What is the basis for updating of information on departure screens at 
unmanned stations?

Waited at Matlock station (end of single branch, unmanned, but with 
departure screen) this morning for 08.59 departure (due in at 08.54).

Screen confidently announced train as On Time even as 08.59 came and 
went without arrival of train.

So, the screen is clearly not benefiting from detailed realtime information.

Train arrived about 10.03. As it arrived (ie about five minutes before 
departure), the screen changed to show that the first departure was the 
10.39

So, the screen is clearly not updating the "next train" automatically on 
the basis of "all-assumed-ontime unless-otherwise-informed", or it would 
have shown next departure the moment that scheduled departure time had 
come and gone.

So what IS the basis for updating of information on screens at unmanned 
stations? Can passengers rely on an "on time" indication to mean that 
the train will run close enough to time to catch them mainline 
connections? (From Matlock, if you want to drive to Derby to catch a 
connection you need to set off before the train would have left).

With this morning's "on time" train running nine minutes late (which 
would have made some connections at Derby a bit dodgy) it seems that the 
information is not reliable enough to be useful, and the cautious 
passenger would be better advised to ignore the screen and call National 
Rail Enquiries on their mobile every time.

John Geddes
Derbyshire

John Geddes wrote:

> What is the basis for updating of information on departure screens at 
> unmanned stations?
> 
> Waited at Matlock station (end of single branch, unmanned, but with 
> departure screen) this morning for 08.59 departure (due in at 08.54).
> 
> Screen confidently announced train as On Time even as 08.59 came and 
> went without arrival of train.
> 
> So, the screen is clearly not benefiting from detailed realtime 
> information.
> 
> Train arrived about 10.03. As it arrived (ie about five minutes before 
> departure), the screen changed to show that the first departure was the 
> 10.39
> 
> So, the screen is clearly not updating the "next train" automatically on 
> the basis of "all-assumed-ontime unless-otherwise-informed", or it would 
> have shown next departure the moment that scheduled departure time had 
> come and gone.
> 
> So what IS the basis for updating of information on screens at unmanned 
> stations? Can passengers rely on an "on time" indication to mean that 
> the train will run close enough to time to catch them mainline 
> connections? (From Matlock, if you want to drive to Derby to catch a 
> connection you need to set off before the train would have left).
> 
> With this morning's "on time" train running nine minutes late (which 
> would have made some connections at Derby a bit dodgy) it seems that the 
> information is not reliable enough to be useful, and the cautious 
> passenger would be better advised to ignore the screen and call National 
> Rail Enquiries on their mobile every time.
> 
> John Geddes
> Derbyshire

it might depend on whether intermediate stations, or routes between 
stations update the running time on the train. If the train is delayed 
but there is no monitoring point the live system won't know about it. Of 
course manned stations can always find out and announce delays. You're 
lucky you have a information screen at all.

On Sat, 30 Jul 2005 11:10:16 +0100, John Geddes
 wrote:


>What is the basis for updating of information on departure screens at 
>unmanned stations?


Usually the signalling/train reporting system, which is the same
information NRES (online and phone) will have.  On single-track branch
lines with no proper signalling system, there will usually be no
reports so the information will be inaccurate.

You can identify such situations by looking for "no report" on the
online departure boards.  This, unfortunately, is usually shown as "on
time" on station boards.

Neil

-- 
Neil Williams in Milton Keynes, UK
When replying please use neil at the above domain
'wensleydale' is a spam trap and is not read.

On Sat, 30 Jul 2005 10:10:16 UTC, John Geddes 
 wrote:

: and the cautious 
: passenger would be better advised to ignore the screen and call National 
: Rail Enquiries on their mobile every time.

I use the text service: "Station name" to 484950 and find it generally
very reliable. It can be a bit over pessimistic sometime - trains seem
quite able to make up a bit of time between the previous reporting 
point and Dumfries, for example, so a "+9mins" train can be "+4mins" 
when it appears in the flesh.

Ian


--

I suspect the following has happened.

The departure screen is probably plugged into the train reporting system,
which is where the live departure boards on National Rail get there info
from. The last reporting point for the is probably Ambergate Junction, where
the Matlock line leaves the Midland Mainline.

The train can pass this point on time but be delayed somewhere else on the
branch, and since there is no other reporting point on the branch until
Matlock, there is no way of the system knowing if the train is late until
the time of arrival has come and gone and wont know when its due until it
shows up.

