-
Our new ticketing site is now live! Using either this or the original site (both powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!
Passenger train consists as Open Data
- Thread starter Poggs
- Start date
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R
RailUK Forums
trainmania100
Established Member
I've seen a few trains where one/two units start a service, and then along the way, both units are removed and a completely different train continues the trip
As an example:
156 and 163 to reading, and then 166 only to Paddington
www.realtimetrains.co.uk
live.rail-record.co.uk
In fact there was a 166 service I can't remember where possibly a Cardiff to somewhere. The unit changed 3 times along the route !
Do TOCs actually expect passengers to get up out of their seats and swap units multiple times on long journeys?
== Doublepost prevention - post automatically merged: ==
As an example:
156 and 163 to reading, and then 166 only to Paddington
Realtime Trains | 1K09 0919 Newbury to London Paddington | 24/09/2025
Real-time train running information for 1K09 0919 departure from Newbury to London Paddington on 24/09/2025. From Realtime Trains, an independent source of train running info for Great Britain.
www.realtimetrains.co.uk
Train P08445 On Is No Longer Valid
Live Train Times for Train to departing on | Rail Record
live.rail-record.co.uk
In fact there was a 166 service I can't remember where possibly a Cardiff to somewhere. The unit changed 3 times along the route !
Do TOCs actually expect passengers to get up out of their seats and swap units multiple times on long journeys?
== Doublepost prevention - post automatically merged: ==
Didn't want to copy RTT and railchecker with the train graphics so it's a little basic but it does the job I guess
pokemonsuper9
Established Member
Potentially, yes.
I've never had a double set swap on a journey before but I've had single set swaps a couple times.
Bletchleyite
Veteran Member
A common one where this happens is where a non-ETCS fitted unit (or one with a fault in it) is allocated on a service that is supposed to be going onto the Cambrian from Shrewsbury. They are brought into adjacent platforms, people swap over and the unit that has come in from the Cambrian goes back onto it with the other one heading back to Birmingham. Often that does involve swapping a pair.
trainmania100
Established Member
Probably a stupid question but do the blank allocation messages have any significance?
Eg one comes in with no resource groups, but it does have the train ID.
I assumed this was a deallocation (IE remove the units associated with a the train ID), but I've noticed when I do that, RTT continues to show the original allox and doesn't remove it. So I'm guessing it means something else and I only update the allocations if a new set is present
(As an example, P10503 5P37 on Wednesday 1st October got an allocation of 802007 around 1am (although I can't check right now), but it got removed by my code , presume because a new allocations message came through, with blank numbers like I've seen some already).
Eg one comes in with no resource groups, but it does have the train ID.
I assumed this was a deallocation (IE remove the units associated with a the train ID), but I've noticed when I do that, RTT continues to show the original allox and doesn't remove it. So I'm guessing it means something else and I only update the allocations if a new set is present
(As an example, P10503 5P37 on Wednesday 1st October got an allocation of 802007 around 1am (although I can't check right now), but it got removed by my code , presume because a new allocations message came through, with blank numbers like I've seen some already).
A message with no allocations means the allocation has been deleted.
Any new consist message should entirely replace any existing data stored.
I cannot see any consist messages for this service that have come through the feed whatsoever. My message archives for midnight, 1am and 2am (UTC) have no messages with that UID.
Sorry, my bad, I checked the wrong side of 1am. Allocation came through 00:28.
Deleted, then replaced with... the same unit... at 04:53.
Deleted, then replaced with... the same unit... at 04:53.
trainmania100
Established Member
Thanks. For some reason my logging, which simply dumps the messages into hourly batches as they come in, has the deallocation message AFTER the other two.
(Although the message Date Time field does correspond to Tom's post times). Although mine are seemingly in BST going by the time string on the actual message.
Could be my system lagging behind perhaps.
Yes, Gemini is local time for the timestamps in the messages themselves, as I understand it.
Likewise, my archive shows them in the correct order from RDM.
trainmania100
Established Member
Thanks. Comparing my incoming messages to the messages on the RDM "Preview" screen, I'm getting them processed accordingly as they come in, so no clue how I managed to fudge up that schedule.
Could be my process of storing the XML in hourly batches wasn't particularly healthy and caused some sort of delay.
Recon I'll just be grateful it's working at all, Im not really a hardcore coder LOL
Last edited:
While there are certainly some things that the UK does best in Europe (in terms of railway open data), I would argue that Finland is better in most respects, even in terms of the third-party services that use said open data.
Allocations for all passenger and empty-stock trains have been available, including unit numbers for commuter services (so not long-distance or regional), for over 10 years. For almost as long, GPS and speed traces from every train (even freight and departmental), routes set for trains, track section reservations, traffic restrictions (TSRs, closures, etc.), passenger information (special notices displayed at stations), even network maps and speed limit data (for all train types) covering the entire country are also part of open data. Oh, and all train information (including speed traces and track section information) that has ever been available is still available (although commuter train data is only stored for 14 days, because there are so many of them). It's cute that RTT now has up to 3 months' history if you pay; we have 11 years' for free.
We're a small country, so we only have one main open data website (Julia) using the data, but it has the same following as RTT, for example. It also, for over 20 years, has had a public gen-sharing function to fill in the allocations of non-public trains.
