‘Congestion on the line’

[Bantamzen]

Trouble with all this is if you want to get information out to the crews, you want it to be accurate, and as things are ever changing in a live incident you don't want to be too quick sending details out which can quickly prove to be quite wrong when full information comes to light or the situation changes, and it takes a bit of time to come up with a plan of action. A suspected fatality can quickly change into a minor animal strike, and vice versa. Socials get give pretty well a running commentary if they want, and they can delete posts that might be misleading if things change.

There's only so much you can speed up giving out info before you lose accuracy, and that causes problems, potentially dangerous ones.

I agree, but I think we can also agree that more can & should be done to improve comms to the staff with boots on the ground.

 

[Annetts key]

Quality of information during disruption to all sources (including both passengers, internal and external staff) is always a challenge because when a big incident happens you suddenly want loads of info / service and fleet controllers to support workload, but 99% of the time all those extra people would be twiddling their thumbs, so no right person is going to fund those additional roles. So there has to be a balance.

And here lies the problem. As automation and "information technology" has advanced and been adopted by the railways, and as "control" centres and signalling centres (under whatever trendy name is the flavour of the month, e.g. ROC) have concentrated the operational information into one place that covers a very large area, while everything is working as intended, it's fine.

But if there is a large problem, there are now insufficient resources to handle the resulting disruption and disseminate information to all the railway employees that need updating.

So, what happened before, how were things run before the railways adopted this large control centre model?

Unfortunately I have to say this, this large control / ROC model in my humble opinion (as a former railway employee) is not fit for purpose. Having experienced how mad some of the decisions that the control that I interacted with made (note I have absolutely no problem with the individual members of staff at control).

It's hard given the current railway economic circumstances to come up with a solution. But until and unless this is done, often when incidents or train / infrastructure failures occur, the same problems with poor communication will occur. Time and time again, over and over again.

I also take issue with control deciding to issue instructions to train crew to no longer call at stations. If the timetable is already completely wreaked, unless they have taken the time to ensure that a suitable connecting train is running and will still be stopping at it's scheduled stations, what real material advantage is there in saving a handful of minutes per station in removing stops? Unless of course the issue is driver / traincrew hours related.

I say this, because I have seen a train that has had stops removed, where all that happened, is that it caught up with the (delayed) train running ahead and hence then followed on single yellows... Hence made up hardly any time. Or it got held at a junction because when the timetable is completely wreaked, it's harder to get the regulation at junctions "right".

Again let me be clear, these problems are not caused by individuals, but by the system not being fit for purpose. The system needs to be able to cope with a major incident, train or infrastructure failure and be able to get the required information and instructions to the railway employees that need it and be able to properly consider the effects of cancelling trains or station calls.

Leaving passengers on stations without properly considering their onward journey is also not acceptable (yes, I've seen this happen). In the worst case that I saw, the guard of the train that informed passengers to alight at the unstaffed station told them to catch the next service which would be arriving soon. But there was no "next" service (a report of a body on the lineside had resulted in no services on the branch line). So A bunch of passengers (more than twenty) were left without any current or useful information, no onwards transport and this went on for over one hour.

 

[Krokodil]

Yes, the view is that stacking trains up in open countryside isn't ideal, so station platforms are used incase the view is to detrain at some point. Milton Keynes, Rugby, Nuneaton etc would fill up pretty quickly.

I can understand this. What I can't understand is why a train which was not going anywhere near the fatality would be caught up in it. It cannot be beyond the wit of man to retain a platform at Stafford free of stopped trains to allow trains heading to/from Birmingham to continue to call. If you've got no choice (say if too many trains have already left Stoke and Crewe to be accommodated in the remaining five platforms) then it would surely have been better for the train referenced above to skip the stop and pass through on a through line than to be held at Crewe.

 

[SCDR_WMR]

I can understand this. What I can't understand is why a train which was not going anywhere near the fatality would be caught up in it. It cannot be beyond the wit of man to retain a platform at Stafford free of stopped trains to allow trains heading to/from Birmingham to continue to call. If you've got no choice (say if too many trains have already left Stoke and Crewe to be accommodated in the remaining five platforms) then it would surely have been better for the train referenced above to skip the stop and pass through on a through line than to be held at Crewe.

