Mr Speaker, world class infrastructure is not experienced in masterplans or at ribbon-cuttings. For the vast majority of Singaporeans, it is experienced at the bus stop downstairs, on the platform at the MRT station, in the daily journey to work and back home. It could be excellent service and smooth journeys, as articulated in the LTA’s service charter. Or it can be poor service and frustrating journeys if things don’t work well.
Whither our public transport ridership goals?
To begin with, the LTA aims to have 75% of all peak hour journeys carried out via public transport by 2030. Ten years ago, in 2016, the figure stood at 62%. As of FY2024, it is a mere 66%, and with just four years to go, I wonder if this target will soon be quietly revised. I believe we should instead be ambitious and aim to meet and exceed it. To do so, our public transport network must be as convenient as, if not more convenient than, private road transport.
To that end, my speech proposes three thrusts. First, reforming the Bus Contracting Model as it approaches its ten year anniversary, to raise service standards rather than merely minimise costs. Second, accelerating the development of our rail network through a renewed rail financing framework and a single, publicly-owned national rail operator. Third, accelerating the rollout of autonomous vehicles, not as a threat to employment, but as a catalyst for anchoring good jobs in Singapore.
Mr Speaker, I am sure many of us in this House have received feedback from constituents regarding bus connectivity in their estate, whether it be increasing bus frequencies or introducing a new bus route altogether.
This also comes at a time of increased public transportation demand, with new BTO projects springing up islandwide and the densification of existing HDB estates. Commuters feel this during their daily commutes, from squeezing with other commuters on the train early in the morning to missing their bus home as it is full.
Ramping up public transportation capacity is no easy feat. Schedules need to be reshuffled, while additional buses and manpower will need to be deployed. In particular, new blood needs to be pumped into the industry. From 2021 to 2025, the proportion of local bus captains dropped from 54% to 41%, with 2 local bus captains leaving the industry for every new local hire.
10 years on: Bus contracting model 2.0
Allow me to first touch on the Bus Contracting Model. Ramping up public transportation capacity presumably comes at a cost, and as the Government often reiterates, our public transport services are heavily subsidised by the Government, with former Transport Minister S Iswaran stating in 2022 that this stands at more than 2 billion dollars per year, and this includes a subsidy of around 1 billion for trains and another 1 billion for buses.
The Bus Contracting Model (BCM) introduced in 2016 splits up our bus network into packages, which Public Transport Operators eventually bid to operate. Aside from service quality, the LTA looks at the bid price set by the operator, which is also the service fee payable to the operator by the LTA. The unique feature here is that the BCM shifted the industry from an asset-heavy framework to an asset-light contracting model, and that the Government assumed full fare-revenue risk while operators can earn a cost-plus margin for their services.
The BCM is a big step forward from the completely privately-run bus network of yesteryear. Subjecting profit-motivated public transport operators to a competitive tender incentivises them to keep operating costs low, which might be an issue if it adversely impacts service standards. However, is there room for us to require even higher standards on the PTOs, for the benefit of our commuters and broader public transport ridership goals? Ten years on from the launch of the BCM, I think it is time for us to undertake a thorough review to evaluate if the model continues to deliver good value for commuters. I have three concrete asks of the Minister when it comes to the BCM.
Firstly, bus services are scheduled at headways of 15 minutes or less during the morning and evening peak periods, compared to the pre-BCM standard of 30 minutes or less. From an extremely low bar of 30 minutes, we have made significant progress but 15 minutes is still a long time, especially if the commuter ends up not being able to board the first bus that arrives. While the Government may say that we have at least half of the bus services having shorter scheduled headways of no more than 10 minutes, I believe this should not be a ceiling but a floor. Can we aim for a shorter scheduled headway of 10 minutes or less, period? Moreover, in a bid to spread demand out, commuters are being encouraged to travel outside of the traditional peak periods, and we ought to at least have specific schedule targets in place for off-peak periods.
On the cost side of the equation, the recent Tampines bus package tender results provide us with much reason to be optimistic. Firstly, the LTA awarded the contract to the Go-Ahead Group, the company that placed the highest bids, for $646 million instead of offering it to the incumbent operator who placed the lowest bid of $596 million. This suggests that there is significant room for improvement when it comes to bus service quality.
