CarrotPie
Member
Which leads to the obvious question: should they be, or should they have more slack snd flexibility built into them to allow for easier recovery from disruption, and how might this be achieved?
-
Our booking engine at tickets.railforums.co.uk (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!
Should timetables allow for easier recovery from disruption?
- Thread starter CarrotPie
- Start date
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R
RailUK Forums
You need to have more trains and crew, more platforms (at Waterloo particularly) and sidings ...
306024
Established Member
The best timetables and associated resource plans get the balance right between journey time, and operational cost and reliability. How many people on hear moan when there is slack in the timetable, for whatever reason, yet complain when their train is late.
The current GA timetable is a good example. It can cope with minor delays, yet is approaching Swiss levels of reliability and punctuality.
The current GA timetable is a good example. It can cope with minor delays, yet is approaching Swiss levels of reliability and punctuality.
DMckduck97
Member
Everyone wants the diagrams to extremely tight for productivity, but then wonder when an incident like this happens why half the crew for the next working is in one place miles away from the other and the service can't be recovered.
Sometimes its not as simple as running shuttles between two stations
Its not so much the timetables, as the resource plans that support them. It’s no coincidence that train operators that have only one or two types of train (that all drivers can drive), simple routes where most crew sign most routes and diversions, and plenty of spare crews, are those that are the most reliable.
306024
Established Member
Indeed, and traincrew costs in particular are a big cost in running a railway. Hence the growth in computer systems to produce ever more mathematically efficient diagrams. But the old maxim, keep the crew with the train wherever possible, is the best way to aid service recovery. Much easier (in theory) on short routes than long distance however.
miklcct
On Moderation
In my opinion, the best way for timetable recovery is to make it such that recovery is not needed, i.e. run all trains on the same route with the same calling pattern, so no one actually care if his train is delayed. No one actually cares if the West Hampstead to Stratford train is the on-time 12:00 or the 11:45 delayed by 15 minutes.
For example, for a service which is supposed to run every 7-8 minutes, if a delay occurs, just simply let the delay propagate until all trains on the line are delayed by 7 minutes, and declare one train as cancelled by that point. If a train is broken down, just run the line at a reduced frequency until a replacement is added.
In order to achieve the above, interworking should be avoided at all costs, while crew / rolling stock diagrams should be self-contained for a particular route.
For example, for a service which is supposed to run every 7-8 minutes, if a delay occurs, just simply let the delay propagate until all trains on the line are delayed by 7 minutes, and declare one train as cancelled by that point. If a train is broken down, just run the line at a reduced frequency until a replacement is added.
In order to achieve the above, interworking should be avoided at all costs, while crew / rolling stock diagrams should be self-contained for a particular route.
CarrotPie
Member
No, but they do care if they're waiting for the 1145 and it doesn't turn up.
Or cancel a service at the start, don't let the delay propagate and everything's on time again.
How many routes in the UK operate at such a frequency though? And how many routes have sufficient extra capacity to just absorb the passengers from a cancelled service and remain running to time?
stevieinselby
Member
Which is related to one of the reasons that the LU Circle Line became the Paperclip Line ... with trains running in continuous service (and interacting with other lines) there was no opportunity to recover or rebalance services in the event of disruption. By having a terminus at Hammersmith and another at Edgware Road, it has become much easier to take a train out of service and to build slack into the diagrams.
Bletchleyite
Veteran Member
That's how Merseyrail used to be, and you'd fix it by cancelling one* and stepping everything up. They seem to prefer running semifast now, though, which was almost unheard of when I was a kid.
There are other approaches, though, such as the classic south WCML one of "run everything into Euston in the morning peak even if it's late, and sort the mess out between the peaks with a few cancellations". That makes sense with a long distance commuter operation where each unit and crew (bar a couple of Trings) is only going to be doing one peak journey into the main terminus. This seems largely still to be the case - cancellations between 0600-0900 south from Bletchley are near enough unknown even if there are delays.
On the general point I have long held the view that simple, non-interworked diagrams are best, Northern do the total opposite of this so there's all manner of knock-on all over the network. As an example, the classic single-unit 1h10 ish frequency Ormskirk-Preston service was what you might term "watch-settingly reliable", but now the "Lancashire triangle" group drops services all the time as there's too much knock-on between the branches. The downside is that they are typically less efficient, and you've got stuff like the bizarre Avanti union agreement that a given member of staff can't do a given line twice in a day (which only really applies to Birminghams anyway).
* i.e. popping it out at Kirkdale and putting it back in on time having missed a round trip out, or whatever the Wirral equivalent was.
CarrotPie
Member
Having said that, the Yamanote Line in Tōkyō sorts this out by running every couple of minutes (so delays are imperceptible) and having its main maintenance depot at Ōsaki, where all trains are timetabled to wait a few minutes, which means slotting trains in and out of service is easy peezy.
Last edited:
stevieinselby
Member
The only way to have slack and flexibility built into the timetables, eg in order to allow better recovery from disruption, is not to use the network to its full capacity.