-- 
http://europeanrailways.fotopic.net/
Any views or opinions expressed and presented are not those of the author
and do not represent those of his employers, they belong to the voices in
his head.

In message , at 11:10:16 
on Sat, 30 Jul 2005, John Geddes  
remarked:

>So what IS the basis for updating of information on screens at unmanned 
>stations? Can passengers rely on an "on time" indication to mean that 
>the train will run close enough to time to catch them mainline 
>connections?


It can be pretty dodgy. After years of observation at Royston Station it 
became apparent that any train flagged up as [exactly] 3 minutes late 
was in effect "lost", and the system in fact had no idea when it would 
arrive.

Similarly, when the online departure boards were first available (mid 
1999) it was apparent that a great number of GNER departures from 
Peterborough would get a prediction which became one minute later every 
minute that went by.
-- 
Roland Perry

BlueGhost wrote:

> it might depend on whether intermediate stations, or routes between 
> stations update the running time on the train. If the train is delayed 
> but there is no monitoring point the live system won't know about it. Of 
> course manned stations can always find out and announce delays. You're 
> lucky you have a information screen at all.


Indeed. These days, there really is no excuse for all stations (apart 
from maybe the very smallest halts) not to have information screens, so 
that at least the waiting passengers can be made aware if a train is 
delayed or cancelled.

Dave

On Sat, 30 Jul 2005 19:42:27 +0100, Dave Roberts
 wrote:


>Indeed. These days, there really is no excuse for all stations (apart 
>from maybe the very smallest halts) not to have information screens, so 
>that at least the waiting passengers can be made aware if a train is 
>delayed or cancelled.


I'd argue that, with the widespread use of mobile phones and the fact
that the LDBs can be called up in a number of ways using one, it is
actually *less* important than it used to be, though obviously the
technology is cheaper than it was hence it is becoming more
widespread.

What annoys me[1] is the reverse-billed charge for the use of NRES
departure boards by text.  It should be a free service where it acts
as a replacement for a proper departure board or long-line PA.

[1] Actually, I believe that reverse-billed texts should be banned
full-stop, certainly those charged at greater than whatever your
operator's sending charge is, but that's a bit OT for here.

Neil

-- 
Neil Williams in Milton Keynes, UK
When replying please use neil at the above domain
'wensleydale' is a spam trap and is not read.

On Sat, 30 Jul 2005 19:42:27 +0100, Dave Roberts wrote in
, seen in uk.railway:

> BlueGhost wrote:
> > it might depend on whether intermediate stations, or routes between 
> > stations update the running time on the train. If the train is delayed 
> > but there is no monitoring point the live system won't know about it. Of 
> > course manned stations can always find out and announce delays. You're 
> > lucky you have a information screen at all.
> 
> Indeed. These days, there really is no excuse for all stations (apart 
> from maybe the very smallest halts) not to have information screens, so 
> that at least the waiting passengers can be made aware if a train is 
> delayed or cancelled.


I wish we lived in a country where such things could be placed and
left without them being vandalised almost immediately. 

I can think of three stations near me where screens were placed and
repeatedly destroyed by the local darlings, to the extent that the
local authority (who'd paid for them) refused to spend any more money
on them.

And even if they are there, as BlueGhost says, you then have the
problem of the quality of data supplied to them. In many cases all
they actually do is display the scheduled data (because that is the
only data available), so any delays won't show, and thus the
information isn't necessarily all that valuable.

If there's a decent system which can be fitted, I'd love to see it
used everywhere, because I think it's incumbent upon the railway
industry to keep its customers fully informed; unfortunately there are
too many problems with the systems in use today for them to fulfil
that rle adequately.

-- 
Ross, Lincoln, UK

We're *not* afraid
http://www.werenotafraid.com

>If there's a decent system which can be fitted, I'd love to see it
>used everywhere, because I think it's incumbent upon the railway
>industry to keep its customers fully informed; unfortunately there are
>too many problems with the systems in use today for them to fulfil
>that rôle adequately.



I assume that in addition to the basic system, there are also options
which human beings can choose to use or not.  This seems to be the case
on SET.  Until recently, it was common for the expected time of a train
that had broken down two minutes away to update continually to being
expected in two minutes from the current time for half an hour or more.

More recently, they seem to have been using the "big problem; don't
know" option (which comes up as "delayed"), but that may depend on who
is on duty.  At least then you know that there is a problem.

On Silverlink, as I've ranted before, the displays never seem to show
anything other than the timetabled times, regardless of what is
running, eg showing the next train coming in eighteen minutes, while
the two-minute-delayed previous train is about to pull in.

