Headcodes and train IDs: ongoing discussion.

[whole diary]

[grahame]

Under "Significant disruption" on Journey Checker ...

Avantix^▸ (Ticket Issuing System used on board trains) Failure on: 2J15 08:37 Reading to Basingstoke

Surprising thing to report widely as "Significant disruption", or is there something I don't know?

I do like to see train reporting numbers - as a programmer, it's a valuable unique key rather than having to work out that the 07:04 from Westbury and the 07:20 from Melksham are actually the same thing!

[brizzlechris]

Although not as unique as one might hope...

[Brucey]

When issues are entered, staff are given the option of making the report public via JourneyCheck or just private on TyrellCheck.  I assume this message was intended to be private...!

[bobm]

Although not as unique as one might hope...

A few years ago I was lucky enough to be shown around Reading Panel Box and I talked to one of the signalmen about reporting numbers.  At that time the trains from Waterloo were class 1 trains and they had two trains in the space of 20 minutes at the end of the morning rush hour which were both 1C13.  One was from terminating from Waterloo, the other was from Paddington to Weston-super-Mare.

[inspector_blakey]

You'll note though that, as far as is humanly possible, trains with the same reporting number are either separated widely geographically or temporally.

For TRUST^▸ (Train Running System TOPS) reporting purposes, each train is actually allocated a ten-character identifier which includes the reporting number with various other bits and pieces of extra information including codes relating to the originating area, originating time, date, type of train... This will identify a train uniquely within a time period of one month.

[EBrown]

I do like to see train reporting numbers - as a programmer, it's a valuable unique key rather than having to work out that the 07:04 from Westbury and the 07:20 from Melksham are actually the same thing!

Train reporting numbers (aka headcodes [slightly wrongly]) aren't completely unique. are only used for timetabling and aren't (really) used on the operating railway in their current form. This is why there can be an overlap between different operators. There shouldn't be an overlap between RSIDs.

As inspector_blakey says, trains are assigned a ten-character number, which is unique over around a 3 month period^{(citation needed, memory iffy)}, so for example the PNZ-PAD^» (Paddington (London) - next trains) sleeper tonight is:
Train reporting number: 1A40
Headcode: 1276
Train ID: 851A40M724

A London Victoria to Brighton service tomorrow has the same TRN... but a different Train ID.

[inspector_blakey]

As inspector_blakey says, trains are assigned a ten-character number, which is unique over around a 3 month period^{(citation needed, memory iffy)}, so for example the PNZ-PAD^» (Paddington (London) - next trains) sleeper tonight is:
Train reporting number: 1A40
Headcode: 1276
Train ID: 851A40M724

Close, but not quite correct. Firstly, "reporting number" and "headcode" are used interchangeably on the operating railway to refer to the same thing, which is 1A40. In the example above, 1276 is most definitely not the headcode. Actually I have no idea what it is, unless it's derived from the number you see printed on the bottom of a seat reservation tag, in which case it's purely commercial (ticketing, seat reservations etc) and has absolutely no currency in terms of train operations.

For the sake of completeness, the ten-character ID is made up as follows:

Characters 1 and 2
Every railway location in the country used as a timing point (stations, junctions, sidings...) is allocated a five-digit number called, for reasons no-one has been able to explain to me, a Stanox^▸ (Station Number (5 digit code)). As a very coarse rule of thumb, Stanox numbers count up as you go down the country geopgraphically, NW-SE, with the lowest at Thurso and the highest at Dollands Moor. Characters 1 and 2 of the ten-character ID are the first two digits of the Stanox of the location at which the train originates.

Characters 3 - 6
Headcode/reporting number.

Character 7
A character denoting the type of train: M is used for passenger trains, with other letters/numbers used for different trains depending on whether they're passenger or freight, working timetable or short-term planned etc/

Character 8
A letter or number that denotes the scheduled starting time of the train. Every letter and number is used over the course of the day, although the way in which they are allocated isn't 100% logical as some refer to 60 min time periods and others to 30 min intervals, starts off counting 0 - 6, then A - Z, then 7 - 9. In the case of 1A40 the 7 refers to an originating time of 2100 - 2159.

Characters 9 and 10
The date on which the train originates (with a leading zero if it's a single digit), so the example above refers to a train originating on 24 Feb.

From this it will therefore be apparent that, assuming the timetable stays the same, a given ten-digit ID will uniquely identify a train within a period of one month, not three. This is overkill however, as the principal use of the ten-character codes is for identifying trains in TRUST^▸ (Train Running System TOPS), which only holds data for 9 days before it is archived.

Train reporting numbers (aka headcodes [slightly wrongly]) are only used for timetabling and aren't (really) used on the operating railway in their current form. This is why there can be an overlap between different operators.

I'm not sure what you mean by this but I would argue that that statement is very wrong: I can tell you've never set foot in a signalling centre. The four character headcodes are the basic currency in signal boxes, signalling centres and route controls: you're unlikely to hear ops staff (signallers, controllers, train crews) refer to a train by anything other than its headcode. Although there will in some cases be multiple trains running on the network with a given headcode on one day, great care is taken to separate these geographically and temporally to avoid any possible confusion.

[IndustryInsider]

Characters 1 and 2
Every railway location in the country used as a timing point (stations, junctions, sidings...) is allocated a five-digit number called, for reasons no-one has been able to explain to me, a Stanox^▸ (Station Number (5 digit code)). As a very coarse rule of thumb, Stanox numbers count up as you go down the country geopgraphically, NW-SE, with the lowest at Thurso and the highest at Dollands Moor. Characters 1 and 2 of the ten-character ID are the first two digits of the Stanox of the location at which the train originates.