Stafford invariably becomes an Avanti Train Park in severe disruption. Even having trains stood on the through lines.

It will depend on the time of the hour, but there are points when all platforms are already occupied, and if they know the delay will be an hour or more then the signaller will always look to occupy a platform otherwise the risk of an uncontrolled evacuation becomes high, especially when people know they are a signal outside a station.

Yes it will have been short sightedness by the signaller to occupy 1, 3, 4 & 6 with southbound trains leaving just 5 to work all services to and from Wolverhampton/New St but that does happen when Nuneaton and Rugby are also fully occupied so as not to have trains stopped on the fast lines of the WCML

 

[Horizon22]

And here lies the problem. As automation and "information technology" has advanced and been adopted by the railways, and as "control" centres and signalling centres (under whatever trendy name is the flavour of the month, e.g. ROC) have concentrated the operational information into one place that covers a very large area, while everything is working as intended, it's fine.

But if there is a large problem, there are now insufficient resources to handle the resulting disruption and disseminate information to all the railway employees that need updating.

So, what happened before, how were things run before the railways adopted this large control centre model?

Information was just as poor then? Or it was all manually done at stations, which was very inconsistent.

Unfortunately I have to say this, this large control / ROC model in my humble opinion (as a former railway employee) is not fit for purpose. Having experienced how mad some of the decisions that the control that I interacted with made (note I have absolutely no problem with the individual members of staff at control).

Not sure that is a symptom of large control centres / ROCs. But having one TOC's control mostly in one place is definitely beneficial. Also decisions can appear "mad" but there's rationale behind them but in the heat of it you're going to be able to explain it all. I've found operationally quite knowledgable people still not understand controllers decisions even when explained to them. I've had disagreements on these very forums, with detailed explanations and people just go "oh that's just an excuse".

I also take issue with control deciding to issue instructions to train crew to no longer call at stations. If the timetable is already completely wreaked, unless they have taken the time to ensure that a suitable connecting train is running and will still be stopping at it's scheduled stations, what real material advantage is there in saving a handful of minutes per station in removing stops? Unless of course the issue is driver / traincrew hours related.

It can be much more than a "handful of minutes". You also have to consider any conflicts en route and the importance of presenting at key junctions on or as close to on-time as possible. Also if the following service is right behind anyway, then the actual impact is not as severe as it seems.

I say this, because I have seen a train that has had stops removed, where all that happened, is that it caught up with the (delayed) train running ahead and hence then followed on single yellows... Hence made up hardly any time. Or it got held at a junction because when the timetable is completely wreaked, it's harder to get the regulation at junctions "right".

It's always a judgment call, and depends on a whole host of circumstances. And of course nobody gets its right every time and is reliant sometimes on good regulation from a signaller. But "do nothing" will also lead to residual disruption for much longer.

Again let me be clear, these problems are not caused by individuals, but by the system not being fit for purpose. The system needs to be able to cope with a major incident, train or infrastructure failure and be able to get the required information and instructions to the railway employees that need it and be able to properly consider the effects of cancelling trains or station calls.

Well it is a public transport network, ultimately there is going to be disruption and occasionally major incidents. The focus needs to be on looking at performance figures, resolution times and response capacity in order to have fewer and shorter disruptive incidents in the first place. Controllers are just trying to work with the cards they have been dealt.

Leaving passengers on stations without properly considering their onward journey is also not acceptable (yes, I've seen this happen). In the worst case that I saw, the guard of the train that informed passengers to alight at the unstaffed station told them to catch the next service which would be arriving soon. But there was no "next" service (a report of a body on the lineside had resulted in no services on the branch line). So A bunch of passengers (more than twenty) were left without any current or useful information, no onwards transport and this went on for over one hour.

