Where it all started
Below is a reproduction of my Reddit post that got this whole thing started. Leaving it here for your interest.
Original title: My thoughts on the recent signalling fiasco
TL;DR: Don’t try to run if you don’t know how to walk.
I’m just a student who read a few manuals and had some conversations with fellow enthusiasts, but I’d like to drop my 2 cents on the matter. If anyone has any better information, I would appreciate your input.
So everyone must be very familiar about the recent issue with the “signalling upgrade” along the North South Line and the mayhem that it brought
This is not something new to the rail industry. London has done it before us (even installing the same Thales solutions) and so have KL and New York, who are still grappling with problems at the time of this writing.
Resignalling projects tend to be difficult to manage already, even when done slowly and carefully. In fact, the resignalling project on the NSL should have been done last year, but they missed the deadline. However, it would seem that LTA is being very rash about the matter, choosing to deploy (relatively) untested technologies into the system considerably ahead of its time.
However, I believe that LTA do have their reasons. Firstly, the urgency to open the Tuas West Extension, which should have been opened last year but delayed. And secondly, the new C151Bs (white trains). Both these projects are only equipped with the new CBTC system (“CBTC” in short), and it would be important to have test coverage on the system to ensure the most can be made out of these new improvements.
Most other systems have chosen to take a phased approach towards resignalling projects. A few stations at a time would be the best approach — that minimizes inconvenience while allowing test coverage to still be as expansive as possible. However, it seems that LTA has not chosen this approach, instead deciding to move 40% of the entire system at one go.
And to add fuel to the fire, as we all know they decided not to slowly and progressively expand the testing to Saturdays, then off-peak hours, and only then do they turn the system on to serve peak hour traffic. This caused considerable mayhem in the first few weeks, and things are probably going to get worse before they can get better.
We already know that a phased, one-step-at-a-time approach is possible, since under the current arrangements all trains must switch at Pioneer into CBTC territory, even if they are not continuing into Tuas. This is because the turnback siding east of Joo Koon station is located in CBTC-only territory, and it cannot be entered under the old Westinghouse Fixed Block system. This causes delays while trains have to “reboot” into the new system, which takes about one or two minutes.
The first phase would ideally be the stretch between Tanah Merah and the airport. This is an isolated line, where delays on this line tend to not have a lot of effects on the rest of the network. However, as far as I’m aware, the nearest depot (Changi) did not receive a full upgrade to CBTC, since it will be replaced by a new facility before 2027. Due to this, the new trains cannot run on the airport line since there’s nowhere else for them to go should one be disabled. Going west, the nearest place to store a disabled train is Outram Park, 30 minutes away. And the CBTC system must be ready for these trains to hobble there on their own power. It won’t be so until next year.
So perhaps they should have phased the North South Line upgrade first. Better to piss off only everyone living between Jurong East and Kranji, for example, than those on the entire line. This is the approach London Underground took with the Jubilee and Northern lines, and it has helped them considerably. London Underground also took weekend closures to install and test the system, but that’s something that might be politically unpalatable here.
Alternatively, they could have run the CBTC system in a sort of “shadow” mode, where the old system still controls the train and the CBTC system “learns” how to run the railway before the CBTC system actually takes over (did this actually happen? Please let me know). London did this, and they managed to reduce quite a fair bit of bugs even before the first passengers boarded a CBTC-equipped train.
However, this phased approach also does not allow them to make use of their new trains. Under a phased approach, SMRT’s 45 brand new trains will be gathering dust in Tuas Depot for a year or two, very limited in where they can go, while the existing ones are strained more than necessary. Also, there would have been even more delays to when the Tuas extension would be able to open (perhaps next year, or the year after? These stations were completed in 2016!)
In response to that, the KL Ampang/Sri Petaling Line also faced a similar problem, where they had a set of trains with a legacy system only and another set of trains with a new system only. They ran a shuttle forcing people to change, which perhaps could be what we could use to make the most out of a difficult situation.
Ultimately, to me, this sounds more of a failing of LTA’s project management abilities. I’m fairly sure that I’m one to talk as a commuter, not knowing of the challenges they face “inside”, but I believe they could have done better. All this mess could have been avoided (to a certain extent) had the entire project be planned better.
Hindsight is always 20/20, of course.
DISCLAIMER: This is a purely opinion writeup based on my knowledge of the matter at hand, with additional context provided so one can keep up with the writeup, and by no means should any of the information provided here be considered accurate. Should any stakeholder in the project wish to contact me regarding any misrepresentation in the writeup, my inbox is open.
Please don’t cite my views in any future articles, as I’ve said this is merely an opinion piece.