Londoners depend on Tube drivers every day. (Image: Getty)
On January 10, 1863, the London Underground opened its first line between Paddington and Farringdon, marking the start of what has now firmly established itself as an indisposable part of the UK’s capital city. The following decades saw a litany of developments and advances, as the metropolis expanded and Londoners and visitors took to the subterranean system. Electric traction was rolled out in 1890, and the formation of the Underground Electric Railways Company of London in 1902 led to yet more expansion. The first trains were powered by steam, producing copious amounts of soot and smoke.
The last of these classic models was withdrawn from service in 1971, and the Tube was henceforth fully electric. Both generations of locomotives needed drivers, upon whom millions of people depend every single day. But the future of the world’s most iconic subway may look a lot different, as it has been suggested that the network could see fully automatic trains, in a move that would ditch the man or woman at the controls. Reform UK’s only London Assembly Member, Alex Wilson, told the Express that his party would, in the long-term, work towards “progressive automation of Tube lines”.
The Victoria, Jubilee, Central, Northern, District, Circle, Hammersmith and City and Metropolitan lines are already semi-automated, but still require a human operator for essential functions, such as opening and closing the doors and assisting in emergencies.
In order to make them driverless, it was estimated in 2021 that Transport for London (TfL) would have to fork out at least £10billion, which means the figure would most likely be much higher today.
The first London Underground line was opened in 1863. (Image: Getty)
Indeed, Sir Sadiq Khan, who is the Chair of TfL, said in December that, including “optimism bias”, the Department for Transport (DfT) believes that the total would be nearly £20billion for just three sections of the Tube: £4.9billion would be required to convert the Piccadilly line, £4.4billion to alter the Bakerloo line and around £10billion to adapt the Central line.
The Mayor of London added: “The most practicable conversion to driverless operation would be to coincide with the introduction of new rolling stock, signalling and platform edge doors as part of a line upgrade.
“The entirety of this work would cost billions of pounds on each line.”
One of the conditions of TfL’s 2021 funding agreement with the Government was to examine the feasibility of producing a viable business case for implementing driverless trains on the Waterloo and City and Piccadilly lines.
A 2023 feasibility report from the DfT stated that there are “no inherent technical impediments to the conversion of most lines” to being fully driverless, although “unique features” of the London Underground would need to be considered.
It also described the economic case for doing so as strong.
Introducing the necessary technology, including new platform edge doors (PEDs), extra cameras, potentially deploying AI to help safeguard the public; new train signalling and reengineered stations, would require significant Government investment, in addition to existing capital requirements, and would take many years to implement, it is thought.
This is not to mention depot changes that would need to be finalised, such as train control systems and ensuring staff are protected from automatic train movements if they are working trackside.
The 2023 DfT report noted that the installation of PEDs would “improve safety and ambience”.
It also mentioned that “more robust fortification” would be needed to ensure the segregation of customers and the automated network.
Sadiq Khan has said that making three Tube lines fully automatic would cost nearly £20billion. (Image: Getty)
It seems that transport authority officials would currently rather focus on implementing other upgrades across its network, including greater automation where there is a clear customer and reliability benefit, such as through the four lines modernisation programme on the Circle, District, Hammersmith and City and Metropolitan lines.
The Express understands that Sir Sadiq is not currently pursuing driverless trains due to the cost in terms of new trains and technology.
A spokesperson for TfL said simply: “There are currently no plans to introduce driverless trains on London Underground.”
However, Professor Sabih Khisaf, an Institute of Civil Engineers (ICE) fellow and infrastructure lead at Hyperloop Transportation Technologies told the Express that making the Tube fully automated is not only feasible but already a reality in many cities worldwide, with metro systems in Dubai, Doha, Vancouver, Singapore, Copenhagen and Paris having driverless trains.
The French capital is putting London to shame on the automation front, as it has three fully driverless metro lines – 1, the first to be converted in 2012; 4, adapted in 2024; and 14, driverless since it was finished in 1998 – with 13 currently being converted.
This work is expected to be complete in 2035.