Again, nobody is saying mistakes haven't been made. But even with best intentions during certain incidents, there can still be long gaps in service because there simply is no alternative, particularly when trains cannot run. Work on stranded trains and holding trains in platforms has seen improvements in the past few years.

== Doublepost prevention - post automatically merged: 22 Oct 2025 ==

I can understand this. What I can't understand is why a train which was not going anywhere near the fatality would be caught up in it. It cannot be beyond the wit of man to retain a platform at Stafford free of stopped trains to allow trains heading to/from Birmingham to continue to call. If you've got no choice (say if too many trains have already left Stoke and Crewe to be accommodated in the remaining five platforms) then it would surely have been better for the train referenced above to skip the stop and pass through on a through line than to be held at Crewe.

Depends how many trains are stuck. Ideally you start turning trains closest to the block back Northbound and run to a pre-agreed contingency plan which involves shuttle services, but this can take some time to implement.

 

[Falcon1200]

Also decisions can appear "mad" but there's rationale behind them but in the heat of it you're going to be able to explain it all.

Yes indeed.

But "do nothing" will also lead to residual disruption for much longer.

A Control colleague used to say that the best course of action was to make the right decision, the second best was to make the wrong decision, and the worst was to make no decision at all. And Controllers, just like anyone else, can only work with the information they have at the time (despite the subsequent reviews often believing that we should have been able to foretell the future....)

 

[Horizon22]

A Control colleague used to say that the best course of action was to make the right decision, the second best was to make the wrong decision, and the worst was to make no decision at all. And Controllers, just like anyone else, can only work with the information they have at the time (despite the subsequent reviews often believing that we should have been able to foretell the future....)

Oh the hindsight police are the worst!

 

[SuspectUsual]

This always made me laugh - there was a period about 8 years when Northern were in utter meltdown and their Twitter feed frequently had references to “congestion of the line” which to me sounded like a medieval illness

 

[Annetts key]

A Control colleague used to say that the best course of action was to make the right decision, the second best was to make the wrong decision, and the worst was to make no decision at all. And Controllers, just like anyone else, can only work with the information they have at the time (despite the subsequent reviews often believing that we should have been able to foretell the future....)

Of course making no decision at all is not helpful. I don't think anyone has asked for that. I certainly haven't.

Foretell the future is what control do try to do, often when they have incomplete information. That's part of the problem. It's a dilemma, how long do you wait to see if there is any new, more accurate information?

And the trouble is, there may not actually be any later, more accurate information. For example, if a report comes from a member of the public or indeed anyone else who cannot be contacted in order to ask for more information.

Making decisions with incomplete information, that is understandable, but it will be mean that they will get things wrong. Especially if/when the later information is different to the initial report.

And the railway is sometimes guilty of managing to lose important details when information is passed on. Especially when the information is passed through a number of members of staff.

But my point is that under the current system, control are far too often trying to handle too many variables, unable to have time to give enough attention to the outcome of their decisions and unable to keep all the railway employees who need to be updated, up to date.

It's a high pressure job when there are things going wrong. Especially when it's causing major disruption.

Oh the hindsight police are the worst!

There will always be hindsight and after the fact, we should have done it my way, people.

Just as there will be people who make decisions under pressure that turn out to have been less than optimal.

The trouble is, the system is not working well enough. The railway doesn't appear to be learning from the mistakes that have now been occurring for many, many years.

The problems discussed in this topic are not new.

 

[Horizon22]

There will always be hindsight and after the fact, we should have done it my way, people.

Just as there will be people who make decisions under pressure that turn out to have been less than optimal.

The trouble is, the system is not working well enough. The railway doesn't appear to be learning from the mistakes that have now been occurring for many, many years.

The problems discussed in this topic are not new.

Of course. But fundamentally, they are intrisic problems without spending loads of money on the situation and ultimately being very inefficient with colleagues' time. That is not to say new approaches with technology and integrated systems can't help more.

Already you have people (certain managers) turn up to a control room, see controllers having a chat when everything is going well and wonder why all these people are employed on pretty decent (relative) salaries doing very little. Ultimately, control teams earn their wage when things do go wrong.