But I bet it's the same system that SET uses in terms of the hardware
and software (not the people).

On 30 Jul 2005 15:22:52 -0700, "MIG" 
wrote:


>More recently, they seem to have been using the "big problem; don't
>know" option (which comes up as "delayed"), but that may depend on who
>is on duty.  At least then you know that there is a problem.


It wouldn't be hard to code around that problem - Merseyrail's system
does.  All you need is logic stating that if the delay time increases
twice in a given time period (say, 3 minutes), change the display to
"Delayed" rather than a time until such time as the delay time does
*not* increase in the same given time period, at which point it is
reasonable to display the time instead.

On a slightly related matter, does anyone find it annoying that many
of the newer PISs always display an expected time instead of "On
time"?  That way you have to look at the time and decide if it is or
not.

Neil

-- 
Neil Williams in Milton Keynes, UK
When replying please use neil at the above domain
'wensleydale' is a spam trap and is not read.

>On a slightly related matter, does anyone find it annoying that many
>of the newer PISs always display an expected time instead of "On
>time"?  That way you have to look at the time and decide if it is or
>not.



SET has this in some places; I think mainly where the systems have been
introduced more recently and are perhaps a later version, such as
Waterloo East.

Sometimes the expected time is now earlier than the due time.  This may
be because a while ago, all published journey times from Charing Cross
to Waterloo East were increased from two to three minutes.  The system
presumably notes the train at Charing Cross about to depart and knows
that it's really two minutes away.

On Sat, 30 Jul 2005 22:33:18 GMT, wensleydale@pacersplace.org.uk (Neil
Williams) wrote:


>On a slightly related matter, does anyone find it annoying that many
>of the newer PISs always display an expected time instead of "On
>time"?  That way you have to look at the time and decide if it is or
>not.


Yes, very.

Almost as bad as the Silverlink ones which take up a large part of the
LED display displaying the platform number (which is of course also
painted on fixed signs, and never changes), and then don't have enough
space to display both the scheduled time and the expected time, so
alternate between the two every 5 seconds or so (so you have to stare
at it for a while to get both pieces of information).

"Ross"  wrote in message
news:7mqne1dfkrpbj5n779utbnuujcg4kj22bv@4ax.com...

> If there's a decent system which can be fitted, I'd love to see it
> used everywhere, because I think it's incumbent upon the railway
> industry to keep its customers fully informed; unfortunately there are
> too many problems with the systems in use today for them to fulfil
> that rle adequately.


Well, Arriva Trains Wales seem to have an "information point" at all their
stations, even the smallest. It's pretty vandal-proof, just a box with a
button and a speaker - I've not seen a broken one yet. When you press the
button, it reads out the time of the next train and how late it is (or just
the train details if it is on time). I've also seen these at other
stations - Central seem to have quite a few.

Mike

"Mike Humphrey"  wrote:


>Well, Arriva Trains Wales seem to have an "information point" at all their
>stations, even the smallest. It's pretty vandal-proof, just a box with a
>button and a speaker - I've not seen a broken one yet. When you press the
>button, it reads out the time of the next train and how late it is (or just
>the train details if it is on time). I've also seen these at other
>stations - Central seem to have quite a few.


Indeed - it seems to be a very good system. The "buttons" don't seem
to have moving parts in, which presumably makes the system more robust.

However, it would be nice if you didn't have to listen to a list of
every single station a train stops at, followed by details of catering
arrangements, *before* you are told if the train is running late, or
even that it's cancelled.

The system did fail me once, when waiting for a Heart of Wales line
train at Craven Arms - after being described as late for some time,
the system then 'forgot' about the train completely and started
describing the following train.

Rail enquiries didn't do much better; they told me that it must have
failed part way and been cancelled and were just trying to find me an
alternate route when the train finally pulled in. Not very impressive.

Adam

A Woodcraft wrote:

> "Mike Humphrey"  wrote:
>
> >Well, Arriva Trains Wales seem to have an "information point" at all their
> >stations, even the smallest. It's pretty vandal-proof, just a box with a
> >button and a speaker - I've not seen a broken one yet. When you press the
> >button, it reads out the time of the next train and how late it is (or just
> >the train details if it is on time). I've also seen these at other
> >stations - Central seem to have quite a few.
>
> Indeed - it seems to be a very good system. The "buttons" don't seem
> to have moving parts in, which presumably makes the system more robust.
>
> The system did fail me once, when waiting for a Heart of Wales line
> train at Craven Arms - after being described as late for some time,
> the system then 'forgot' about the train completely and started
> describing the following train.


