Train Planning TIPLOCs

[Martin Swanson]

Hi 

Is anyone actually using the location data from the Train Planning feed where the TIPLOC code does not refer to a station ?

The data seems to be virtually useless. It does not seem to be used on Peter's http://www.opentraintimes.com website either.

Is there a reliable location data source for TIPLOCs that are not stations ?

Regards,

Martin

[Matthew Macdonald-Wallace]

naptan data from Atoc is a better bet, see the wiki for more information.

Matt

--
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 post to this group, send email to openrail...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Martin Swanson]

I have NapTAN, but the RailReferences.csv file only has the stations. I am looking for non-station location data.

Martin

[Chris Northwood]

Basically, no. There's a crowdsourcing effort to get the data onto OpenStreetMap though: http://nrodwiki.rockshore.net/index.php/AddingJunctionsAndSidingsToOsm, but it's got fairly low coverage atm

[Martin Swanson]

In which case, why not delete the LOC records in the Train Planning file?

[Matthew Macdonald-Wallace]

Ah, yeah, that's an unsolved problem as far as I'm aware :(

[Shaun Flynn]

Hello there folks,

There are other online resources for TIPLOCs and other bits and bobs.

This is a good one:

http://www.railwaycodes.org.uk/CRS/CRSa.shtm

It is A to Z gazeteer (I think that's the right word) of every known railway location.

Shaun

[Matthew Macdonald-Wallace]

That site has been an excellent resource for hunting down missing TIPLOCs etc however it doesn't have any GPS data (which may or may not be a problem for you!)

Matt

[petermount]

I'll also add my own as well which went live last week: http://uktra.in/Api lists the available endpoints but http://uktra.in/Api/StationPosition might be of interest as that has geographical coordinates for most stations - I use it to present a local map for each station.

It's based on the geographical list posted here a while back but I've made some mods to it including associating TIPLOC's to them.

Not all tiplocs are in there yet & I need to handle stations that have multiple tiplocs (e.g. London Victoria) but it's getting there.

Peter

[Martin Swanson]

Thanks Peter.

I use the lat/long for stations direct from NaPTAN, this seems complete, accurate and is maintained.

Do you have non-station TIPLOC lat/long? This is my gap.

Martin

[Shaun Flynn]

I think the problem maybe that it is easy to get a GPS read for each railway station when the train has stopped.

Much harder to get a GPS read for an intermediate location from a moving train?

[Marcus]

I have been doing this via a very old school method of maps, books and
general research. I suspect, as mentioned, the station information is
easy to come by but the junctions are hard and the signals are well on
nigh impossible without some 'inside knowledge'/cab ride videos.

I've mapped just over 22% of the list:
(http://www.trainspots.co.uk/codes-index.php)
but it is slow work. This one:

Location Stanox Stanme CRS nlc TIPLOC
Baileyfield C.E. 4637 BAILEYFLD 932462 PORTBCE


Took about 5 minutes to work out as I looked it up in a rail atlas, then
looked it up on the google map and then found it. It's in Edinburgh for
what it's worth.

But with over 12,000 dots to map I've got a LONG WAY to go.



One question I have that does arise from this thread is what will/why do
people need different geographical locations that are a few meters apart
or indeed the same place? For example London Bridge has over 10
TIPLOCS, 4 STANOX, 3 STANME and one CRS code associated with it.
Platforming might be handy but I can't see why one would differentiate a
lat long between platforms 2 and 3 which are probably the same piece of
concrete. On the other hand, Clapham Junction Overground and Clapham
Junction Brighton lines could be a useful differentiation.

Just my curiosity

Cheers,

Marcus

On 04/01/2015 11:01, Shaun Flynn wrote:
> I think the problem maybe that it is easy to get a GPS read for each
> railway station when the train has stopped.
>
> Much harder to get a GPS read for an intermediate location from a moving
> train?
>
> On 4 Jan 2015 10:58, "Martin Swanson" <martind...@gmail.com

> <mailto:martind...@gmail.com>> wrote:
>
> Thanks Peter.
>
> I use the lat/long for stations direct from NaPTAN, this seems
> complete, accurate and is maintained.
>
> Do you have non-station TIPLOC lat/long? This is my gap.
>
> Martin
>
> > On 4 Jan 2015, at 10:37, petermount <peter...@gmail.com

> <mailto:peter...@gmail.com>> wrote:
> >
> > I'll also add my own as well which went live last week:
> http://uktra.in/Api lists the available endpoints but
> http://uktra.in/Api/StationPosition might be of interest as that has
> geographical coordinates for most stations - I use it to present a
> local map for each station.
> >
> > It's based on the geographical list posted here a while back but
> I've made some mods to it including associating TIPLOC's to them.
> >
> > Not all tiplocs are in there yet & I need to handle stations that
> have multiple tiplocs (e.g. London Victoria) but it's getting there.
> >
> > 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

> <mailto:openraildata-t...@googlegroups.com>.