 

[Annetts key]

Already you have people (certain managers) turn up to a control room, see controllers having a chat when everything is going well and wonder why all these people are employed on pretty decent (relative) salaries doing very little. Ultimately, control teams earn their wage when things do go wrong.

Yes, and there should be no place in the railway for such managers.

I've had to point out to a certain manager, which would you prefer, staff not being busy because everything is running well or delays racking up like a demented gambling machine and insufficient staff to attend and deal with all the problems?

Funny enough they did not look best pleased at my question.

A good manager should be able to understand what happens during the "bad" times as well as what happens during the "running normally" days.

And, does anyone ask a firefighter what they do when there are no fires to fight or no car accidents that need fire fighters to attend?

 

[Bald Rick]

And, does anyone ask a firefighter what they do when there are no fires to fight or no car accidents that need fire fighters to attend?

They do, actually, hence the closure of many fire stations!

Re control. I have covered many roles in control(s) in my time. There are many things that could be done to improve comms, but not much that could be done to improve process. Ultimately in very simple terms the process is, and always will be, person(s) on ground reporting > incident controller > service controller > person(s) on ground actioning. The issues get tougher when there is more than one incident occurring, and or lots of people needing to action.

 

[Horizon22]

Ultimately in very simple terms the process is, and always will be, person(s) on ground reporting > incident controller > service controller > person(s) on ground actioning. The issues get tougher when there is more than one incident occurring, and or lots of people needing to action.

And in fact that is quite a normal process for most emergency / other public transport control roles, because there is just a necessity as to the minimum possible communications.

 

[Annetts key]

They do, actually, hence the closure of many fire stations!

My point was do the public (the primary users) ask.
How many fire stations there are and the service provision depend on meeting the minimum requirements (determined in part by central government and associated funding arrangements like grants) plus funding from local authorities or regional government, or for Fire authorities, funding directly via a Council Tax levy. In my town there has been no reduction, and in the wider area, it was years ago that some part time stations in small towns were affected. Since then there have been no significant changes. And in some places additional services now exist, for example my local station has a hovercraft.
But this is going off topic.

Re control. I have covered many roles in control(s) in my time. There are many things that could be done to improve comms, but not much that could be done to improve process. Ultimately in very simple terms the process is, and always will be, person(s) on ground reporting > incident controller > service controller > person(s) on ground actioning. The issues get tougher when there is more than one incident occurring, and or lots of people needing to action.

Processes can be improved. Communications without doubt do need improving.

IMHO, part of the problem with centralised control is:

1. Can create single points of failures. Have you seen what happens when the building where control is located has to be evacuated due to the fire alarm going off? Or what happens when the telephone system used by control fails?
2. Everything (well, almost) has to go via control, whereas in some situations, a more distributed approach may work better.

 

[WAB]

1. Can create single points of failures. Have you seen what happens when the building where control is located has to be evacuated due to the fire alarm going off? Or what happens when the telephone system used by control fails?

The benefits of colocating route and service controls outweigh the very rare occasion where the control office is incapacitated.

1. Everything (well, almost) has to go via control, whereas in some situations, a more distributed approach may work better.

What do you propose?

 

[Krokodil]

What do you propose?

Don't GWR have someone based at Exeter? Titled "Area Operations Manager" or something like that, who can make a lot of decisions regarding local services without needing to get permission from Swindon first. I don't think that it would be appropriate for this case, but it can work quite well where you do have fairly self-contained parts of an operator.

 

[Bald Rick]

IMHO, part of the problem with centralised control is:

1. Can create single points of failures. Have you seen what happens when the building where control is located has to be evacuated due to the fire alarm going off? Or what happens when the telephone system used by control fails?
2. Everything (well, almost) has to go via control, whereas in some situations, a more distributed approach may work better.

And yet almost every railway, power grid, gas grid, air traffic control etc in the developed world is centralised to some degree. They cant all be wrong.

 

[Falcon1200]

Already you have people (certain managers) turn up to a control room, see controllers having a chat when everything is going well and wonder why all these people are employed on pretty decent (relative) salaries doing very little.