New lines 15, 16, 17 and 18 are also planned, which would run automatically as part of the around £34billion Grand Paris Express project.
Dubai’s rail system is completely driverless and has two sections, the Red and Green Lines, featuring air-conditioned, Wi-Fi-enabled stations and trains.
Denmark’s fully automated network has been in operation for over two decades, with the first lines opening in 2002.
Professor Khisaf added: “However, for older and more complex systems, the transition requires significant investment and negotiation with stakeholders. While not commonly practical in the immediate future, the long-term trend suggests that automation will become the standard for urban underground transport systems.”
The cost of a fully automated driverless metro system is large and depends hugely on local conditions, the expert added, as well as design choices and how much of the work is underground.
Professor Dr Sabih Khisaf (Image: ICE)
For a modern new line that is between 20 and 30 kilometres long in a developed city, costs are likely to be in the range of around £74million to £148million per kilometre, possibly more, Professor Khisaf said.
He added: “A fully automated driverless metro system is more than just trains without
drivers, it requires a holistic rethinking of urban transit using cutting edge technology.
“From Communication-Based Train Control signalling and AI traffic management to predictive maintenance and cybersecurity, each component must work seamlessly together.
“The ultimate goal is to create a transport network that is safe, efficient, resilient, and capable of adapting to the dynamic needs of a modern city.
“While many of these technologies already exist in early forms, their full integration into a large-scale underground system remains the next frontier of urban mobility.”
Professor Khisaf also highlighted that a new mode of automated transport system for passenger and freight is being developed, known as the Hyperloop System.
The high-speed transportation system involves specialised pods being accelerated through a low-pressure tube to achieve speeds of up to 745mph.
Another expert, Dr. Piers Connor, calls the prospect of driverless Tube trains a “political myth”.
He wrote in 2021: “Operating trains is complex. It involves a vast range of micro-calculations and micro-decisions on everything from the speed of approach or departure, to when doors should open, to how trains on the same (and sometimes other) lines should flow around each other.
“Both human beings and computers are capable of doing all these things. Humans are, ultimately, just natural computers. But the more complex the decision-making required on a particular transport system, the more knowledge and calculations are required for it to be done within acceptable safety parameters.”
He added: “What this means is that the more complex the transport system, the more learning required and thus the less suited it is to automatic operation.
“Because, at least until the invention of ‘true’ Artificial Intelligence, learning is an area in which ‘wetware’ (that is, humans) remains superior to software. Wetware learns things faster and is quicker to adapt when things go wrong. And the more complex – and the older – your transport system is, the more potential there is for things to go wrong.”
Dr Connor told the Express: “The annual interest on the billions of pounds TfL would need to borrow to put driverless trains on the Underground would cost more than paying for drivers.”
Customers want a driver or pilot in the hot seat when the technology goes wrong, Mr Whelan said. (Image: Getty)
He added: “Eliminating drivers wouldn’t stop strikes by other staff on stations, in maintenance depots and control rooms who could quickly stop the system, like Docklands.”
In September, a strike on the Underground brought the capital to a halt, as RMT members, drivers as well as station staff, walked out in a dispute over pay and conditions.
On Monday, the RMT announced that it had received an improved pay offer of a three-year deal, consisting of a 3.4% uplift in year one and a straight RPI increase in year two and three, from London Underground.
The union’s general secretary Eddie Dempsey said: “This offer will now be discussed by the national executive committee and in consultation with the membership.
“As a result, the union remains in dispute with London Underground subject to the will of the membership under our democratic structures.”
Mr Dempsey told the Express: “Reform claims that they would bring in driverless tube trains are not serious as it would require a combined upgrade of rolling stock, signalling, and platform infrastructure across the entire network.”
Mick Whelan, the general secretary of ASLEF, said: “Driverless trains are simply not viable on the infrastructure of Britain’s railway network – a testament to the imagination and skill of our Victorian engineers, but now very old.”
He added that even the automated Docklands Light Railway (DLR) is staffed with a train captain who can drive the train when needed.
Mr Wheelan added: “Planes are fly-by-wire, but all the evidence shows that passengers – on trains as well as planes – want a driver or pilot in the hot seat when the technology goes wrong, as it always, inevitably, does.”