> To post to this group, send email to openrail...@googlegroups.com

> <mailto:openrail...@googlegroups.com>.

[petermount]

I do occasionally use NaPTAN but didn't use it this time around. I've got it but not yet exposed it.

As for non-station's no I've not got those, but for most of what I'm using it for I don't need them.

[petermount]

On Sunday, 4 January 2015 12:28:06 UTC, Marcus Dawson wrote:

One question I have that does arise from this thread is what will/why do
people need different geographical locations that are a few meters apart
or indeed the same place?  For example London Bridge has over 10
TIPLOCS, 4 STANOX, 3 STANME and one CRS code associated with it.
Platforming might be handy but I can't see why one would differentiate a
lat long between platforms 2 and 3 which are probably the same piece of
concrete. On the other hand, Clapham Junction Overground and Clapham
Junction Brighton lines could be a useful differentiation.

I'm not certain why other's need to have them, I don't personally need that although possibly for those stanox entries that are reported in the trust feed could be useful - an example I can think of right now would be Otford which isn't in the trust feed but Otford Junction which is to the south (& other side of the M26) is. Knowing that location woult then be useful as I could then use that to present delays visually on a map.

peter

[Peter Hicks]

On 04/01/15 12:27, Marcus wrote:
> One question I have that does arise from this thread is what will/why
> do people need different geographical locations that are a few meters
> apart or indeed the same place? For example London Bridge has over 10
> TIPLOCS, 4 STANOX, 3 STANME and one CRS code associated with it.
> Platforming might be handy but I can't see why one would differentiate
> a lat long between platforms 2 and 3 which are probably the same piece
> of concrete. On the other hand, Clapham Junction Overground and
> Clapham Junction Brighton lines could be a useful differentiation.

The further you drill down in to the "Where is a TIPLOC/STANOX?"
problem, the more difficult it gets.

Consider the network of TIPLOCs and the network of STANOXes as a graph
(as in node/edge, rather than 'chart'). Trying to align this with the
actual geography on the ground is tricky:

* The TIPLOC 'WATFDJ' (Watford Junction) refers to platforms 6-11 at
Watford Junction. Platform 11 is some way from the main lines. Do you
put the point mid-way along platforms 6-9, and somewhere between
platforms 7-8? Or do you put the point further toward platforms 10 and 11?
* The TIPLOC 'WATFJDC' (Watford Junction DC Lines) refers to
platforms 1-4 (there is no 5) at Watford Junction. These are adjacent,
however where do you put the point? Do you put it between platforms 2
and 3, mid-way down the platform? It's a terminal platform, so it may
make sense to put it at the buffer stops, but you might want it at the
end of the platform.
* The TIPLOC 'EUSTON' (London Euston) refers to the whole of London
Euston. Platforms here are different lengths. Do you put the point
somewhere in the middle of the area covered by the platforms?

Assuming you've put the TIPLOC for London Euston somewhere near
platforms 8-11, if you're working with a track-level diagram, then the
TIPLOC is 'locked' to a single line. However, if you're working with
centre-line geometry (such as the OS Meridian2 data), you can snap this
point to the nearest part of the closest line.

There is no one solution which will satisfy everyone. The way the
Corporate Network Model (CNM) in Network Rail works is to use TIPLOC and
STANOX lines. This allows you to position a TIPLOC over multiple lines
- in the case of Watford Junction, the TIPLOC line may stretch from
mid-way down platform 6 all the way across to platform 9 perpendicular
to the railway, then continue at a different angle to cross platforms 10
and 11. Regardless of whether you're working with a centre-line or
track-level diagram, you can locate a TIPLOC at any intersection.

Unfortunately, and I want to challenge this, I've been able to get very
little traction on getting any geospatial data released, so we may have
to make our own model until such time as the wind changes - as it has
done with Darwin.


