Transport has always had its timetables. The stage and mail coaches that ran up until the mid nineteenth century used basic timetables but, as we shall see later in this article, the spin-off resource diagrams would have been as complicated as anything used in the modern day railway.

The coaches of old, and our modern buses, have the advantage that they can overtake each other and they have – or had – to some degree, quite reasonable stopping power. Trains on the other hand have great difficulty overtaking and their stopping power, or lack of it, is a significant issue.

The National network will be introducing a new timetable on 13 December and so Rail Engineer thought it would be the right time to talk to Fiona Dolman, capacity planning director at Network Rail to see what is involved.

The December timetable is not really new. It’s the evolution of years of development that is revised to cater for a multitude of external influences twice a year. Many trains will not be changed. At the other end of the scale, whole new services will be introduced as a result of new infrastructure being commissioned or new rolling stock being introduced.

First, sharpen your pencil

How does it all begin? There are many inputs into a service. There are franchise obligations that fix the level of service – the number of trains, the frequency and the number of seats available being but a few of the parameters.

There are franchisee aspirations which overlay the minimum obligations. There are influences from outside the industry from a wide variety of stakeholders. Put everything into the melting pot and a basic set of train graphs can be constructed.

The train graph, at its simplest, is a visual representation of the progress of a train showing distance travelled against time. The trains appear as lines on a sheet of paper, or more recently, on a computer screen. To the practiced eye an express train will have a completely different appearance to a stopping train or a freight train. One of the golden rules to be followed is that the lines must not clash. Indeed they should be kept apart from each other by a minimum distance. This represents the headway between the trains – the distance determined by the signalling of the line. The ‘more practiced eye’ will be able to make sense of some of the more complex train graphs accompanying this article.

Fitting in all the trains needed to run a service, at the same time as satisfying timetabling rules, is the job of a train planner. The timetable rules include such issues as station dwell-time, terminus turn-round time, headways on plain line and at junctions. Coupled with the fact that different trains have different acceleration and braking characteristics and that timing is further influenced by line gradients, this can all get a little complicated.

image006 [online]

It would be tempting to look at the timetabling rules and to expect that these could be expressed digitally, so that the whole exercise could be plumbed into a grand timetable computer programme. After all, we have had road route planning software for years.

GPS software in a car will sort out routes and timings against embedded rules to produce what is, in effect, a timetable for a journey. Can that technology not be aligned with the railway network? Well, no is the short answer. The slightly longer answer is that there are timetabling programmes that do a great deal of the work, but not all.

Manual intervention

It is possible to build a basic, fully annotated train graph on a computer, but the finer practical details still need human intervention. The problem that has bedevilled train planners, throughout not only the UK but the whole of the planet, is that some of the timetable rules are inconsistent and erratic.

Some, for example terminus turnround times, involve simple integers for the most part, but then they disintegrate into a complicated string of words that are dependent on a long list of factors influenced by other illogical constraints elsewhere on the network. There have been valiant attempts to crack the problem, but the human brain is still the best tool to sort out all the practical intricacies of the complex national timetable.

Fiona is realistic. “It’s one of our aspirations to have a system that will not only populate a timetable but also test it for practicality with the further ability to feed in a variety of ‘what ifs’ into the mix.”

The basic timetable is just the start. It’s simply a collection of figures, a scattering of lines on a screen. Downstream is a welter of resource detail. There is little point having a bonny timetable if there are not enough trains available to fit all the paths. The trains need to match the likely demand for seats. There needs to be enough train crew available to staff the trains. Trains need to be in the right place at the end of the day so that the service can be started the next day. Rolling stock has to be inspected, serviced, and overhauled at specific intervals. All this has to be built into the timetable proposals.

The compilation of any timetable can stretch back to around five years before the press date. Train operating companies submit their bids to Network Rail, the custodian of the national timetable, so that the proposals can be checked against all the other demands by all other interested train operators.

Conflicts at junctions need to be resolved. Platforming at terminals – often a critical factor – has to be checked so that incoming services don’t block in those about to depart.

After a series of iterations/negotiations, an agreement is reached on the likely shape of the services. These need to be checked back with rolling stock arrangements and also checked against the conditions of service of the train crew.


This is all fine so long as everything goes to plan. But, of course, real life can get in the way. There is the expected and the unexpected. Expected issues include bad weather and engineering work. Unexpected issues could involve the sudden loss of a route.

Looking right back to the way that a new timetable is compiled, it is pretty obvious that any changes because of planned engineering work would demand similar timescales in planning. A timetable catering for the loss or heavy modification of a route for engineering is in effect a new timetable. Downstream alterations to stock and crews are just as complicated and there’s the additional issue of planning the transition from normal working to engineering working and back again so that a normal service can resume without anyone noticing.

Train planners are sanguine about engineering. It happens, and it’s the job of a train planner to… plan trains (and buses for that matter). Some engineers might think that cancelling a possession might be doing the world a favour. This is not necessarily the case. In fact, the nearer to the possession, the more difficult it is to disentangle all the rolling stock and train crew arrangements. There is no quick ‘undo’ button in train planning!

image007 [online]

There are times of major disruption when one of a suite of pre-planned special timetables are brought in. These are developed well in advance and are switched on in the event of the loss of a complete route or terminus, for example. For any sudden emergency the initiative for running a service is handed directly to the route controller who retains control until matters stabilise.

Fiona makes the valid point that “for the railway to work right across the network and across administrative boundaries requires complete commitment from everyone, from the front line right through to the architects of the service.”

It’s an enormous logistical exercise complicated by trains’ innate inability to shuffle their relative positions. And it is in the gift of engineers – if they’re not careful – to thoroughly stitch up the service with untimely ambushes.

The stage and mail coaches of old had basic timetables. They had complicated staffing and motive power issues (horses) but they were able to defend themselves from ambushes. Their drivers were issued with blunderbusses – items that don’t seem to appear on train rosters these days.