It is true that the threat of a strike would still loom over the Tube, as humans would still be involved at stations to some degree.
This is unless some kind of robot is developed that can do the job of station staff.
The Bombshell Plan to Ditch London Underground Drivers – and Sadiq Khan’s £20bn Warning
The idea of removing drivers from London’s Underground trains has returned to the political debate, reigniting an argument over technology, transport costs, industrial relations and the future of one of the world’s largest metro systems.
For years, driverless Tube trains have been discussed as a possible way of modernising London’s transport network. Supporters argue that automation could improve reliability, reduce the impact of industrial disputes and potentially allow trains to operate more efficiently. Critics, however, point to the enormous cost of converting existing infrastructure and question whether removing drivers would deliver enough practical benefits to justify the investment.
The debate has particular significance because London Underground already operates alongside a fully automated railway. The Docklands Light Railway has operated without conventional train drivers since its early years, demonstrating that automated rail transport is technically possible in London.
But transferring that model to the Underground would be considerably more complicated.
The Tube network is much larger and includes infrastructure that has developed over more than a century. Different lines use different trains, signalling systems and operational arrangements. Introducing automation across the network would therefore require more than simply removing the driver from the front of an existing train.
Transport for London has previously examined the possibility of introducing driverless trains on parts of the network. During the pandemic, TfL was required as part of a government funding settlement to investigate the possibility, including the Piccadilly and Waterloo & City lines.
However, that work was eventually abandoned.
In December 2024, Mayor Sadiq Khan said the costs involved would be extremely high. He stated that introducing driverless trains would cost billions of pounds on each line, with major expenditure required for new trains, signalling and platform-edge doors. TfL subsequently ended its investigation into the proposal.
That history explains why the latest debate has attracted so much attention.
The financial question is central.
One recent argument has put the potential cost of full automation across London’s Underground at more than £20 billion. That figure has been cited in commentary about the possibility of replacing conventional train operation with fully automated systems.
Such an investment would be enormous even by the standards of London’s transport system.
Supporters of automation argue that the initial cost should be considered alongside the long-term benefits. Driverless technology could potentially reduce staffing requirements, improve operational flexibility and reduce the effect of strikes involving train operators.
Industrial action has repeatedly disrupted London Underground services.
In 2026, the RMT announced strike action over proposed changes to the working arrangements of Tube drivers. The dispute centred on a proposed compressed four-day working week, with the union raising concerns about shift lengths, fatigue, safety and working conditions.
TfL, however, presented its proposal differently.
The transport authority said the four-day working pattern would be voluntary, would not reduce contractual hours and could improve reliability and flexibility without additional cost. TfL said the proposal was designed to bring London Underground closer to working arrangements used elsewhere in the rail industry.
The disagreement illustrates the complexity of London’s transport debate.
Automation is sometimes presented as a solution to strikes because a driverless train does not depend on a human operator sitting in the cab. But removing drivers would not necessarily remove the need for staff.
Passenger safety, station management, maintenance, security and emergency response would still require human workers.
The Docklands Light Railway provides an example. Although trains operate without conventional drivers, staff remain part of the system and can assist passengers and respond to operational problems.
This is an important distinction.
“Driverless” does not necessarily mean “staffless”.
A fully automated Underground could still require substantial numbers of employees, although their responsibilities could change.
There would also be major technical challenges.
A driverless system needs sophisticated signalling and control technology capable of managing trains safely under a wide range of conditions. Stations may require platform-edge doors or other safety measures to prevent passengers from falling onto the tracks.
Existing trains may not be suitable for full automation.
Consequently, replacing drivers could require replacing rolling stock, upgrading signalling and rebuilding parts of stations.
This is precisely why Khan previously described the challenge as requiring billions of pounds of investment. In 2020, he said that automated trains would require new trains, major station changes, new signalling and changes to other parts of the infrastructure.
The debate therefore cannot simply be reduced to the question of whether technology exists.
The technology exists.
The real question is whether it makes financial and operational sense to transform London’s existing Underground network.