Secondly, and more importantly, Go-Ahead Group's proposal offered cost savings for the LTA compared to the current contract, even at the highest bid price. Considering that the incumbent operator is willing to earn $50 million less and still make a profit suggests that there are significant cost savings to be made from the current expenditure levels incurred by the Government. The latest tender award for the Serangoon-Eunos bus package on 16 July presents a similar picture, where SMRT was awarded the contract for a fee of $599.5 million, and offered cost savings for LTA compared to the current contract. Looking at the latest financials of SBS Transit, I estimate that their Return on Equity is 9% while their Return on Invested Capital (ROIC) is 18%. For an oligopolistic business, particularly where the operator takes on no fare revenue risk for buses, this appears very high.
I hope that the Government can accelerate the calling of tenders for the remaining four of fourteen bus packages that have never been put up for competitive tender but were instead directly negotiated between the operator and the LTA. As the Tampines and Serangoon-Eunos examples have shown, there is significant room for cost savings and quality improvement to be made.
Thirdly, the LTA collects Excess Wait Time, lost mileage, safety, and commuter satisfaction data by operators under existing contractual frameworks. Publishing this annually, as Transport for London does quarterly, would cost nothing additional and would allow Parliament and the public to assess the performance of operators and whether performance extensions are justified.
Most importantly, I hope that a thorough review can be conducted by the LTA, to evaluate if the model continues to deliver good value for commuters and if we can raise the bar for better public transport outcomes.
A renewed rail financing framework: one network, one operator
Mr Speaker, if buses are the capillaries of our public transport network, then rail is its backbone. Our MRT and LRT network carries some 3.7 million rides daily, and the expansion pipeline is substantial.
Former Transport Minister Lui Tuck Yew announced all the way back in the Land Transport Master Plan 2013, that by 2030, eight in ten households will be within a ten-minute walk of a train station. This was again highlighted in a gov.sg Facebook post in 2020, stating “Do you know that 8 in 10 households will be within a 10 minutes’ walk of a train station by 2030?
Recent commentary however seem to have shifted to the 2030s with an “s”, and it gives the impression that the target has been quietly tweaked. Regardless of whether the Government intends for this target to be reached in 2030 or 2039, if we are serious about 75% mode share by 2030 and beyond, rail is where the heavy lifting must happen.
But even as we pour tens of billions into rail infrastructure, the framework governing how our rail lines are financed and operated has become a patchwork. Allow me to briefly trace its history. The New Rail Financing Framework, or NRFF, was announced under the 2008 Land Transport Master Plan, with the Downtown Line the first to come under it in 2011. Under NRFF version one, the operator collected fare revenue and paid the Government a licence charge, bearing significant commercial risk. In 2016, a decade ago, the LTA bought over $1.06 billion of operating assets from SMRT and moved its lines onto NRFF version two, which introduced fare revenue shortfall sharing and an EBIT cap and collar that effectively caps the operator's margin at around 5%. SBS Transit's North East Line and LRT lines followed in 2018, and the Downtown Line only completed its transition to NRFF version two in January 2022.
Meanwhile, the Thomson-East Coast Line was awarded to SMRT in 2017 via a limited tender on an entirely different basis: a service fee model where the Government collects all fare revenue, bears all fare revenue risk, and pays the operator a fee of about $1.7 billion over nine years. The Jurong Region Line, awarded in November 2024 to a joint venture between SBS Transit Rail and RATP Dev for approximately $750 million over up to eleven years, is on the same service fee model. The Cross Island Line and future lines such as the Seletar Line will in all likelihood follow suit, with the LTA bearing all development expenditure and fare revenue risk.
So today we have one increasingly intertwined network operating under multiple concurrent financing frameworks. As more interchanges open and more journeys span multiple lines and operators, ever more effort is expended on apportioning fare revenue between operators and reconciling different frameworks. I believe that our rail network should be planned, financed and viewed holistically.