Peter

[Matthew Macdonald-Wallace]

The idea behind www.lunchtimetrains.co.uk is that it brings up all the services running through a given point on the network for spotters to spot.

At the moment, stations are fine, but there are some yards etc which are good for spotters but I can't recommend them automatically because I don't know where they are in relation to the the user's location.

I'm not massively concerned with GPS data for every TIPLOC/stanox but having it for each "area" (as opposed to individual platforms or shunting roads) would be great!

Matt

On 4 Jan 2015 12:28, "Marcus" <g...@rail-net.co.uk> wrote:

I have been doing this via a very old school method of maps, books and general research.  I suspect, as mentioned, the station information is easy to come by but the junctions are hard and the signals are well on nigh impossible without some 'inside knowledge'/cab ride videos.

I've mapped just over 22% of the list:
(http://www.trainspots.co.uk/codes-index.php)
but it is slow work. This one:

Location                Stanox  Stanme          CRS     nlc     TIPLOC
Baileyfield C.E.        4637    BAILEYFLD               932462  PORTBCE


Took about 5 minutes to work out as I looked it up in a rail atlas, then looked it up on the google map and then found it. It's in Edinburgh for what it's worth.

But with over 12,000 dots to map I've got a LONG WAY to go.



One question I have that does arise from this thread is what will/why do people need different geographical locations that are a few meters apart or indeed the same place?  For example London Bridge has over 10 TIPLOCS, 4 STANOX, 3 STANME and one CRS code associated with it. Platforming might be handy but I can't see why one would differentiate a lat long between platforms 2 and 3 which are probably the same piece of concrete. On the other hand, Clapham Junction Overground and Clapham Junction Brighton lines could be a useful differentiation.

Just my curiosity

Cheers,

Marcus



On 04/01/2015 11:01, Shaun Flynn wrote:

I think the problem maybe that it is easy to get a GPS read for each
railway station when the train has stopped.

Much harder to get a GPS read for an intermediate location from a moving
train?

On 4 Jan 2015 10:58, "Martin Swanson" <martind...@gmail.com

<mailto:martindswanson@gmail.com>> wrote:

    Thanks Peter.

    I use the lat/long for stations direct from NaPTAN, this seems
    complete, accurate and is maintained.

    Do you have non-station TIPLOC lat/long? This is my gap.

    Martin

     > On 4 Jan 2015, at 10:37, petermount <peter...@gmail.com
    <mailto:peter...@gmail.com>> wrote:
     >
     > I'll also add my own as well which went live last week:
    http://uktra.in/Api lists the available endpoints but
    http://uktra.in/Api/StationPosition might be of interest as that has
    geographical coordinates for most stations - I use it to present a
    local map for each station.
     >
     > It's based on the geographical list posted here a while back but
    I've made some mods to it including associating TIPLOC's to them.
     >
     > Not all tiplocs are in there yet & I need to handle stations that
    have multiple tiplocs (e.g. London Victoria) but it's getting there.
     >
     > 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-talk+unsubscribe@googlegroups.com
<mailto:openraildata-talk+unsub...@googlegroups.com>.
To post to this group, send email to openraildata-talk@googlegroups.com
<mailto:openraildata-talk@googlegroups.com>.

For more options, visit https://groups.google.com/d/optout.

--
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-talk+unsubscribe@googlegroups.com.
To post to this group, send an email to openraildata-talk@googlegroups.com.

[Marcus]

The succinct answer to my reply is 'it all depends on the level of
detail you require', the long version is below.

Watford is a good example and it has the oddness of Platform 11 being a
different northbound destination to 6-10.

As you say I suspect it depends on the level of accuracy one would
require. There is no tiploc/stanox for Watford Junction Platform 9. So
as far as a tiploc/stanox system is concerned the working is at Watford
junction. A geographic blob at Watford would probably work.

If you need to differentiate platforms then in most examples GPS
wouldn't really be helpful as people can read signs and the like and the
accuracy of gps might be off for differentiating platform 9 and 8.

From my point of view if it gets you to within 20 meters then it will work.

> Assuming you've put the TIPLOC for London Euston somewhere near
> platforms 8-11, if you're working with a track-level diagram, then the
> TIPLOC is 'locked' to a single line. However, if you're working with
> centre-line geometry (such as the OS Meridian2 data), you can snap this
> point to the nearest part of the closest line.