I've had trouble with this before, Central's are a telephone in a
sealed unit and dial out when you press it, however I used it once (and
I can't remember where) and it told me I was at a completely different
station to where I actually was and gave train times for there instead!

Chris.

"CJ Bennett"  wrote:


>A Woodcraft wrote:
>> "Mike Humphrey"  wrote:
>>
>> >Well, Arriva Trains Wales seem to have an "information point" at all their
>> >stations, even the smallest. It's pretty vandal-proof, just a box with a
>> >button and a speaker - I've not seen a broken one yet. When you press the

....

>> The system did fail me once, when waiting for a Heart of Wales line
>> train at Craven Arms - after being described as late for some time,
>> the system then 'forgot' about the train completely and started
>> describing the following train.
>
>I've had trouble with this before, Central's are a telephone in a
>sealed unit and dial out when you press it, however I used it once (and
>I can't remember where) and it told me I was at a completely different
>station to where I actually was and gave train times for there instead!


I'm pretty sure the (ex) Wales and Borders units aren't telephones;
there is no pause when you press the button. And apart from this one
occasion, I've used them many times without any trouble, and as I said
above, the one time they did fail me, phoning Rail Enquiries was
equally useless.

Adam

On Sun, 31 Jul 2005 16:56:34 +0100, Mike Humphrey wrote in
<42ecf59b$0$3478$ed2619ec@ptn-nntp-reader03.plus.net>, seen in
uk.railway:
[...]

> Well, Arriva Trains Wales seem to have an "information point" at all their
> stations, even the smallest. It's pretty vandal-proof, just a box with a
> button and a speaker - I've not seen a broken one yet. 


We used to have one of those at Ruskington.

After the fourth time the box was crowbarred [1] off the wall and the
cables torn to shreds, they gave up repairing it.


[1] I assume crowbarred, as I can't see any other (easy) way of
getting them off.
-- 
Ross, Lincoln, UK

We're *not* afraid
http://www.werenotafraid.com

On 30 Jul 2005 15:22:52 -0700, MIG wrote in
, seen in
uk.railway:


> >If there's a decent system which can be fitted, I'd love to see it
> >used everywhere, because I think it's incumbent upon the railway
> >industry to keep its customers fully informed; unfortunately there are
> >too many problems with the systems in use today for them to fulfil
> >that rle adequately.
> 
> I assume that in addition to the basic system, there are also options
> which human beings can choose to use or not.  


I assume so, but the basic system itself relies on data supplies which
are themselves problematic.

To take one example, trains from the Peterborough direction towards
Lincoln get reported on departure from Peterborough, at Werrington
Junction and then (manually) on departure from Spalding. After that
there are no operative reporting points, so train effectively vanish
(for information purposes) between Spalding and Lincoln, which is the
best part of an hour's journey - and a lot can happen in an hour.


[...]

> More recently, they seem to have been using the "big problem; don't
> know" option (which comes up as "delayed"), 


999 minutes delay shows up as "delayed". Took us ages to discover how
to get that "unknown delay" to work, although that may well have been
because our amazing information clerks were quite proud of their
failure to pay any attention during the training. <sigh>


[...]

> But I bet it's the same system that SET uses in terms of the hardware
> and software (not the people).


It's very likely the same system, but the way it is operated can be
different. The system can be used as a stand-alone (which relies on
the staff to update it manually) or it can be tied into TRUST and take
data from that - which is when it tends to go for the "keep adding two
minutes" displays which annoy everyone. 

In either mode it relies on data being available, although that
shouldn't be a problem on the WCML for Silverlink, although I don't
know about the other routes of theirs.

-- 
Ross, Lincoln, UK

We're *not* afraid
http://www.werenotafraid.com

"Mike Humphrey"  wrote in message
news:42ecf59b$0$3478$ed2619ec@ptn-nntp-reader03.plus.net...

> "Ross"  wrote in message
> news:7mqne1dfkrpbj5n779utbnuujcg4kj22bv@4ax.com...
> > If there's a decent system which can be fitted, I'd love to see it
> > used everywhere, because I think it's incumbent upon the railway
> > industry to keep its customers fully informed; unfortunately there are
> > too many problems with the systems in use today for them to fulfil
> > that rle adequately.
>
> Well, Arriva Trains Wales seem to have an "information point" at all their
> stations, even the smallest. It's pretty vandal-proof, just a box with a
> button and a speaker - I've not seen a broken one yet. When you press the
> button, it reads out the time of the next train and how late it is (or
just
> the train details if it is on time). I've also seen these at other
> stations - Central seem to have quite a few.