We always considered that there are two kinds of railway manager; Those who, seeing Controllers drinking tea and chatting, know the railway must be running well, and those who think 'what the hell am I paying all these people for?'. The latter often tried to find extra work for us, which our boss went along with up to the point where he asked 'what part of our work do you want us to abandon so that we can take on this extra duty?'. All this, BTW, dates from BR days so is nothing new.

 

[Adrian1980uk]

And yet almost every railway, power grid, gas grid, air traffic control etc in the developed world is centralised to some degree. They cant all be wrong.

We also have to remember we're not in the 70s 80s and 90s anymore, there is process to be followed and not a person on the scene who can totally call the shots. The process keeps everyone safe although may not be the most efficient in times of disruption, things like the preferences to hold trains at stations rather than in the middle of the country are there for good reason. The teams dealing with the incident don't need the pressure of seeing train after train queuing as they get to the incident.

Local control rooms may have more local knowledge but again the autonomy to use that knowledge is generally not there anymore in any industry, you see that in why small private companies can be more agile than large corporates for similar reasons, has to be a process to follow.

Communication should be much easier now but the pressure for information mostly outstrips the actual information available, I work in the IT industry and have seen many system issues where you have a team of people asking when something is going to be fixed and the support teams are giving best guesses where the real answer is that if we knew what was wrong in most cases the fix is measured in minutes if the problem is known but again the actual fix is minutes but the process to test and release can be hours or even days.

 

[Bantamzen]

We always considered that there are two kinds of railway manager; Those who, seeing Controllers drinking tea and chatting, know the railway must be running well, and those who think 'what the hell am I paying all these people for?'. The latter often tried to find extra work for us, which our boss went along with up to the point where he asked 'what part of our work do you want us to abandon so that we can take on this extra duty?'. All this, BTW, dates from BR days so is nothing new.

It's not just the railways where that applies, for many years I worked as IT Support for a large office, part of a nationwide public sector body. And yes, there were days when literally everything ran smoothly and I wouldn't be needed once. Some managers did ask what I did with all this time, and the truth was sometimes there was nothing to do. That's how some jobs just are, and it's a cultural thing that managers need to wrap their heads around. The key isn't what staff are doing when systems are running smoothly, and all work to keep that going has been done. It's about having resources that can be stood up immediately for when things don't go smoothly. In my experience it costs more to not have that available resource, because urgent issues cannot be immediately be addressed, and that often leads to a pile up of subsequent issues which can overwhelm systems, leading to a longer resolution time.

And if some managers still find it hard to swallow that staff may have little to do in working hours, there's always the opportunity to upskill, even run simulated issues to fine tune contingencies, discuss better ways to work etc, etc. In my time in IT Support, I often used quiet times to make improvements to team, such as building databases to better track assets / incidents (this was before we had a centralised system) and using that to build a site map, so that we could quickly identify individual machines, their associated ports / switches / servers etc. This allowed us to improve response times, knowing what build a reported faulty PC had, it's location and even if other similar incidents were being reported. Just simple little things to do when it's quiet that lead to improvements make it very hard for even the most zealous manager complain (though some will try).

 

[Horizon22]

Don't GWR have someone based at Exeter? Titled "Area Operations Manager" or something like that, who can make a lot of decisions regarding local services without needing to get permission from Swindon first. I don't think that it would be appropriate for this case, but it can work quite well where you do have fairly self-contained parts of an operator.

GWR have multiple of these at various hubs I believe. They are sort of like auxillary controllers, but they will still need permission / direction from route control on many aspects. They are also not the only company to emply such an approach, although more common on long-distance and intercity operators. It's a useful role to have with an element of autonomy to relieve some pressure.

The process keeps everyone safe although may not be the most efficient in times of disruption, things like the preferences to hold trains at stations rather than in the middle of the country are there for good reason. The teams dealing with the incident don't need the pressure of seeing train after train queuing as they get to the incident.

