Driverless taxis are coming to London's streets in the Spring
Plus: The mathematical solution to catching a bike thief, how to ensure your Deliveroo doesn't come from a ghost kitchen, and a property with a very public en suite.
The de facto legalisation of low-level theft is one of the most annoying aspects of life in London, whether it’s the tens of thousands of stolen mobile phones being wrapped in tinfoil before being sent abroad, or the failure to prosecute bike thieves. There are certain non-violent incidents that elicit a shrug from the police, a crime number, and an email closing the case — if you can be bothered to report the theft in the first place.
One of the main reasons the police give for not investigating such thefts is the time it would take officers to review hours of CCTV that could catch the perpetrator. There is a solution — and a Cambridge professor has spelled out for us how police could process hours of footage in just a few minutes, if they would just adopt a simple mathematical trick.
Scroll to the end to find out how — or first read about the driverless taxis coming to London.
I’ve been working on an investigation into a scam that has turned into the single most bizarre story I’ve reported in my journalism career. Paying subscribers will get to read it first this weekend — if you’re not a member, sign up now to ensure you get it straight to your inbox. Your support enables all this reporting to happen.
Get ready to ask “shall we get a Waymo home?” as robotaxis come to London.
Londoners will be able to pay to use self-driving taxis to get around the capital from as early as next Spring, after Google-backed Waymo confirmed plans to launch in the capital.
The first test vehicles, complete with boxy cameras on the roof and a human supervisor in the driving seat, will start to appear on the roads of London in the coming weeks. Paying customers are expected to follow at some point early next year, depending on when central government gives its approval.
Waymo told London Centric they would launch with a very small fleet of autonomous electric Jaguar I-Pace cars. A broader roll out of autonomous taxis in the capital is then likely to take place in late 2027, although don’t expect the streets to be immediately flooded with thousands of vehicles.
Waymo is in a race against its (confusingly similarly named) British rival Wayve to be the first to launch in London. Wayve, which has formed an alliance with Uber, has been testing its autonomous vehicles on the capital’s streets for the last year.
We’re intrigued to see how Waymo’s vehicles, designed for the wide grid system of Californian cities, deal with central London’s narrow, congested street plan complete with winding medieval alleys. While jaywalking when there’s a risk of collision is criminal offence in San Francisco, London is teeming with pedestrians willing to dash across roads, as well as Lime bike riders with a death wish. In theory, autonomous vehicles might be safer for London’s pedestrians and cyclists than human drivers, as Waymo’s cars are known in San Francisco for being less aggressive than some real drivers and less willing to break the law by nipping through a red light.
Also, don’t necessarily expect a Waymo to be substantially cheaper than an Uber. Despite removing the cost of employing a driver, its kit is far more expensive and there are substantial costs to run its systems. Instead, Waymo says it sells itself on selling a sense of safety and privacy by removing the need to interact with another human.
“I don’t think the demand is there,” insists black cab boss.
Driverless taxis aren’t necessarily great news for the drivers of traditional taxis, whether they’re behind the wheel of an Uber or a hackney carriage.
We got on the phone to Steve McNamara, general secretary of the Licensed Taxi Drivers’ Association, who represents black cab drivers. He’s confident his members can face down Waymo’s threat: “The press release is incredibly misleading. When they talk about having driverless cars in London, what they’re actually going to have is a driverless car with a driver in it. They’re not actually launching driverless cars as such. It’s a headline-catching press release.”
(Waymo says otherwise and insists it will have true driverless cars, with an empty space in the driving seat, charging fares in competition with black cabs from next year.)
McNamara suggested self-driving vehicles are a novelty: “I’ve got a robot lawnmower. Every time people come round, they say, ‘What’s that?’ And I tell them, ‘It’s a robot lawnmower that cuts the grass.’ Nobody else I know has one. Why is that? It’s because they don’t trust a robot to cut their grass. Yet somehow, we’re meant to believe that people who won’t trust a robot to cut their lawn are suddenly going to put their kids in a driverless car to be taken to school. I don’t think the demand is there.”
In a quote that we’ll revisit in twelve months time, McNamara told us: “Am I concerned about driverless cabs? In 10 to 15 years time, I can see them becoming an issue. But I certainly don’t see it becoming a major problem any time soon.”
Preposterous property of the week
There’s nothing wildly unusual about this three bedroom £1.75m flat overlooking the Thames in Wandsworth, aside from its location squeezed between a recycling centre and a bus garage. That’s until, as spotted by Madeline Odent, you get to the remarkably open en suite. When we rang the estate agent they accepted that the feature was “unusual” and believed the en suite to feature clear glass without any ability to frost the panes. If you do buy the flat, think carefully about who you invite into your bedroom.
Deliveroo men with ven
The comments section under our pieces is consistently brilliant and Saturday’s piece on the rise of the “host kitchen” elicited this tip from subscriber Nick Duffy on how to game Deliveroo: “At risk of them getting wind of this hack and ruining it: you can filter out basically all the ghost kitchens by filtering for pick-up orders — tends to be only legit restaurants that allow actual customers to collect food themselves. You can then change it to delivery when ordering.”