The Chiltern ones are even better. Rather than giving you recorded or
computer-generated information they put you through to Chiltern Control at
Banbury so you can ask a real person where your train is.

Roger

http://rpm-railpics.fotopic.net/
http://therailwaystationgallery.fotopic.net/
http://therailticketgallery.fotopic.net/

All opinions expressed are my own and do not necessarily reflect the views
or policies of my employer.

--- Ross said...


> It's very likely the same system, but the way it is operated can be
> different. The system can be used as a stand-alone (which relies on
> the staff to update it manually) or it can be tied into TRUST and take
> data from that - which is when it tends to go for the "keep adding two
> minutes" displays which annoy everyone.


What's even more annoying is when the display says a "Terminates Here"
arrival is delayed, but still thinks the subsequent departure as a new
service is on time, even though that means the train would be departing
_before_ it arrives!

 And then, when the departure time comes and goes and the display
finally does notice that the train hasn't departed, it goes into that
"keep adding two minutes" mode, although it already knows the
"Terminates Here" train won't arrive for another 10 minutes at least!

Is it _really_ that hard to link a "Terminates Here" to its later
departure, so that the display knows that they stay in sync?

wrote

>
> What's even more annoying is when the display says a "Terminates Here"
> arrival is delayed, but still thinks the subsequent departure as a new
> service is on time, even though that means the train would be departing
> _before_ it arrives!
>
>  And then, when the departure time comes and goes and the display
> finally does notice that the train hasn't departed, it goes into that
> "keep adding two minutes" mode, although it already knows the
> "Terminates Here" train won't arrive for another 10 minutes at least!
>
> Is it _really_ that hard to link a "Terminates Here" to its later
> departure, so that the display knows that they stay in sync?
>

There was a time when FGW and Thames each used there own versions, but
Thames showed all trains at Paddington. Thames assumed that a late inward
FGW train would turn round in 1 minute; FGW assumed that it would use its
full terminus allowance. Result - Thames was overoptimistic; FGW was over
pessimistic.

Peter

On 31 Jul 2005 14:49:47 -0700, paul.ingerson@gmail.com wrote in
, seen in
uk.railway:
[...]

> What's even more annoying is when the display says a "Terminates Here"
> arrival is delayed, but still thinks the subsequent departure as a new
> service is on time, even though that means the train would be departing
> _before_ it arrives!
> 
>  And then, when the departure time comes and goes and the display
> finally does notice that the train hasn't departed, it goes into that
> "keep adding two minutes" mode, although it already knows the
> "Terminates Here" train won't arrive for another 10 minutes at least!
> 
> Is it _really_ that hard to link a "Terminates Here" to its later
> departure, so that the display knows that they stay in sync?


The software used is, believe it or not, over 20 years old and
(needless to say) proprietary, so I suspect the answer is "Yes, it
really _is_ that hard". :(

And, also unbelievably, not everywhere has that software yet. Lincoln
only got it last year, and I suspect (based on the old monitors used
and the display style) that Sheffield still doesn't have it.

-- 
Ross, Lincoln, UK

We're *not* afraid
http://www.werenotafraid.com

On Sun, 31 Jul 2005 23:13:53 +0100, Ross 
wrote:


>And, also unbelievably, not everywhere has that software yet. Lincoln
>only got it last year, and I suspect (based on the old monitors used
>and the display style) that Sheffield still doesn't have it.


The similarity to a Mode 7 Teletext display would suggest that
Sheffield is still on the old BBC Micro-based setup.

Anyone ever seen what it runs on?

Neil

-- 
Neil Williams in Milton Keynes, UK
When replying please use neil at the above domain
'wensleydale' is a spam trap and is not read.

--- Ross said...


> On 31 Jul 2005, paul.ingerson@gmail.com wrote
> >
> > Is it _really_ that hard to link a "Terminates Here" to its later
> > departure, so that the display knows that they stay in sync?
>
> The software used is, believe it or not, over 20 years old and
> (needless to say) proprietary, so I suspect the answer is "Yes, it
> really _is_ that hard". :(


Thanks Ross.  Twenty years old?  Wow!

