The London Cycling Campaign has urged caution concerning the prospect of driverless taxis roaming the streets of London this year, after autonomous driving tech firm Waymo announced that it hopes to be operating a ‘robotaxi’ service in the capital perhaps as soon as September.
Waymo, formerly known as the Google Self-Driving Car Project, said this week that a pilot service for the scheme will launch in April, in preparation for the UK government’s plans to change its regulations on driverless vehicles at some point in the second half of 2026.
In November 2019, Waymo – owned by Google’s parent company Alphabet – secured permission from the California Department of Motor Vehicles for its vehicles to carry passengers without the need for a safety driver who could intervene in the case of a potential collision, making it the first company in the world to secure such clearance.
It has since established itself as the market leader in the United States for self-driving taxis, with commercial operations in San Francisco, Phoenix, Los Angeles, and Austin, and began testing its autonomous ‘robocabs’ in New York City last year.
> Waymo self-driving car lets cyclists pull out of blocked cycle lane
At London’s Transport Museum on Wednesday, Waymo showed off the fleet of taxis it hopes will be operational on London’s streets by September. The vehicles are currently being driven around London by a safety driver mapping the capital’s roads.
But when the scheme eventually launches, it will be driver-free, with customers able to hail a robotaxi through an app, with fares at a “competitive but premium” price, the company says.
According to Waymo, their cars use four sensor systems to gather data from the world outside – radar, lidar, vision, and microphone – enabling the vehicles to be “aware” of their surroundings up to a distance of three football pitches, and including during bad weather.
A powerful computer in the boot processes the data obtained by the sensor, determining how the car acts and reacts in “real time”.
> Uber warned over safety issues with its self-driving cars days before cyclist killed
Riva companies with similar systems, such as Uber and Lyft, also say they are ready to bring their own autonomous taxi services to the UK when the law permits.
Speaking at the launch event this week, local transport minister Lilian Greenwood said she expects self-driving cars like Waymo’s will make London’s roads safer.
“We’re supporting Waymo and other operators through our passenger pilots, and pro-innovation regulations to make self-driving cars a reality on British roads,” she said.
“We know that unlike human drivers, automated vehicles don’t get tired, don’t get distracted, and don’t drive under the influence.”
Nevertheless, Greenwood insisted that, before they become operational in London, self-driving cars will have to meet strict safety standards, “including protection from hacking and cyber threats”.
And speaking to road.cc, the London Cycling Campaign has expressed reservations about their ability to adapt from the wide, straight roads of California to London’s winding lanes.
“As with all new innovation, it’s really early days for Waymo and other autonomous ride-hailing services in London,” the campaigns chief executive Tom Fyans told us.
“Waymo claims they’re far safer in the US than traditional taxi services. But whether that is still the case on London’s infamously complex, congested and contested streets, remains to be seen.
“At LCC, we talk to political leaders, innovators and private companies of all stripes all the time – to make sure everyone’s working hard to make London a better place for healthy, safe cycling for everyone. We hope new ride-sharing services will add to that, rather than detract from it.”
When it comes to how its cars operate around cyclists, it’s fair to say that Waymo has something of a chequered history.
When it first launched as Waymo back in 2016, the firm said its cars are programmed to recognise cyclists as “unique users of the road”, drive conservatively around them, and recognise common hand signals.
In 2019, the company also released a video showing one of its vehicles predicting that cyclists will move out onto the road to pass a car blocking a cycle lane, with the taxi slowing to allow them to safely move across.
However, in February 2024, a San Francisco cyclist was left with “non-life-threatening injuries” after one of the company’s taxis failed to detect his presence and struck him.
According to the company, “the cyclist was occluded by the truck and quickly followed behind it, crossing into the Waymo vehicle’s path. When they became fully visible, our vehicle applied heavy braking but was not able to avoid the collision.”
And last June, another cyclist in San Francisco sued the company after she was seriously injured when one of the brand’s driverless taxis stopped in a cycle lane and a passenger opened its back door, striking the cyclist and causing her to smash into another Waymo car that was also illegally blocking the bike path.