It has been a decade since SMRT was privatised, and NRFF version two was rolled out in 2016. I therefore call for a renewed rail financing framework: to complete the convergence of all rail lines onto a single, unified service-fee framework, and go one step further by consolidating rail operations under a single, publicly-owned national rail operator. This draws on the Workers' Party's proposal in our manifesto for a National Transport Corporation: a publicly-owned, non-profit entity that reinvests profits for commuters' benefit, ensures fare sustainability, optimises infrastructure revenue, attracts top talent, achieves economies of scale, and enables the LTA to focus on its regulatory role.
Now, Mr Speaker, the Minister responding to my speech may well point out an apparent contradiction: on one hand, I call for more competitive tendering of bus packages, and on the other hand, I call for a single rail operator. The answer is that the right market structure follows the nature of the asset. Buses are asset-light for the operator, the Government owns some fleet and depot assets and they can be parcelled out into distinct geographic regions; there are many capable bus operators globally; contracts turn over every five to seven years; and critically as the recent Tampines bus tender showed, contestability delivers real value for money.
Rail is the opposite: massive, lumpy capital expenditure over decades, deep system-specific engineering knowledge, and in practice only two incumbent operators. The TEL itself was awarded by limited tender, not open competition. Even for the JRL, foreign operators can only participate as minority partners of SMRT or SBS Transit. Contestability in rail is thus limited to a duopoly by design, so we should instead capture the economies of scale that a single operator brings.
Looking at foreign jurisdictions with larger rail networks than us is instructive. Transport for London, the very jurisdiction whose quality incentive contracts inspired our own Bus Contracting Model, operates the London Underground directly through a wholly-owned subsidiary while franchising its buses to private operators through competitive tender. Rail in-house, buses by tender: precisely the structure I am proposing. In Hong Kong, the MTR Corporation operates the entire heavy rail network as a single integrated operator with on-time performance of 99%. Taipei's metro is likewise run by a single city-owned corporation. Across the jurisdictions we most often benchmark ourselves against, the single integrated rail operator is the norm, not the exception.
Indeed, Mr Speaker, we are already halfway there in practice, just not in name. The Rail Reliability Taskforce formed in September 2025 is chaired by the LTA Chief Executive and comprises the Group CEOs of both SMRT and SBS Transit, precisely because rail reliability cannot be managed operator by operator. Its recommendations, which the Ministry accepted in February this year, call for accelerated renewal of trains, signalling and power systems, better management of spare parts, and more comprehensive, standardised condition‑monitoring across the rail network. The Singapore Rail Academy is also being positioned as a sector‑wide platform to build and deepen workforce competency in rail.
As with buses, I therefore have three concrete asks of the Minister when it comes to the NRFF. First, commit to a renewed rail financing framework that converges all existing lines onto a single, unified service-fee model under one licence, ahead of the Cross Island Line's opening in 2030 and upon the gradual expiry of existing licenses. Second, study the consolidation of rail operations into a single publicly-owned national rail operator, drawing on best practices from London, Hong Kong and Taipei. Third, in the interim, publish regular updates on progress made in implementing the recommendations of the rail reliability task force. Particular attention should be paid on the costs of the current fragmented structure, and the potential synergies from network-wide harmonised data standards, pooled management of spares and supply chains, and coordinated renewal of power, signalling and trains across the network.
Accelerating autonomous vehicle rollout
Mr Speaker, let me now turn our eyes to the future, and that is autonomous vehicles or AV technology.
Frontier technologies will be critical to strengthening our economic competitiveness and anchoring good jobs. To quote the MOT, AVs will be the next game changer. My concern is that we appear to be more like an observer than a player in this rapidly evolving game, and risk being left behind in the global push for AVs at a time when physical AI is being framed as the next wave after generative AI. As Jensen Huang, CEO of Nvidia shared in January this year, "The ChatGPT moment for physical AI is almost here", and that "Our vision is that someday, every single car, every single truck, will be autonomous".
As I have shared in my MOT COS speech earlier this year, Singapore has been involved in research and small-scale testing of AVs since at least 2013. Yet, over a decade later, our AV rollout remains largely in the trial phase, with only around 20 AVs involved in short, fixed route services in Punggol. We also have six autonomous buses procured at a contract value of $8.14 million, due to start operating on Bus Services 400 and 191 in the second half of 2026. These are 16-passenger seat vehicles in a pilot that will run for an initial three years.