I suppose there's not a wide demand for railway information display
software, so it's not like you can just walk into PC World and buy the
latest version.  OTOH you'd think they would've fixed these bugs in all
that time!


> And, also unbelievably, not everywhere has that software yet. Lincoln
> only got it last year, and I suspect (based on the old monitors used
> and the display style) that Sheffield still doesn't have it.


Well, if Sheffield's system works better than the standard software,
they're probably in no hurry to change it!

"paul.ingerson@gmail.com"  wrote:


>> The software used is, believe it or not, over 20 years old and
>> (needless to say) proprietary, so I suspect the answer is "Yes, it
>> really _is_ that hard". :(
>
>Thanks Ross.  Twenty years old?  Wow!
>
>I suppose there's not a wide demand for railway information display
>software, so it's not like you can just walk into PC World and buy the
>latest version.  OTOH you'd think they would've fixed these bugs in all
>that time!
>
>> And, also unbelievably, not everywhere has that software yet. Lincoln
>> only got it last year, and I suspect (based on the old monitors used
>> and the display style) that Sheffield still doesn't have it.
>
>Well, if Sheffield's system works better than the standard software,
>they're probably in no hurry to change it!


I liked the system they had at Sheffield in the 70's - a camera
pointing at a table where the departures were laid out on bits of
card.

Occasionally you would see a giant hand appear and re-arrange them.

I don't know how unusual this was, but I never saw it anywhere else.

Adam

A Woodcraft wrote:


> I liked the system they had at Sheffield in the 70's - a camera
> pointing at a table where the departures were laid out on bits of
> card.


A bit like the delightfully low-tech 'live' Liverpool St departure
boards, using a webcam:

http://www.vicinitee.com/docs/travel/webcams/webcams_01.cfa

which are particularly pleasing since LS has one of the last big Solari
boards.

> I wish we lived in a country where such things could be placed and
> left without them being vandalised almost immediately.


There is such a place. Where the line is electrified. That might deter
the little darlings!

On 1 Aug 2005 06:22:32 -0700, 1501 wrote in
, seen in
uk.railway:


> > I wish we lived in a country where such things could be placed and
> > left without them being vandalised almost immediately.
> 
> There is such a place. Where the line is electrified. That might deter
> the little darlings!


Doesn't seem to work at Lea Hall.
-- 
Ross, Lincoln, UK

We're *not* afraid
http://www.werenotafraid.com

John Geddes wrote:


> Waited at Matlock station (end of single branch, unmanned, but with
> departure screen) this morning for 08.59 departure (due in at 08.54).
> 
> Screen confidently announced train as On Time even as 08.59 came and
> went without arrival of train.


That happens at staffed stations too.

>John Geddes wrote:
>> Waited at Matlock station (end of single branch, unmanned, but with
>> departure screen) this morning for 08.59 departure (due in at 08.54).
>
>> Screen confidently announced train as On Time even as 08.59 came and
>> went without arrival of train.
>
>That happens at staffed stations too.




When I ask staff to explain the nonsense on the screens, they generally
have no better information than the customers do.  But on pestering
they may make a phone call and then get enough information to make an
announcement.

On Tuesday 02 August 2005 20:54, MIG wrote:


>>John Geddes wrote:
>>> Waited at Matlock station (end of single branch, unmanned, but with
>>> departure screen) this morning for 08.59 departure (due in at
>>> 08.54).
>>
>>> Screen confidently announced train as On Time even as 08.59 came and
>>> went without arrival of train.
>>
>>That happens at staffed stations too.
> 
> 
> 
> When I ask staff to explain the nonsense on the screens, they
> generally
> have no better information than the customers do.  But on pestering
> they may make a phone call and then get enough information to make an
> announcement.


To be fair, I find that most of the time the screens in Manchester are
right.

John Geddes wrote:

> What is the basis for updating of information on departure screens at 
> unmanned stations?
> 
> Waited at Matlock station (end of single branch, unmanned, but with 
> departure screen) this morning for 08.59 departure (due in at 08.54).
> 
> Screen confidently announced train as On Time even as 08.59 came and 
> went without arrival of train.
> 
> So, the screen is clearly not benefiting from detailed realtime 
> information.


They didn't fire up the Matlock line loudspeaker system then? There's 
one at Duffield which gets the cobwebs shaken off (mostly by the 
feedback) when the people at Derby 'box (I think) remember to use it, 
and the announcement seems to imply it's feeding every station.

