The key to all of this lies in the train planning; that much has been determined as the cause many times over and isn't really open to discussion any more, no matter how some posters (and commentators elsewhere) may like to think otherwise.
It's the sheer complexities of train planning that interests me here, which seems to be vastly more complicated than bus scheduling, where almost everything will return to the same location overnight, with only one crew member to worry about at a time (although it may be two or three over the course of the day), no need to worry about any other road users timetable and by and large, as long as they have a map not much else.
I'd be delighted if people who actually know about these things would correct me, but what I'm picking up about trains, and why there is no such thing as "just run here" is that the following things need to be taken into account:
- Diagrams for each particular train (which may have additional complexity if units divide/join en route or for peak strengthening, and may not go back to the same depot as they started from)
- Diagrams for each particular driver
- Diagrams for each particular guard/conductor/On Board Supervisor (which are unlikely to be the same pattern as the driver, meaning where both are required, two separate diagrams have to be taken into account).
Of those, the train, driver and "second person where necessary" all have to be suitable for the particular stretch of line: merely having a spare driver doesn't necessarily mean they can fill a gap in the rota "because they are there."
These patterns may also have to take into account the workings of another company, which will also be subject to the same constraints, and may not be able to make the small tweak to their timetable to match the small tweak you wish to make to yours. There are also "annoying" other things such as freight paths, which may or may not be used.
There is also the need to take into account engineering work - unlike roadworks, or even road closures, where it is usually fairly easy to take the appropriate diversions, it's not so easy to divert a train (especially where you need to take the route knowledge into account).
Units, drivers and "second persons" all have to be suitable for the relevant stretches of line, which may include operational requirements (e.g. DOO/DCO), line speed, selective door opening, unit lengths to avoid fouling crossings or junctions at stations, etc.
Then, to top this all off, in an ideal world you have some regularity for the customer, even in times of disruption, trying to balance things out on a fairly even spacing for each service pattern (even if the timetable as a whole may not be even) - it's not just a case of moving the xx01 by two minutes, but also the xx31 to so you don't have a 28/32 pattern, and the xx31 has a clash that the xx01 doesn't.
I guess there could also be effects on station opening times and staff, especially late evening? (When that "just" 5 minutes may affect finishing or locking up times?)
Anything I've missed?
I have to say, I hadn't realised there was quite so much additional work that went into train planning compared to other "logistics" situations. It's also clear why you can't just drag someone out of the finance department and given them a crash course in the scheduling software (For the avoidance of doubt I'm being facetious with this suggestion; not actually thinking anyone has done that).
(Mods: No concerns if you want to split this off into a new topic or add it to an existing one if it exists. I posted it here as there appears to be some misunderstanding of what is/was involved, and I'm keen to know more about it in relation to this topic as well as more generally).