According to the lawsuit, the Safe Exit system employed by Waymo, which aims to alert passengers of surrounding dangers and hazards, failed – leading 26-year-old Jenifer Hanki to claim that Waymo knows its cars are ‘dooring’ cyclists.
Following the “violent” crash, which left her with a brain injury, as well as spine and soft tissue damage, preventing her from working or riding her bike, Hanki sued Waymo and Google’s parent company Alphabet in San Francisco County Superior Court alleging battery, emotional distress, and negligence, while seeking unspecified damages.

18 thoughts on “Safety of driverless taxis on London’s “infamously complex, congested, and contested streets remains to be seen”, say cycling campaigners – as ‘robotaxi’ service set to launch this year”
Waymo wrote:
As wycombewheeler commented on the original article:
The car had sight of the HGV and waited, but it failed to anticipate the occluded zone around the HGV as a possible space for other, vulnerable road users. Rather than assessing the space as clear, it made an assumption in the absence of appropriate data that it was clear (or that the probability of somebody being there was sufficiently low as to take the risk).
How very human.
GMBasix wrote:
Actually more head in the sand ostrich like, if I can’t see you then you don’t exist!
Ravenous Bugblatter Beast of
Ravenous Bugblatter Beast of Traal?
Well, given that taxi drivers
Well, given that taxi drivers are the second-most dangerous to cyclists and pedestrians, after HGVs, as measured by KSIs-per-mile*, this is an interesting development. The key question is whether these Waymo ‘drivers’ are more or less dangerous than human drivers. I suspect a key issue with translating from the USA to the UK is that the USA has much more car-centric laws and planning (e.g. the existence of jay-walking as a crime).
*to be fair to taxi drivers, this could be because they are dangerous idiots who perform dodgy u-turns and other manouvres because they are looking for a fare and driving whilst knackered. Or it could be because they are more likely to be on busy city centre roads at times when there are drunk careless pedestrians.
Talking of taxi drivers :
Talking of taxi drivers :
https://www.gloucestershirelive.co.uk/news/cheltenham-news/taxi-driver-caught-using-ghost-10784607
Taxi driver caught using ghost plates. Now what would they want to do that for?
If one of these kills you,
If one of these kills you, your family can take comfort that you “live on” as part of a company’s proprietary AI training data. ?
A ghost in the machine!
A ghost in the machine!
Yes UK roads aren’t as car
Yes UK roads aren’t as car friendly as those in the US, however computing power has increased exponentially since trials first began.
Both as a driver and a cyclist, I can’t wait for driverless cars. Removing the driver (and the necessity to drive) can only be a good thing. The information I see coming from the US (independently compiled) shows that driverless cars are way safer than driven cars.
If I have a concern it is that traffic levels will go up as “driving” somewhere becomes so cheap and easy.
Both as a driver and a
Both as a driver and a cyclist, I can’t wait for driverless cars
What about the inevitable survival of the fittest ‘hacked’ driverless cars, where the safety features are bypassed to bully other cars (which still have the safety features and are therefore ready to have ‘sand kicked in their faces’ like other weaklings) into coming to a halt and deferring to the ubermensch? There will no policing of these modifications, just as the police can’t be bothered to pull the illegal electric motorcycles off the roads for confiscation.
NickSprink wrote:
Careful what you wish for. You may be unknowingly uttering the modern equivalent of “I’m so glad for these new horseless carriages! Taking the nervous horse out of driving will make it so much safer and more pleasant for all us other road users!”
I think your latter point is very much a concern. And it doesn’t even have to be cheap – this new future may offer more possibilities for fewer, larger cartels, which doesn’t necessarily make it “too cheap to walk” at all. And look at how drivers are prepared to pay now (and complain, even when the general taxpayer is picking up part of the costs) because they “have to drive”.
Businesses effectively “capturing” public space, new businesses in a sector not being well- motivated to provide a safe pathway to the future, motoring businesses discounting the “externalities” (eg. the fact of “safe” transport feeling very unsafe and unpleasant to those on the outside of the vehicles… surely *that* couldn’t happen (again)…?