Sam

On Tue, 02 Aug 2005 21:37:27 +0100, Adam Funk <a24061@yahoo.munged>
wrote:


>To be fair, I find that most of the time the screens in Manchester are
>right.


I wouldn't rely on the ones at Oxford Road of an evening when the
control staff have gone home.  For them to be correct is an exception,
which is pretty dire when it could end up with people missing the last
train.

Neil

-- 
Neil Williams in Milton Keynes, UK
When replying please use neil at the above domain
'wensleydale' is a spam trap and is not read.

MIG wrote:


> Sometimes the expected time is now earlier than the due time.


LDB often shows this on the final station of a route.
York, on time. Thirsk, on time. Northallerton, on time. Yarm, on time.
Thornaby, on time. Middlesbrough - oh, three minutes early!

-- 
                          Stevie D
    \\\\\       /////     Bringing dating agencies to the
   \\\\\\\__X__///////    common hedgehog since 2001 - "HedgeHugs"
___\\\\\\\'/ \'///////_____________________________________________

John Geddes wrote:


> Waited at Matlock station (end of single branch, unmanned, but with 
> departure screen) this morning for 08.59 departure (due in at 08.54).
.... 
> Train arrived about 10.03. As it arrived (ie about five minutes before 
> departure), the screen changed to show that the first departure was 
> the 10.39


The problem is most likely to be that the computer doesn't link the
inward working with the outward working. So although it knows that the
0854 arrival is still out there, running late, it isn't able to
transfer that delay to the 0859 departure. This is a particular
problem for stations at the end of the line.

As others have said, in this situation LDB shows the fractionally more
accurate, if no more helpful, "No report".


Having said that, being a terminal unstaffed station is not the only
way to get inaccurate information. Selby is both staffed and a through
station, with westbound trains starting from Hull, 30 miles away. Yet
the screens rarely show accurate information when trains from Hull are
delayed.

Surely writing the software can't be _that_ big a job?


> So, the screen is clearly not updating the "next train" automatically on 
> the basis of "all-assumed-ontime unless-otherwise-informed", or it would 
> have shown next departure the moment that scheduled departure time had 
> come and gone.
> 
> So what IS the basis for updating of information on screens at unmanned 
> stations? Can passengers rely on an "on time" indication to mean that 
> the train will run close enough to time to catch them mainline 
> connections? (From Matlock, if you want to drive to Derby to catch a 
> connection you need to set off before the train would have left).
> 
> With this morning's "on time" train running nine minutes late (which 
> would have made some connections at Derby a bit dodgy) it seems that the 
> information is not reliable enough to be useful, and the cautious 
> passenger would be better advised to ignore the screen and call National 
> Rail Enquiries on their mobile every time.
> 
> John Geddes
> Derbyshire



-- 
                          Stevie D
    \\\\\       /////     Bringing dating agencies to the
   \\\\\\\__X__///////    common hedgehog since 2001 - "HedgeHugs"
___\\\\\\\'/ \'///////_____________________________________________

Stevie D wrote:

> MIG wrote:
> 
>>Sometimes the expected time is now earlier than the due time.
> 
> LDB often shows this on the final station of a route.
> York, on time. Thirsk, on time. Northallerton, on time. Yarm, on time.
> Thornaby, on time. Middlesbrough - oh, three minutes early!


Isn't that the difference between the working timetable and the public 
timetable - in the above example there's three minutes of recovery time 
built in?

"Stevie D"  wrote in message 
news:c222f1popo3rkj72qoojqhq8trorkicg80@4ax.com...
| John Geddes wrote:
|
| > Waited at Matlock station (end of single branch, unmanned, but with
| > departure screen) this morning for 08.59 departure (due in at 08.54).
| ...
| > Train arrived about 10.03. As it arrived (ie about five minutes before
| > departure), the screen changed to show that the first departure was
| > the 10.39
|
| The problem is most likely to be that the computer doesn't link the
| inward working with the outward working. So although it knows that the
| 0854 arrival is still out there, running late, it isn't able to
| transfer that delay to the 0859 departure. This is a particular
| problem for stations at the end of the line.
|
| As others have said, in this situation LDB shows the fractionally more
| accurate, if no more helpful, "No report".
|
|
| Having said that, being a terminal unstaffed station is not the only
| way to get inaccurate information. Selby is both staffed and a through
| station, with westbound trains starting from Hull, 30 miles away. Yet
| the screens rarely show accurate information when trains from Hull are
| delayed.
|
| Surely writing the software can't be _that_ big a job?