Not all of our trains have open data: sometimes the drivers' tablets which provide the GPS and speed data don't work properly (and it doesn't work in tunnels) and military trains are omitted from open data, for obvious reasons.
RTT's now become 5 years, with future expansion to include back to 2012 planned.It's cute that RTT now has up to 3 months' history if you pay; we have 11 years' for free.
Northern158170
Member
Now that I will absolutely pay almost whatever is asked for, provided the unit diagrams are includedRTT's now become 5 years, with future expansion to include back to 2012 planned.
Class15
Established Member
Still “if you pay” though. I’d say the closest we have to the Finland one is Charlwood House.RTT's now become 5 years, with future expansion to include back to 2012 planned.
mathstrains19
Member
I would assume that there will be no units all the way back to 2012, and there will only be a handful included in the long term subscription that aren't in the 5 year one, as RTT only started KYT in 2020 or so with just one operator. There are no historical allocations from before RTT receiving them.
Quite, but it's worth remembering we run a few more trains than Finland does. That additional data does cost money to process and store.
Yeah, I’m trying to support hosting costs by displaying ads (I know, nobody likes ads, but I have to try and cover costs somehow!) and offering optional subscriptions to remove them. The ads alone won’t keep the project alive with the current user base, but I’m hoping the critical mass can be reached within a year. Time will tell!
Good luck to you. I have passed on your app details to a few people I know and hopefully they will start to use it over time. Unfortunately with the nights now about to start drawing in and winter just around the corner it may be a slow start but hopefully from February onwards it will start to pick up. Also with the new timetables and the new units (398s, 701s etc) hopefully coming on stream sooner rather than later, this app will take off in a great way. Especially the full days diagrams that you show.
Discussion relating to Trainmapper can now be found in its own thread here.
trainmania100
Established Member
Looking at set NL26 (91105) on 1st November, I got allocations for 3A07 in the early AM, and then nothing at all until 1D29 late evening, despite RTT showing workings for the full day.
I'm guessing RTT predicted the future workings based on the CIF associations between successive schedules? If not, and a direct feed from Gemini I guess some sort of RDM outage around the time these allocations were sent out?
Charlwoodhouse logs both allocation and deallocation to schedule and has no record at all on the day workings so I double checked what I have on there to make sure my own code wasn't an issue...
(Same data on trainmapper)
I can't see any immediate issue with filling in allocations for the rest of the day based on an earlier schedule (one that has association records) and then filling them in (with fresh data) again when the consist message comes through but seems unnecessary
I'm guessing RTT predicted the future workings based on the CIF associations between successive schedules? If not, and a direct feed from Gemini I guess some sort of RDM outage around the time these allocations were sent out?
Charlwoodhouse logs both allocation and deallocation to schedule and has no record at all on the day workings so I double checked what I have on there to make sure my own code wasn't an issue...
(Same data on trainmapper)
I can't see any immediate issue with filling in allocations for the rest of the day based on an earlier schedule (one that has association records) and then filling them in (with fresh data) again when the consist message comes through but seems unnecessary
Last edited:
I don't believe RTT does anything like this. I also can't see any general issues with the Gemini feed on this date.
Raileasy saw allocations set for NL26 on (Y23523) 3A07, Y23252 (1D29) and Y23661 (5D29).
For the sake of making sure there wasn't an issue interpreting the data, there are no results for P00075/1A39 in the feed that day.
Unsurprisingly Traksy has the same three allocations for the day: https://traksy.uk/new/unit/gb/NL26/20393
We don’t do anything like that because it would be silly in instances of set swaps.
Is that Raileasy page publicly available please? Thank you
Sorry, it's not. It's part of our internal data lookup tool, so has much more data than just that which cannot be shared publicly!
I have longer term plans to put together a similar lookup tool for Gemini data in the future, though.
trainmania100
Established Member
Thanks, doesn't look like I missed any data then. I don't use the diagram ID but if I wanted to fill out the missing schedules alloxes I could match up to the diagram and fill in the blanks I guess
I have a question, maybe it should be in it's own thread entitled 'phantom trains?'.
This morning I was looking at the live maps around Cardiff Central on both Traksy and OpenTrainTimes, and spotted a train on both with headcode 5N11 heading in from the east. I watched it stop briefly at P4 at CDF and then disappear into Canton depot. While I was watching it I was also trying to find out what it was. OTT said it didn't know, Traksy, RTT and live.rail-record all gave me a choice of 8 5N11's, all at least 100 miles away, and there aren't any matching VSTPs.
How can this be? And is there some other place I could have looked for the answer?
This morning I was looking at the live maps around Cardiff Central on both Traksy and OpenTrainTimes, and spotted a train on both with headcode 5N11 heading in from the east. I watched it stop briefly at P4 at CDF and then disappear into Canton depot. While I was watching it I was also trying to find out what it was. OTT said it didn't know, Traksy, RTT and live.rail-record all gave me a choice of 8 5N11's, all at least 100 miles away, and there aren't any matching VSTPs.
How can this be? And is there some other place I could have looked for the answer?
CR165022
Established Member
Sometimes trains aren't run under schedules, in instances such as depot moves & shunts etc.