Richard Brown: Robotaxis are coming and London must be prepared

The Automated Vehicles Act 2024 received royal assent on 20 May last year, but was rather overshadowed by a rain-soaked Rishi Sunak announcing, two days later, that he was calling a general election. While the Act leaves a lot of detail to secondary legislation (an explainer by Scarlett Milligan of 39 Essex Chambers  is helpful), it sets out a structure for regulating the use of self-driving cars, which could see them on Britain’s roads as early as next year, according to a Department for Transport press release.

The year 2026 is suddenly very near, and it seems unlikely that regulations will be in place by then. But there have already been London trials of autonomous vehicles – in the Queen Elizabeth Olympic Park and in Woolwich – and Waymo self-driving taxis are commonplace in several US cities, including San Francisco, where the urban environment is as almost complex as London’s. This particular future seems to be drawing closer.

You can debate whether self-driving cars (more likely to operate as “robotaxis” than as individually-owned vehicles, for the next few years at least) are a good thing. Advocates say they could reduce accidents and congestion, and offer liberating point-to-point transport for people who are currently excluded by price, location or physical conditions – a case argued in an excellent blog by James O’Malley, which set me thinking about what abundant robotaxis would mean for London.

Not everyone is so positive. Sceptics argue that robotaxis could divert people from healthier walking and cycling, lead to more sprawl, congestion and pollution, reduce social interaction, and undermine the viability of mass public transport. Transport for London’s 2019 position statement takes a fairly guarded view, pointing to the Mayor’s strategic priority – that 80 per cent of journeys should be made by walking, cycling and public transport by 2041.

Progress towards that target has been slow: around 63 per cent of journeys were made by walking, cycling and public transport in 2023, a level that has barely changed over ten years. But in any case, it seems to me that readily-available and cheap robotaxis would be very attractive to people – including those who do not like walking, cycling or using public transport, or are not able to use those ways of getting around.

So, unless the government chooses a highly restrictive approach (essentially banning autonomous vehicles), the question becomes one of what policies cities such as London should adopt in order to manage their impact on urban mobility, on the built environment and on public health.

At the moment, it costs more – in time and money – to use Waymo taxis than to use Ubers in San Francisco, but that gap is likely to narrow and reverse as the technology improves; McKinsey estimate that by 2030 in the US the cost per mile will be little more than  using your own car. If the cost fell that low or further, the challenge for public transport systems and for urban streets would be acute. People could switch to robotaxis en masse – causing gridlock and stripping London’s trains and buses of paying customers (even if gridlock would eventually choke demand).

One inevitable policy response to the arrival of autonomous vehicles is the (much discussed and much can-kicked) introduction of road user charging. As my former colleagues at Centre for London have suggested, this could be targeted so that using roads (and robotaxis) is made more expensive where journeys can be easily made by public transport, walking or cycling, and cheaper in places where those options are not available, and for people to whom they are not accessible.

Switching from car ownership to robotaxi use could also potentially free up road space. Car ownership is already declining faster than car use in London, and abundant robotaxis could accelerate this, with many privately-owned cars disappearing from kerbsides and suburban driveways, opening up new public spaces across the city.

But where would the robotaxis go, while we’re asleep in our beds? A dystopian, polluting and congesting, scenario would see empty cars cruising London roads until summoned by passengers, like electric Marie Celestes. A better option would be to place limits on empty circulation, and to provide local parking hubs with charging facilities and waiting space.

This would require some adaptation of existing car parks (which could hold many more cars if no human drivers needed to get in and out), and perhaps transport hubs being incorporated into new developments; again, Centre for London has undertaken research in this area.

From a public health perspective, you might go further, requiring or incentivising most users to walk to a transport hub to pick up a car, rather than to expect door-to-door service. This might be too nanny-ish for some tastes. But not only would it incorporate a (minimal) level of walking into most journeys, it would also discourage people from using cars for shorter, walkable or cyclable, trips.

The 20th Century city was shaped by private car ownership. If robotaxis become commonplace, their impact on the 21st Century city could be just as great. Policy action will be needed, and soon, to make that impact positive.

OnLondon.co.uk provides unique coverage of the capital’s politics, development and culture. Support the website and its writers for just £5 a month or £50 a year and get things that other people won’t. Details HERE. Follow Richard Brown on Bluesky. Image from Waymo video.