Writing the software is no longer the problem.  Perfectly good software 
exists and is on display at my local unstaffed "shack" on the edge of the 
New Forest.

It's the supply of information that is the problem.  People have a really 
hard job understanding that the computer is *not* a Magic Box, capable of 
producing perfect and outstanding performance under all circumstances, but 
is basically a glorified adding machine which processes information supplied 
according to the rules it is programmed with.  If the information supplied 
is patchy or non-existent, the best software in the world will not get you 
the right answer.

In the case of Customer Information Systems, the source of information 
normally comes either from the signalling system or from the TRUST reporting 
system.  The relative qualities of these may be gleaned from the fact that 
there is only one TRUST reporting point (Worting Junction) between 
Basingstoke and Winchester, but probably about 20 signal sections - 
admittedly this is an extreme example.

Away from the main lines, TRUST reporting points and signals are much more 
widely spaced, so the quality of information available to the CIS is 
greatlly reduced.  Some examples from my part of the world:

Between St Denys and Fareham there is ONE "stop" signal for every TWO 
stations and NO intermediate TRUST reporting points at all!  The CIS system 
at the stations fortunately works off the signalling, not TRUST, but even so 
can be in error by the 5 minutes or so it takes stopping trains to pass 
through the signal sections.  This is a real problem at the first station of 
these pairs, where the train will continue to be displayed for many minutes 
after departure, indeed until it has left the other station of the pair, 
which is the point at which the system can next "know" where the train 
actually is.   If the system ran off TRUST, it would presumably have to 
"guess" where the trains were by applying a time delay to the last report, 
so if trains were delayed in section they could be cleared from the displays 
and the next put up before actually having arrived at the station.  For 
reference, a "stopping" train is allowed about 25 minutes to pass between 
the adjoining TRUST reporting points at St Denys and Fareham.

The Salisbury to Exeter line is single-line with passing places.  The only 
intermediate signals are "gate" signals protecting level crossings and are 
controlled locally, not reporting either to TRUST or to the master 
signalling system which only needs to know whether a section is occupied or 
not to prevent another train being signalled in.  With such a poor quality 
of information available from whatever source, it is a testimony to the 
dedication of those who have set up the system in that area that it can 
perform any useful function at all.

* * *

Go around the country and this is doubtless the cause of many of the "poor 
performances" from these information systems.  The Live Departure Board feed 
is worse in many areas because for the most part it only uses the TRUST 
feed, although it seems that steps are now being taken to provide it with 
the higher quality information available in areas where CIS works off the 
signalling.  But there still exist some antiquated systems which have no 
automatic input at all but merely repeat the timetable with whatever delays 
the human operators can find out about and enter in.

At the end of the day, forecasting the arrival time of your train is like 
forecasting the weather - incomplete information input to the computer gives 
disappointing results.
-- 
                - Yokel -
    oo                               oo
  OOO                           OOO
   OO            0                OO
    ) (              I                  ) (
    ) (              /\                 ) (

"Yokel" now posts via a spam-trap account.
Replace my alias with stevejudd to reply.

In article , John Geddes
 writes

>What is the basis for updating of information on departure screens at 
>unmanned stations?
>
>Waited at Matlock station (end of single branch, unmanned, but with 
>departure screen) this morning for 08.59 departure (due in at 08.54).
>
>Screen confidently announced train as On Time even as 08.59 came and 
>went without arrival of train.
>
>So, the screen is clearly not benefiting from detailed realtime information.
>
>Train arrived about 10.03. As it arrived (ie about five minutes before 
>departure), the screen changed to show that the first departure was the 
>10.39
>
>So, the screen is clearly not updating the "next train" automatically on 
>the basis of "all-assumed-ontime unless-otherwise-informed", or it would 
>have shown next departure the moment that scheduled departure time had 
>come and gone.
>
>So what IS the basis for updating of information on screens at unmanned 
>stations? Can passengers rely on an "on time" indication to mean that 
>the train will run close enough to time to catch them mainline 
>connections? (From Matlock, if you want to drive to Derby to catch a 
>connection you need to set off before the train would have left).
>
>With this morning's "on time" train running nine minutes late (which 
>would have made some connections at Derby a bit dodgy) it seems that the 
>information is not reliable enough to be useful, and the cautious 
>passenger would be better advised to ignore the screen and call National 
>Rail Enquiries on their mobile every time.
>
>John Geddes
>Derbyshire

Noticed a similar problem with screens on the Robin Hood line Nottingham
to Worksop.

-- 
Al