This is mainly to prevent stranded passengers who then might need rescuing. Of course it is infinitely easier to help someone "stranded" at a station than it is in the middle of nowhere with poor acess.

Communication should be much easier now but the pressure for information mostly outstrips the actual information available, I work in the IT industry and have seen many system issues where you have a team of people asking when something is going to be fixed and the support teams are giving best guesses where the real answer is that if we knew what was wrong in most cases the fix is measured in minutes if the problem is known but again the actual fix is minutes but the process to test and release can be hours or even days.

You are right that now people are hungry for information that arrives sooner, is more detailed and more bespoke than ever before. No doubt this is because of social media, digital devices and connectivity being almost instanteous now. What used to cut it just doesn't sometimes any more. Also it may be cultural but people seem more 'on edge' with a lack of information compared to 20 or even 10 years ago and assume the worst. There have been several - reasonably well publicised - cases on the London Underground where people have started pulling egresses, smashing windows and generally panicking when trains have been at a stand for mere minutes. Most metro operators than run DOO request the drivers make announcements within 30-60 seconds of coming to a stand and are assessed against this metric.

Sometimes however the full information is an unknown, and in the initial stages you can only provide very basic information. This is of course better than nothing, but people fundamentally want the classic three things:

1. What is the problem/cause?
2. What is the impact going to be to my journey?
3. What is the advice for me as a passenger?

 

[spp]

I don't believe that response was sarcastic, but just wanted to give the the angle the Conductor would have had.

Arrived into Crewe unaware of any issues
Sat in platform on a red signal.
Driver likely informs of being held by signaller (may have misunderstood as simply being held at the signal)
Conductor relays message about being held at red signal, awaiting more information.
Train departs late, conductor now receives information from control with ammended calling pattern.

Please be aware that just because a tweet has gone out, that doesn't mean the conductor has that information. Twitter is not on our work phones.

The app used for Comms on our work phone no longer receives ANY messaging, and likely the phone lines to control would be engaged as they are contacting all adds ted trains on the route, prioritise those in the immediate area and are possibly trapped.

Absolutely communication should be better to train crew, and there should have been better announcements whilst you were stuck at Crewe. Don't presume that train crew have even the same level of information that you had.

Sorry delayed response been away for a few days (more delayed trains ha), won’t go into specifics as not to drag the thread on any further it’s been and gone but;

Appreciate you explanations. However it kind of reiterates my original gripe. You say the train crew may not have had information which I accept but that’s my point. They should have!

If a laminate like me can find it out through some digging (social media RTT etc) then there really needs to be a system where train manager, driver and social media team etc all have the same information ASAP (to pass on the most informative update to passengers).

With today’s modern methods of communication I don’t think that’s too much to expect from our railways fragmented as they are

== Doublepost prevention - post automatically merged: 25 Oct 2025 ==

Trouble with all this is if you want to get information out to the crews, you want it to be accurate, and as things are ever changing in a live incident you don't want to be too quick sending details out which can quickly prove to be quite wrong when full information comes to light or the situation changes, and it takes a bit of time to come up with a plan of action. A suspected fatality can quickly change into a minor animal strike, and vice versa. Socials get give pretty well a running commentary if they want, and they can delete posts that might be misleading if things change.

There's only so much you can speed up giving out info before you lose accuracy, and that causes problems, potentially dangerous ones.

the fatality happened hours before.

If they are happy with the accuracy of the info to give it out by social media they should be happy to give it out to people actually stuck on a train

== Doublepost prevention - post automatically merged: 25 Oct 2025 ==

I apologise as I didn't intend it to sound sarcastic, I was just giving my point of view (and experience).


If you work for any TOC in any remotely customer-facing role then you should understand how poor the information is.


They will be in the same room. They would likely be able to tell the social media team face-to-face in only 10 seconds of walking.

Apology accepted, maybe I misread the tone so I too apologise for calling you sarcastic.

Sorry just as a point of order I don’t work for any TOC the quoted part should have read ‘I and an avanti employee’ had worked out the info being given was incorrect as he too was waited for the cross country train . Me based on RTT him I pressure something similar