Supporters argue that London must think beyond the immediate cost.
A major investment could potentially create benefits over decades. More automated operations might allow trains to be controlled more precisely and could improve consistency in certain circumstances.
There is also the issue of industrial relations.
Every major strike can cause significant disruption to commuters and businesses. London’s economy depends heavily on public transport, and Underground disruption can affect workers, retailers, tourists and businesses across the capital.
Some supporters of automation therefore argue that reducing dependence on drivers could provide greater resilience.
Critics respond that automation itself carries risks.
Large-scale technological systems can fail, and when they do, the consequences can be significant. Human staff remain important when unusual situations occur, including passenger emergencies, equipment failures and evacuations.
There is also the question of employment.
London Underground drivers represent a skilled workforce with established contracts, pay arrangements and union representation. A transition toward automated trains would inevitably raise questions about jobs, retraining and the future structure of the workforce.
Those questions could make any attempt to introduce automation politically difficult.
The history of London’s transport system also demonstrates how expensive infrastructure projects can become.
A £20 billion transformation would compete for funding with other priorities, including new trains, station improvements, accessibility projects, maintenance and expansion of London’s transport network.
That creates an opportunity-cost question.
Even if driverless technology could eventually provide benefits, policymakers would need to decide whether the same money could achieve greater improvements elsewhere.
For example, investment could potentially be directed toward upgrading ageing infrastructure, improving accessibility at stations or increasing capacity on overcrowded routes.
This is why the financial case for automation would have to be demonstrated in considerable detail.
The debate also has a political dimension.
Sadiq Khan has faced repeated criticism over the cost and management of London’s transport system, while his supporters argue that TfL has faced extraordinary financial pressures, particularly following the collapse in passenger numbers during the pandemic.
The mayor has previously been open to examining new transport technologies, but he has also stressed the enormous cost associated with converting the Underground to driverless operation.
His position demonstrates the difference between being open to technological innovation and being prepared to spend billions implementing it.
For passengers, the issue is much more straightforward.
They want trains that arrive reliably, stations that are safe and fares that remain manageable.
Whether the person controlling the train is sitting in a cab or supervising the system remotely is less important than whether the service works.
That means any future proposal for driverless Underground trains would need to demonstrate clear improvements for passengers.
Reliability would be particularly important.
If automation could allow trains to operate more frequently or recover from disruptions more quickly, it could potentially provide a meaningful benefit. But if the transition caused years of disruption and enormous expenditure without producing substantial improvements, passengers could reasonably question the value of the project.
There is also a question about public confidence.
Passengers are already familiar with automated rail systems such as the DLR, but the Underground is different. Many Tube journeys take place in deep tunnels, on complex networks and through extremely busy stations.
Introducing driverless trains would therefore require passengers to trust the technology and the safety systems supporting it.
That trust would not be created by political announcements alone.
It would have to come through testing, transparent safety assessments and evidence that the system can handle unexpected circumstances.
The debate over drivers is therefore likely to continue.
For some politicians and commentators, automation represents an opportunity to modernise London and reduce the disruption caused by industrial disputes.
For others, the enormous cost makes the proposal difficult to justify when London’s transport network faces numerous competing demands.
Sadiq Khan’s previous warning about the billions required for each line demonstrates just how significant the financial challenge could be. The £20 billion figure discussed in more recent commentary illustrates the scale of the transformation being contemplated.
Ultimately, the question is not simply whether London can build driverless Tube trains.
It is whether London should spend the necessary money to do so, how the transition would be managed and what passengers would receive in return.
Technology has already transformed many parts of London’s transport network.
The next stage could bring even greater automation, but moving from a few automated operations to a largely driverless Underground would represent a much more ambitious project.
For now, the idea remains surrounded by technical, financial and political questions.
The £20 billion warning provides a reminder that technological ambition comes with a substantial price tag.
And if London is ever to abandon conventional Tube drivers, the argument will ultimately have to be won not only in the political arena, but through a convincing demonstration that automation can deliver a safer, more reliable and financially sustainable Underground for millions of passengers.