I used to think the first part of 'STANOX' stood for STAtion Number, but you also have the 9-characted STANME codes which are alpha based and not numeric, so the 'N' must stand for something else!

[inspector_blakey]

I did ask our TOPS^▸ (Total Operations Processing System)/TRUST^▸ (Train Running System TOPS) guru who knows everything about the system, has years of experience and runs training sessions at head office, but even he didn't know how the Stanox^▸ (Station Number (5 digit code))/Stanme nomenclature came about!

[JayMac]

STANOX.

Perhaps STA for Station, NO for Number, and the X to denote the plural. Maybe some reason why it was decided STANOS was not acceptable. Could simply be that STANOX sounds more technical.

[SandTEngineer]

From the Railway Group Standards website.
Stanox^▸ (Station Number (5 digit code)) = TOPS^▸ (Total Operations Processing System) Location Code. A five digit number representing a TOPS location.

A list of STANOXES can be found here: http://deaves47.zxq.net/CRS/CRS0.htm and somebody has already asked NR^» (Network Rail - home page): http://www.whatdotheyknow.com/request/stanme_codes

[inspector_blakey]

NR^» (Network Rail - home page) make an updated list of Stanoces*/Stanmes available at https://datafeeds.networkrail.co.uk/.

*Yes, that is the plural that seems to be used generally

[EBrown]

Close, but not quite correct.

Yes and no...

Firstly, "reporting number" and "headcode" are used interchangeably on the operating railway to refer to the same thing, which is 1A40

That may well be the case in the operating railway, but is not the case within timetabling.

In the example above, 1276 is most definitely not the headcode.

I'm afraid in this instance you are incorrect. The headcode forms part of the RSID, made up from the ATOC^» (Association of Train Operating Companies See - here) operator code, so for the above example, it is GW1276.

For the sake of completeness, the ten-character ID is made up as follows:

Characters 1 and 2
Every railway location in the country used as a timing point (stations, junctions, sidings...) is allocated a five-digit number called, for reasons no-one has been able to explain to me, a Stanox^▸ (Station Number (5 digit code)). As a very coarse rule of thumb, Stanox numbers count up as you go down the country geopgraphically, NW-SE, with the lowest at Thurso and the highest at Dollands Moor. Characters 1 and 2 of the ten-character ID are the first two digits of the Stanox of the location at which the train originates.

...

Characters 9 and 10
The date on which the train originates (with a leading zero if it's a single digit), so the example above refers to a train originating on 24 Feb.

Thanks for the info, certainly never knew that before.

Train reporting numbers (aka headcodes [slightly wrongly]) are only used for timetabling and aren't (really) used on the operating railway in their current form. This is why there can be an overlap between different operators.

There I am certainly wrong, I got myself in a pickle it seems. Original post modified suitably.

[inspector_blakey]

Close, but not quite correct.

Yes and no...

Firstly, "reporting number" and "headcode" are used interchangeably on the operating railway to refer to the same thing, which is 1A40

That may well be the case in the operating railway, but is not the case within timetabling.

In the example above, 1276 is most definitely not the headcode.

I'm afraid in this instance you are incorrect. The headcode forms part of the RSID, made up from the ATOC^» (Association of Train Operating Companies See - here) operator code, so for the above example, it is GW1276.

Well, if what you've just said is correct then I think this may be an example of different functions within the rail industry using different terminology. An operator will never, never refer to the RSID as a headcode for the very simple reason that is isn't one. A headcode is exactly that: a code that was carried on the front end (head) of a train, originally to indicate to signallers what type of train it was and/or where it was going. Go far enough back and the headcode was indicated by lamps or discs in a particular position on the front of the locomotive. Later on these were replaced with the four-character headcode system which was displayed on the front of trains until the 1970s. At this point the practice was stopped, in large measure because signallers were much more concentrated in "panel" boxes than they used to be and wouldn't see most (or indeed any) of the trains under their control: because they weren't seeing the trains there wasn't much point in showing the headcode on the front, and the headcode blind equipment was a bit of a pain to maintain. Headcodes (1A40 etc) are now displayed instead on a train describer in every signal box that indicates to the signaller the location of the trains under their control at that moment.

I still stand by my assertion that the four digits of the RSID are most definitely not the headcode, at least by a unanimous opinion amongst the staff who actually run the trains and therefore ought to know, and that whoever is telling you that is using the term incorrectly; it's possible this is common practice with timetabling but in fact having spent a fair bit of time with train planners I've never heard them use the RSID in any context at all, and they have certainly never referred to the four digits therein as a headcode. Even if timetabling staff are doing this, it's absolutely not technically correct to do so.

[EBrown]

I still stand by my assertion that the four digits of the RSID are most definitely not the headcode, at least by a unanimous opinion amongst the staff who actually run the trains and therefore ought to know, and that whoever is telling you that is using the term incorrectly; it's possible this is common practice with timetabling but in fact having spent a fair bit of time with train planners I've never heard them use the RSID in any context at all, and they have certainly never referred to the four digits therein as a headcode. Even if timetabling staff are doing this, it's absolutely not technically correct to do so.

I am telling you definitively that is what a headcode is within timetabling. I'm afraid you are wrong in this instance. If you wish to find out more I'd suggest you talk to ATOC^» (Association of Train Operating Companies See - here)/Network Rail, perhaps Tom Cairns, Peter Higgs or, if you feel brave Rockshore.

An example of the RSID being used is retail with Virgin Trains. (Which probably means retail with thetrainline too - though I cannot back up that assertion).