If you are working to vector type information then I'd snap to lines..
Unless you're in a siding looking for a train then GPS won't be much
use. If you know your train is in the down carriage sidings then you
can walk to them. If it's in Basford Hall Yard, road 9 then you can find
it.

If you are planning signal maps based on geographic data then your map
will be based around fixed paths between two points. Like the OTT maps
but with all the twists and turns in them.

>
> There is no one solution which will satisfy everyone.

That is the key point.

> The way the
> Corporate Network Model (CNM) in Network Rail works is to use TIPLOC and
> STANOX lines. This allows you to position a TIPLOC over multiple lines
> - in the case of Watford Junction, the TIPLOC line may stretch from
> mid-way down platform 6 all the way across to platform 9 perpendicular
> to the railway, then continue at a different angle to cross platforms 10
> and 11. Regardless of whether you're working with a centre-line or
> track-level diagram, you can locate a TIPLOC at any intersection.
>
> Unfortunately, and I want to challenge this, I've been able to get very
> little traction on getting any geospatial data released, so we may have
> to make our own model until such time as the wind changes - as it has
> done with Darwin.

I suspect that a lot of this data does not exist.


Looking at location: Signal D 230
Stanox: 88733
Stanme: HOOJN1611
CRS:
nlc: 521014
TIPLOC: HOOJ230

It will be on signalling diagram that will put it between 'x' and 'y'
and might even say which line/road it is on.

There might be a distance from mp 0 to it listed somewhere.

But if you wanted to, proverbially, drop a smart b0mb on it you would
need to know where it was in the Hoo Junction area.


For my purposes if one can point to it on googly maps and be in the
right 20 meters then it works for me. And that is the logic I've used
when mapping locations for pages like this (made this seven years ago!!):
http://www.trainspots.co.uk/working-report.php?working=381

The closer you zoom in it becomes more obvious this is 'dot to dot' but
it should give most people a clue as to where the working is going. A
progression would be vector lines tied to specific points, but that is a
whole new bag of work.

Cheers,

Marcus

[Juhani Pirttilahti]

Hi all,

I have done some research and you can quite easily find most railway locations from Sectional Appendixes and Timetable Planning rules. Of course, this will require some manual placing. 

It's a tedious task to map thousands of locations by hand (total number being almost 10700 tiplocs!). However, only about 6700 of them are actively used and about 3000 of them are on passenger stations so this leaves us with about 3700 locations to be placed by hand. Unfortunately I can't see any better approach as the current Tplan LOC data is a bit fuzzy.

So i ignored it and started building my own dataset of locations. Currently I have a collection of 3337 tiplocs (includes 2527 stations from Meridian 2) and I'm willing to share this data. It's currently very basic: tiploc code and it's coordinates, but could be extended to cover also unique stanoxes. This data certainly will be posted to the Wiki at some point.

But let's talk about the model first. Any suggestions?

Also, there might be some locations where stanox points and tiploc points should be mapped separately. In some situations, a stanox code seems to cover larger area than a tiploc. Where the location should be when a single stanox covers multiple tiplocs? In the center of an area covered, or maybe at a station if there is one - or at a platform?

On Sunday, 4 January 2015 16:18:20 UTC+2, Peter Hicks wrote:

The further you drill down in to the "Where is a TIPLOC/STANOX?"
problem, the more difficult it gets.

Consider the network of TIPLOCs and the network of STANOXes as a graph
(as in node/edge, rather than 'chart').  Trying to align this with the
actual geography on the ground is tricky:

[Chris Northwood]

The OpenStreetMap model for CRS codes might be a good starting point for the model - it has nodes (single points), ways (lines), closed ways (eg, buildings areas) and relations (when multiple nodes or ways have some shared characteristic).

So, some TIPLOCs (eg, junctions or signals) may be nodes, others which are stations may be closed ways, sidings could be ways (corresponding to the railway track), tiplocs which correspond to multiple platforms might be a relation between the multiple ways that represent each platform.

--
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 post to this group, send email to openrail...@googlegroups.com.

[Martin Swanson]

Peter - Who have you engaged on opening up location data? has anyone spoken to NaPTAN to request adding more stop point types to their data set? Or to the ODI? 

