A crisis of integration
We now know what went wrong last September.
As always, hindsight is 20/20. Full investigations which have taken close to 9 months have revealed plenty. And I find these pretty damning, to be honest. All the more so as the powers that be have spent plenty of money on preventive maintenance systems designed exactly to prevent this sort of incident.
Does that mean such an approach is mistaken? I don’t think so. If anything, the problem is getting all of it to work together. The most obvious symptom of this is when one screen on the NEL tells us about disruptions we already know of elsewhere, instead of next train and destination information for the NEL service itself.
When the monitoring stops
But that’s not just where the integration issues end.
The findings of the Transport Safety Investigation Bureau, a department of the Transport Ministry, said that there was supposed to be a monitoring system that was designed to specifically watch out for the sort of overheating axle box situation that caused this.
But what use is gizmos when they’re not being put to good use, or even kept in working order? While this “hotbox detection system” had problems with detecting the identity of a train with issues passing by the sensor location, a simple sanity check could have been done against the signalling system, which already has to know where all the trains are, instead of relying on the RFID tag of the hotbox detector system alone.
It would be better than nothing, I’d think, as someone could have been activated to check the train quicker. Additional sensors can be implemented to increase the system coverage and provide second opinions; but does that solve the fundamental problem, that such a system is standalone?
The government is doubling down on “preventive maintenance”. A $1 billion investment will go to develop further such preventive maintenance capabilities. But that goes on to the next issue raised by the TSIB. Alarms fatigue — that operational staff were desensitised to a relatively unreliable system constantly crying wolf, and grew to ignore the alarms raised, is a real concern.
Alarms fatigue teaches us that just buying more equipment alone will not work. The more things get added to controllers’ and operators’ workloads, the more likely they are to overlook small things like this and to make mistakes. In fact, I’d go even further and say that while previous Transport Minister Chee Hong Tat claimed that 2,000 new jobs will be created in the rail sector, it’s important to create the right jobs.
Instead of adding yet more equipment and more complexity, starting new projects with more people involved in meetings and more equipment to take care of, what this $1 billion should be spent on is system integration. Instead of building new gizmos, pay Alstom, Siemens, Hitachi, and ST Engineering (as train makers, and communications and train control systems OEMs) to help integrate data from the current on-board systems to enhance the decision support systems used by both SMRT and SBS Transit.
In fact, the OEMs already sell such services. Apart from Alstom’s HealthHub, the LTA is already working with Siemens on such solutions, perhaps it has something to do with Siemens’ own Railigent X technology. Expanding REAMS to other rail lines and integrating all these additional technologies into the system should be of paramount importance.
Maintaining simplicity and resisting the urge to over-engineer may also be a lesson for the LTA. How many times has a backup system failed when the primary system still remained operational? While for one we may never know how many “track circuit failures” in Singapore may have been non-issues with a primary CBTC system still working, we have an example.
Hong Kong’s MTR crashed trains in 2019 when they were testing their new signalling system — an incident blamed on an improperly programmed handover between a simulated total failure of the train control system and its own internal redundancy, with a third level of backup supposed to kick in.
If this is what over-maintenance looks like, when systems that are supposed to aid maintenance and reliability efforts end up causing more problems themselves, or when they don’t even work properly and there’s no point in having them, then it does us a disservice to have it.
A system that doesn’t work
Then what about the response?
It’s not surprising that people make a mockery when SMRT is slapped with *only* a $3 million fine — a fine that the LTA claims will go to the Public Transport Fund to help needy families (read: the yearly public transport voucher exercise). While the LTA claims SMRT has already been sufficiently punished by the $10 million cost of running bridging buses and providing free trips for those affectedd, in addition to any costs paid for the track repair, is that enough?
And yet, public opinion in Singapore must be contrasted with the following headlines in KL and Sydney, cities all familiar to Singaporeans:
We need to remember that in Singapore, people are forced to use public transport. These moves, while noble, give the perception that the system is not working for the average Singaporean. Because incidents like these have their costs, not only in lost hours as more people must also spend time travelling even if they get a free ride, but the costs of surge-pricing private hire as well, if they can’t afford to spend the extra travelling time.
In other words, what this feels like, is like the government forcibly taking a $10 charitable deduction from every bank account. In contrast to the response taken by other agencies and authorities, such moves in Singapore risk underlining the tone-deafness that some have fairly or unfairly associated with the government.
This only worsens the LTA’s existing PR woes, that they’re not doing things to make the average citizen’s life better. Or at least, they aren’t exactly holding SMRT accountable — an observation backed up by the TSIB, which noted in its report that the LTA had a duty to be informed over maintenance decisions made by SMRT — and that SMRT couldn’t even cough up paperwork when audited.
It’s not hard to walk away with the notion that public trust has been squandered here. This policy should be reassessed, as part of rebuilding public trust. Penalties assessed should demonstrate that not only has a harm been done, there must also be appropriate restitution.
Other ways to do something
Of course, we must then remember how we got here in the first place. Delivery woes with new trains forced SMRT to extend overhaul intervals for its ageing fleet, past the 500,000km threshold, all the way out to 750,000km.
While this in hindsight turned out to be a bad idea, did they have much of a choice? Only from October 2024 onward were the LTA able to speed up deliveries of new trains, with a total of 70 trains of 106 delivered so far according to a Zaobao report. 16 entered service since the start of this year and 12 more are under testing. So two years after the first launch of the R151 trains, we now have 58 in service.
Conversely, according to SGTrains Spotters, only 16 old KHI trains remain in active service. 74 have been sent to the scrapyard and another 16 aren’t in active use. That leaves us short of 32 trains, hardly enough to maintain today’s service levels. Furthermore, some newer generation trainsets like 661/662 and 703/704 have not been in service for years; and another trainset 649/650 was sent back to CRRC Qingdao Sifang in China for upgrading. Some more trains may also need to be taken out of use to support testing and commissioning activities for the East Coast Integrated Depot.
As it happens, there’s a win-win solution, to take the load off the legacy MRT lines so there’s breathing room to support these projects, and also look like help is being given to manage the rising cost of living.
The LTA should begin offering Travel Smart Journeys to get people to use the TEL and perhaps the DTL as well. The TEL has so much excess capacity that SMRT has seen fit to sideline 28 of the older trains that have been running since TEL1 opened in 2020. Same goes for the DTL, where about 20-odd trains have been out of service, some since 2023. Of course, some of these may be to support East Coast Depot works as well, but do they really need that many trains?
Such initiatives can perhaps also be extended to CCL6 come 2026; where passengers coming from the west can exit at Prince Edward Road to make better use of CCL capacity.
The goals of building new MRT lines was always to take the load off parts of the existing network. Those goals need to be met, to improve the integration of our network overall.
In conclusion, these findings are damning, and I call on our new Transport Minister, Mr Jeffrey Siow, to do his best to rebuild trust in the railways. It’s clear what needs to be done; all that is left is to do it and get the railways back on track.
Like what you read? Join the Telegram Channel for updates, or follow me on Instagram, Threads, or Bluesky for quick takes!