From a regulatory standpoint, sections 6C, D, and E of the Road Traffic Act provide a regulatory sandbox for AV trials here, with the MOT's public consultation regarding an upcoming piece of legislation specific to the deployment of AV systems having just wrapped up in June. The Government's stated AV ambition, when pressed, was that Singapore would have "many autonomous vehicles" in five years. Notably, the LTA adopts a more cautious approach when implementing such technology locally, choosing to be more "calibrated and data-driven". Now, that is not a negative per se but let us set that against the global reality.
The AV industry in places like China and the US has boomed, with their biggest players, Pony AI, WeRide and Waymo already running fully driverless commercial services in the cities they are operating in. Amidst the growth this industry is facing, the UN has also introduced a regulatory framework for AVs, bridging the gap between the fragmented patchwork of regulations across different countries.
Countries in the Middle East are also showing markedly faster implementation of AVs. The Dubai Autonomous Transportation Strategy aims to have 25% of the total transportation in Dubai to be autonomous by 2030. Saudi Arabia's Transport General Authority has set a target for 25% of all goods transport vehicles to be fully autonomous by 2030.
While Singapore deliberates and conducts protracted trials, these global players and global cities are already operating at scale, creating a new ecosystem of jobs in AV research, operations, and maintenance. Even as we have close to 100,000 Private Hire Cars, there remains a shortage of ride-hail drivers especially during the peak periods and the wee hours of the night. On robotaxis, as Transport Minister Jeffrey Siow has said, even if we bring all the AVs in the world to Singapore, that is less than 10% of our taxi and private hire car population. Having AVs deployed on the specific routes and hours where the driver gap actually bites would materially help. And the severe shortage of local bus captains presents a clear opportunity for autonomous buses to alleviate manpower pressures and satisfy rapidly growing transport demand.
A 2025 survey by the National Transport Workers' Union found that one in three public transport workers expressed concern that autonomous vehicles would impact their jobs. That concern is legitimate and deserves to be taken seriously. However, even as the demand for public bus captains has grown, as of December 2024 only 28% are Singaporeans, with 57% foreign work pass holders required to supplement the driver pool. Whether it's school buses or tour buses, operators consistently cite the challenge of driver shortages. In fact, 42 school bus operators have been allowed to recruit more foreign drivers than under the standard provision to alleviate an acute shortage of drivers. The shortage of local bus captains and the need to expand our public transport network is a clear and urgent impetus for us to accelerate AV deployment. This is also echoed by PAP MP Mr Ang Wei Neng, who said at the debate on the Transport Motion that we need to move with a much greater sense of urgency.
Moreover, to truly anchor good jobs, we must accelerate our AV rollout with clear, ambitious numerical targets. This is not about replacing human jobs; it is about transforming the nature of work. We need to create higher-skilled roles in the AV ecosystem, from AI engineering to remote fleet management, and ensure Singaporeans are trained to take up these roles.
Critically, AVs should be used to grow and plug gaps in our public transport network, connecting underserved neighbourhoods with major transport nodes. This would be especially helpful for those with mobility issues, such as the elderly and persons with disabilities. With the pilot autonomous public buses equipped with ramps for barrier-free access, I hope that the AV roll-out would be carried out with the needs of these communities in mind.
The Minister for Transport shared during the debate on the Transport Motion that “MOT plans to invest $800 million over the next five years on transport research and innovation to support efforts to develop frontier technology”. All that is fine and good, but once again, I reiterate the question I have posed for the Minister of Transport, in that beyond stating that Singapore will have, and I quote "many autonomous vehicles in Singapore" in five years' time, can the MOT provide numerical targets for our AV ambition, as well as a concrete and just path towards our engagement with frontier technologies?
Conclusion
To conclude Mr Speaker, our public transport mode share has crawled from 62% to 66% in eight years, against a 75% target that is now four years away. The three thrusts I have proposed today, a Bus Contracting Model 2.0, a unified rail financing framework and an accelerated AV rollout with numerical targets, share a common thread: they ask the Government to organise our land transport system around outcomes for commuters, and to move quickly, rather than navigate around the increasingly complex but administratively convenient structures in existence today.
