This site tells the bigger story of a troubled rail network through data visualization. It is inspired by a growing trend of data journalism; to distil quality information of public interest from raw data. By presenting the data in a usable and visually engaging format, we hope that the right questions will be asked, so that we can take the first step towards greater transparency and accountability.
While massive delays and outright downtime hog the limelight with their own courtroom drama, smaller delays also add up and frustrate commuters all the same. This site shows you the ugly side of the long tail, by tracking every train service disruption that is known in the public domain.
While there's no such thing as total happiness in a public transport system, there is always room for improvement. The authorities should pay earnest attention to the issues at hand, instead of chiding the public for nitpicking on 'unavoidable' disruptions and making apple-to-orange comparisons with other metros abroad. Even more frustrating, is the tendency to belittle the impact to commuters by spinning the data in their perspective: commuters do not care that a 45-minute disruption accounts for only 0.02 percent of the 2.7 million trips made daily.
Putting on an air of confidence with self-congratulatory reports does not do anyone a favour. A few minutes of delay may seem trivial, but every second counts in the context of train headways. If not, why would the service operators be struggling to cut train headways to 2 minutes during peak hours? In fact, train headway is a key indicator under the Operating Performance Standards set by LTA. Every train that missed its schedule extends waiting time, resulting in the build-up of crowds on the platform - that we do care.
The site is not a real-time news alert for train service disruptions in Singapore. You can get first-hand information from SMRT’s twitter feed or SBS Transit’s SMS Alert service.
In the aftermath of two unprecedented train service disruptions in December 2011 that affected more than 200,000 commuters, Prime Minister Lee Hsien Loong ordered an inquiry to get to the bottom of the matter. A committee of inquiry (COI) was convened, comprising of three members: Chief District Judge Tan Siong Thye; Director of Prisons Soh Wai Wah and Professor Lim Mong King from the School of Aerospace and Mechanical Engineering at Nanyang Technological University. The inquiry heard the evidence of 116 witnesses over a period of 6 weeks from 16 April to 25 May 2012.
In a strongly worded 358-page report, the COI concluded that both the 15 and 17 December 2011 incidents were preventable. Lapses in the rail maintenance regime were found despite claims to the contrary by the service operator SMRT. No words were minced when the committee said that there "appeared to be a gaping disconnect between what was formally on record and what was happening on the ground … it was particularly difficult to reconcile the seemingly robust maintenance regime with the failure of a significant number of items.”
In one of the most glaring lapses, despite a sterling profitability record and repeated assurance from SMRT’s ex-CEO Saw Phaik Hwa on maintenance concerns, SMRT had only one Multi-Function Vehicle (MFV) at their disposal to check for rail defects in the North-South and East-West Lines (a combined rail length of 90 km). Furthermore, outdated software on the MFV (in use since the early '90s) meant that it could not get reliable readings and a cracked mirror on the day of the incident forced a manual check along 900 metres of track in 20 minutes on foot.
While SMRT’s maintenance regime had its shortcomings, Transport Minister Lui Tuck Yew acknowledged that both the Ministry of Transport (MOT), as the supervising ministry, and the Land Transport Authority (LTA), as the regulator - had to shoulder their share of responsibility as well.
SMRT was fined S$2 million for the December 2011 train service disruptions.
The COI was billed at a S$10 million dollar price tag, to be borne by LTA and SMRT.
SMRT is investing in a S$2-million integrated incident management centre for both train and bus services. The centre will be ready by the end of 2013 and it will allow management to monitor train movements and station crowds via a multi-screen system.
Data is collected from online news and the websites of the two rail operators, SMRT and SBS Transit, as well as SMRT’s Twitter feed. Online sources are preferred because every data point can be linked for citation. The data is stored in a public Google Spreadsheet. If you wish to use the dataset, please link back to this site for acknowledgement.
SMRT does not tweet all train service withdrawals, only those that have material impact on commuters. Hence, while complete information on train operating performance is not publicly available, this site's data is a valid subset of the records held tightly to the chest by the service operators. Still, SMRT deserves credit for giving better customer service through transparency. In contrast, SBS Transit does not yet have a Twitter account, and its SMS alert service does not report minor delays -- at times it even has to piggyback on SMRT's tweets to disseminate important information on developing situations. Consequently, SMRT received a lot of flak from angry commuters concerning service disruptions not of their own making.
Could this be the longest delay-free streak since SMRT started their Twitter account in December 2011? Nada, there wasn't even a single tweet to warn the public about the extended delay on the North-South Line on 7 December 2012. Instead, SMRT is tweeting diligently on yet more lift faults, tourism advice and commercial messages. Hopefully, this is not a case of data throttling. SMRT should not renege on their promise to deliver prompt information on delays.
The following events are excluded from the dataset, because they are beyond the control of the service operators:
After SMRT revamped its official website in early 2012, they have removed all press releases dating before 2009. This site's archive has preserved the old docs in Google Drive for posterity.
Check out the code on the Github repo. This site is built on
This site is a work in progress and will be expanded over time. The database is certainly not complete and we will need your help to contribute any data that we have missed out or had reported inaccurately. Feedback and comments are most welcome, please send them to datadope (at) gmail (dot) com