snow]

This exists, it's called CIGP (Customer Information GPS Portal). It was announced in 2018 that it'd
be published on the NRDP, this was never done. It's available internally as I understand it, but
some TOCs apparently don't like the idea of their trains exact positions being available in
real-time, which means we get none of it.

Evelyn

[Matthew Burdett]

Never really understood the reasoning behind it being hidden as TD also gives locations (although not as fine) which, as many have done, Juhani, signaboxes.io and can be pretty easily worked out to a location just very time consuming.

Hopefully when train operators see their numbers out in the wild a bit more and see what the lovely people here do with their data they'll be more towards it

--
You received this message because you are subscribed to the Google Groups "A gathering place for the Open Rail Data community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openraildata-t...@googlegroups.com.

To view this discussion, visit https://groups.google.com/d/msgid/openraildata-talk/aEIeF6o2q0usqjeh%40evelyn.moe.

[Peter Hicks]

On Friday, 6 June 2025 at 06:01, Matthew Burdett <matthewbu...@gmail.com> wrote:

Hopefully when train operators see their numbers out in the wild a bit more and see what the lovely people here do with their data they'll be more towards it

Exactly this.  Initially, NROD only had Class 1, 2 and 5 trains in the TD and TRUST feeds, until I proposed including other train classes and removing the 'sensitive' reporting numbers of other trains.  After years of nothing happening, we now have details of all train movements in the CIF and everything (at least I believe so) in the TD feeds.

Similarly, at one time ATOC (former senior management, nobody current) asserted strongly that Darwin would 'never' be available as an Open Data feed because there's too great a risk somebody will display the data incorrectly resulting in loss of revenue for train operators.  And only a few years later, we have NRDP.

Peter

[Bigbigcheese]

Juhani,

I think I am aiming to do the same thing as you, w.r.t mapping the locations of "all the trains", though I'm using OpenStreetMap railway nodes/ways instead of the NWR track model. I'm a big fan of open source data and plan to slowly work my way around the UK adding the signal locations to OSM as I strongly believe that this is where it belongs, plus it works within my usecase and would allow others more competent than me to do the same. It sounds as if your dataset would be very useful for open sourcing this data and I'm wondering if, despite your reservations about NR being unhappy, and wishing to be reimbursed for your efforts you'd be at all interested in sharing via OSM?

Thanks,

Joe

[Juhani Pirttilahti]

Joe,

I'm familiar with OSM and I like the idea. I'll definitely look into it.

Cheers,

Juhani