I noticed some other sources:

1. DfT recently released rail line data

https://certificates.theodi.org/datasets/3041/certificate

2. geonames.org has concept called "feature" - lots of them for railroad and includes lat/long for UK "railroad"  locations

http://www.geonames.org/export/codes.html

Feature types:

RSD	railroad siding	a short track parallel to and joining the main track
RSGNL	railroad signal	a signal at the entrance of a particular section of track governing the movement of trains
RSTN	railroad station	a facility comprising ticket office, platforms, etc. for loading and unloading train passengers and freight
RSTP	railroad stop	a place lacking station facilities where trains stop to pick up and unload passengers and freight

RJCT	railroad junction	a place where two or more railroad tracks join
RR	railroad	a permanent twin steel-rail track on which freight and passenger cars move long distances

Martin

[Peter Hicks]

Hi Martin

On 04/01/15 23:47, Martin Swanson wrote:

Who have you engaged on opening up location data? has anyone spoken to NaPTAN to request adding more stop point types to their data set? Or to the ODI?

Network Rail's Geospatial team.  TIPLOCs and STANOXes come and go, track geometry changes (e.g. Watford South and Watford North junctions, Sundon Loop) happen and taking data straight from the horse's mouth ensures it's as up-to-date, accurate and consistent as possible.

NR also have other datasets, e.g. the location of each signal, the associated TD berth and other data - but when I worked with this data last Autumn, it was really out-of-date, still showing Bletchley PSB (BY) signals.

I've also been in contact with Ordnance Survey about their products, and NR don't feed them any data - they build it themselves.

How could the DfT or ODI help?  There are licensing and 'derived data' issues here, plus, I imagine, a desire to cover the costs of generating the geospatial data (even though it's required for operating the railway).

I noticed some other sources:

1. DfT recently released rail line data

https://certificates.theodi.org/datasets/3041/certificate

This centre-line data was added back in 2012, and I made an FoI request in October of that year for the data in vector format (see https://www.whatdotheyknow.com/request/map_of_the_railway_network), which didn't go anywhere.


Peter

[Juhani Pirttilahti]

Hi Peter,

Well the centre-line data certainly is available in vector format. I just typed in the WFS address to QGIS and received the whole network in vector format. It actually does have kind of useful information to work with. For example ELR codes and mileages are included.

WFS access is provided at this URL: http://inspire.misoportal.com/geoserver/transport_direct_railnetwork/wfs

[Peter Hicks]

Hi Juhani

> On 5 Jan 2015, at 08:57, Juhani Pirttilahti <juhani.pi...@gmail.com> wrote:
>
> Well the centre-line data certainly is available in vector format. I just typed in the WFS address to QGIS and received the whole network in vector format. It actually does have kind of useful information to work with. For example ELR codes and mileages are included.
>
> WFS access is provided at this URL: http://inspire.misoportal.com/geoserver/transport_direct_railnetwork/wfs

It’s available in vector format now - it wasn’t back in 2012 when it was hosted on a non-Geoserver… the only thing that was available was a set of rendered bitmaps.


Peter

[Dave Butland]

Martin, 

I'm using the NaPTAN data for stations but I'm also looking to locate the non-station locations using the network link data (distance between tiploc nodes) along with the track geographic locations to try and locate them more accurately. It does seem a little odd in this day and age to have such inaccurate geographic information or think that not having a geolocation for a geographic object is sensible. 

I will happily publish this data and make it available for update if people want to help. 

Dave.

[Ben Woodward]

The problem is that it's not a "geographical object" but more often
several objects, or an area, how do you define it then?

[Dave Butland]

The short answer, without wanting to stray too far off topic, is to use topological identifiers like those used by the Ordnance Survey in their Mastermap product. It is a real shame that the "Integrated Transport Layer" in this product set only covers roads and footpaths. The railway data is just in the topography layer but does include a complete connected vector model of the rail network and all rail transport sites. 

I'm not sure what the Open Street Map data uses for referencing locations but I'm sure something similar is used to reference complex polygons and compound shapes. 

[Peter Hicks]

On 5 Jan 2015, at 14:34, Dave Butland <dgbu...@gmail.com> wrote:

> The short answer, without wanting to stray too far off topic, is to use topological identifiers like those used by the Ordnance Survey in their Mastermap product. It is a real shame that the "Integrated Transport Layer" in this product set only covers roads and footpaths. The railway data is just in the topography layer but does include a complete connected vector model of the rail network and all rail transport sites.

I think I see why the ITN layer only covers roads and footpaths - there are a lot more people driving cars and walking around than there are driving trains, so several orders of magnitude more eyeballs on the data!

Even if the ITN layer contained detailed railway data, the licence costs for it are not insubstantial.


Peter