As soon as the network is maxed out, any delay or problem is going to propagate – and the more interconnected the lines are, the further it will propagate.
So the question becomes, do we want to put more money into building and maintaining surplus capacity and redundancy in the system, if it gives us no benefits when things are working well? There needs to be some level of resilience in the network, but too much would be a poor use of resources. I would rather look at investing that extra funding into preventing problems in the first place – maintenance of trains and infrastructure, prevention of people or obstructions on the line, etc.
Bletchleyite
Veteran Member
Yes, absolutely we do. Just go and look at how well SBB works. Plenty of platforms, plenty of line capacity, decent dwells etc lead to a punctual network and easy recovery when things go wrong.
It wasn't previously possible to do that on London commuter routes, but it now probably is with reduced commuting (e.g. the south WCML timetable now is reduced a bit over pre-COVID but is now nicely clockface*), but as an example the idea of 6-7 trains per hour on TPE North is nuts - drop it to 4 and do the work to allow ten car running when it's busy and it'll all work far better. Similarly it might be possble to shove 20 trains per hour through Castlefield in theory (think Thameslink, to use an old slogan) but in reality there are too many outside influences for that to work.
* But not resilient enough as some turnarounds are too tight, to be fair.
As an interesting side note, here are some Google Translated comments from a Japanese website written by a Japanese man who previously lived in London:
I have a lot of respect for the control and planning staff and I have met quite a few over the last few years via work who generally do a good job under difficult circumstances held back back by inflexible and old fashioned systems.
One thing that seems to hold back implementation of amended timetables seems to be that trains have to be added removed or amended individually during times of distruption.
Also a well designed IT system should be able to allow removal, amendment etc en mass and could also help identify problems caused by short term timetable or diagram changes cause for instance a train not getting back to a depot when needed and suggest potential solutions to help reduce knock on effects.
I think one way to reduce disruption would be to invest in IT systems for the control and planning function to allow it to better cope with disruption. This would need to build on the strengths of the people who work in these roles and complement their experience and strengths to achieve the best outcomes.
Basically I think the control and planning functions need modernisation.
One thing that seems to hold back implementation of amended timetables seems to be that trains have to be added removed or amended individually during times of distruption.
Also a well designed IT system should be able to allow removal, amendment etc en mass and could also help identify problems caused by short term timetable or diagram changes cause for instance a train not getting back to a depot when needed and suggest potential solutions to help reduce knock on effects.
I think one way to reduce disruption would be to invest in IT systems for the control and planning function to allow it to better cope with disruption. This would need to build on the strengths of the people who work in these roles and complement their experience and strengths to achieve the best outcomes.
Basically I think the control and planning functions need modernisation.
306024
Established Member
Well compared to when I worked in Control there has been a massive leap in technology. But in those days you couldn’t even see where the trains were!
But seriously, investing in IT is easy to say, more difficult to achieve successfully, especially in our fragmented railway. You have to be aware of software developers who simply want to sell you what they already have, not necessarily what you want. Developing from scratch with the full involvement of experienced staff assumes those staff can be released from the day job, and someone will sign the huge cheque.
There have been technology improvements, but in disruption there is no one size fits all. And whatever your technical plan, it still needs traincrew to carry it out. The geographical position of those staff varies minute by minute, and I‘d suggest logistically is the single most difficult aspect of managing service disruption.
DMckduck97
Member
And whether the punters like it or not, the logistics of having traincrew work these ad hoc shuffles sometimes just doesn't work and outside of the enforced overtime, the last thing you want is a train stuck on a running line with a driver over his/her hours refusing to take it.
I can see why these short notice plans aren't implemented because it can get very messy quickly.
Unfortunately it's not as easy as plucking a guard and driver out of a diagram and saying here work these trains.
306024
Established Member
And before someone gets upset it is not necessarily traincrew being difficult for the sake of it. Many go the extra mile (literally) during disruption. But if you expect to finish work at xx.xx, then have an urgent appointment, school run or whatever, what else can you do.
This is correct but for accuracy, Ōsaki, not Ōmiya.
stevieinselby
Member
Or, in a lot of cases, reaching the legal limit on the hours they can work.
DMckduck97
Member
It's usually a simple case of well if I can work past my hours in the office and help out with something why can't they...
People (on this forum especially) should be well aware of the hidden report and its recommendations and how much limitation it can have on service recovery.
Bletchleyite
Veteran Member
They mostly do!
CarrotPie
Member
What limitation does it have?
DMckduck97
Member
They certainly do aswell in my neck of the woods!
Mostly around working hours and rest periods, and a branch which falls into this is break periods, drivers and guards need to have started and finished a break between a specific set of hours, no ifs buts or maybes.
So therefore limiting service recovery because all of this has to be accounted for by control and resource management whilst also managing the incident and displaced crews.
It's simply not as easy as saying we have this many staff, let's run this from now until now. Even though it sounds easy in practice.
CarrotPie
Member
Righto, thanks for that.