I found the following vision of what we might be welcoming in excessively dystopian, but (given how things have gone in the past) plausible – Notjustbikes video on self-driving vehicles and the future: https://m.youtube.com/watch?v=040ejWnFkj0
Anyone else looking to avoid
As I said yesterday:
Anyone else looking to avoid commuting due to this?
I’m glad they’ve given us warning of when to stay away.
Whilst I appreciate that we can think a computer controlled vehicle should be safer than a human driver, we clearly are not there yet going by the KSIs experienced during testing in various locations.
I would also think that they are only good enough right now on straight easy road systems, which simply does not apply to our roads here in the UK.
And thats before applying anything else such as temporary traffic restrictions etc.
The best case scenario is the trial simply grinds motor traffic to a halt because the computers can’t cope and is quickly stopped.
Anyone else looking to avoid
Anyone else looking to avoid commuting due to this?
I’m glad they’ve given us warning of when to stay away.
Nope. I’ve spent loads of time in Waymos in the USA and outside of Phoenix it’s not a traditional grid structure. Your traditional black cab or minicab driver is invariably on a phone / tired / thinking about dinner / next fare / chatting to the passenger etc.
While I’m certain that there will be accidents involving these things as there are in the USA, what tends to happen is that they’re nearly always low speed. They don’t have the aggression problem endemic to car drivers.
Your post sounds like the traditional regurgitated Daily Mail content also associated with hatred of renewable energy and EVs, and so on.
Fursty Ferret wrote:
I’m no fan of alphabet/google or AI, but for the reasons stated above, I’d feel safer on the roads with driverless taxis than with cab driver taxis. Other bonuses – there’s the possibility that the driverless cabs will learn from mistakes, will follow the law and won’t abuse their passengers (Worboys, for example).
Well, when there are one or
Well, when there are one or two, and the firms are still chasing regulatory approval and chasing market share, you’re probably right.
BUT if they get their foot in the door AND (open question) these does turn out to be something that they get onto roads in numbers, expect a) these companies pushing for commercial desirables for them * and b) as a result of those further pressure on people not to walk or cycle (possibly direct, but definitely passive/ indirect just like happened with standard motor vehicles) c) maybe no dodgy taxi drivers but if the vehicle is autonomous you’re no more in control of where it goes than worth a cabbie – and that could be people who’ve hacked these…
* what might companies lobby for which could be problematic? Increased speed limits for AVs (because these vehicles are so much safer than human drivers of course). Extra access to otherwise motor-traffic-reduced spaces (ditto).
Special spaces and lanes for these vehicles. And perhaps further rules limiting access for things that might limit / impedes these AVs (which companies will say is an issue because the AVs can deliver greater efficiencies / effective higher capacities by being shared not needing a driver and running round the clock). Which means “get those cyclists and pedestrians off the roads!”
I have been in Waymo taxis a
I have been in Waymo taxis a number of times in the US and they are far safer than humans..
I have also been in Wayve self driving test vehicles in London in heavy traffic, narrow complex streets and they had no issues at all.
I don’t think Waymos will have any issues launching a service here and personally would rather have them on the roads than back cabs drivers.
I wonder if they will u turn
I wonder if they will u turn without indicating? They need to programme them to be able to do that – in order to ‘conform to standard’.
Rome73 wrote:
I’ll be disappointed if they don’t have an external speaker with a computerised voice shouting “Do you pay road tax, wanker?” and “Get a farkin’ job!” at me from time to time.
May I also re-recommend the
May I also re-recommend the Dystopian AV prophecy video linked by CoaB below. One entertaining suggestion is totally disabling an AV by putting a traffic cone on the bonnet (hood to the Trumpians on here). I still think they’re likely to be safer than Audi/ BMW drivers etc. and at least they’ll have MOTs! However, the video illustrates the societal catastrophe the widespread introduction of AVs by MegaCorp could bring