A more dystopian take came from the Financial Times’ Bryce Elder, who read the piece and proposed a future Deliveroo model based around “a transit van full of ping meals and a microwave” circling London, on the basis that all most customers care about is that their food arrives piping hot. We’d invest.
Sometimes making people feel awkward gets results.
The restaurants at Elephant and Castle that were evicted over an enormous backdated electricity bill will be allowed to return, following a campaign by local activists that was highlighted by media outlets including London Centric. Eventually the council intervened and they have been offered 12 months to pay back their debts. Don’t quite celebrate yet, cautions local activist group Latin Elephant, who say they want to see the deal in writing and applied to all the outlets — but it’s a good excuse for a trip to south London to sample their food at Castle Square.
Thank you to all the paying supporters who make London Centric possible.
If you want to hear about our plans for the future, London Centric was recently featured on the Media Confidential podcast with Alan Rusbridger and Lionel Barber.
If you want to get in touch, please send an email or a WhatsApp.
The police won’t investigate bike theft because there’s too much CCTV to review. Why won’t they adopt this simple mathematical trick?
By Cormac Kehoe
“Bike thefts at stations ‘decriminalised’” was an eye-catching headline on the BBC this month, regarding a London commuter who was told by the British Transport Police (BTP) that they no longer investigate theft if a bike has been locked up for more than two hours.
As the BTP puts it, this is an active choice to prioritise other crimes rather than having an officer sitting in front of a screen and reviewing a tape — and a clear indicator to thieves that they are likely to get away with the theft.
“If it is available we will review around two hours of CCTV footage to try to identify the incident, but it is not proportionate to review longer periods as it keeps officers from being available to respond to emergencies, visibly patrolling railway stations and trains, investigating crimes with identified lines of enquiry or which cause the most harm to victims – such as violent or sexual offences.”
As a result, unless you can pinpoint the time your bike was nicked, they say reviewing the CCTV simply requires too much work time for today’s policing budgets. Given how many bike thefts at stations happen either overnight or during a working day, that’s a lot of incidents not being investigated.
It’s the latest in a steady stream of reporting about how rare it is for the police to actually go after non-violent crimes in London. (If you’ve been here a while, you will remember London Centric’s own story of how we had to track down our own bike.)
But what if there was a solution that could transform the ability for the police to deal with bike theft, and in turn transform what has become an unpleasant part of life for the London cyclist? We’ve spoken to people who think they have some of the answers.
“This would mean about two minutes to process an eight hour tape.”
The first potential solution we heard doesn’t even require any fancy technology – just a simple mathematical principle that could help police officers review a day’s worth of CCTV and identify footage of a thief in a matter of minutes.
Richard Weber, a professor of mathematics at the University of Cambridge, explained the technique, known as “binary search”, could cut the time to search a tape by 99.5%. The approach, which he describes as “a fast way to find something in a sorted list by repeatedly cutting the list in half” is used by programmers and mathematicians to quickly find items in a list of data.
“Suppose the tape is eight hours long and we want to identify the theft point to within a window of five minutes,” he told London Centric. “Binary search reduces the search space by half each time.”
Essentially, it involves the police officer scrolling to halfway through a CCTV recording and asking themselves “is the bike still visible?”
If the bike is still visible and yet to be stolen, they can ignore the first half of the tape and then scroll to halfway through the remaining section. Apply this principle repeatedly and the officer can identify the moment that a theft took place in almost no time.
“The number of inspections needed to locate the theft time to within five minutes is at most seven inspections,” said Weber. “If a human could do one inspection in 15 seconds and then wind the tape to a new point, this would mean about two minutes to process an eight hour tape.”
Can’t we just automate that?
But surely in 2025 there’s an artificial intelligence solution. This doesn’t have to mean deploying facial recognition but rather using technology to review a CCTV tape and highlight leads for a police officer to follow.
Yes, says David Hogg, professor of artificial intelligence at the University of Leeds. “In principle, it should certainly be possible with current technology to automate the process of reviewing CCTV footage to detect when a bicycle theft occurs,” he said. “The action required to remove a securing device would be sufficiently distinctive for current AI methods to work.”
Even if the thief isn’t seen visibly cutting a lock, AI can compare the appearance of the person who dropped off a bike with the CCTV of who was seen next to it at a later time, he said: “By setting a low threshold of what might constitute a theft, a person could then review short clips of possible thefts to identify the real thing.”
Josef Kittler, a professor at University of Surrey who specialises in pattern recognition and machine intelligence, said he had worked on technology that automates reviews of CCTV: “The University of Surrey demonstrated such technology to several UK police forces during the past two years. One of the forces planned to perform trials of the technology on their historical videos already enriched by ground truth established by laborious manual investigation. Unfortunately, they failed to get funding for this trial.”
“Wider privacy issues”
We put the suggestions to the British Transport Police, in the hope of allaying some of the dread that Londoners feel when they chain their bike to a rack outside a station.
“We are looking at how we can work with the rail industry to modernise the tools we have available to us when it comes to CCTV – whether that’s improving the quality [or] our ability to access it more quickly and effectively,” they told us.
“We’re also working closely with train operating companies to support and advise them in helping to “design out” this type of crime, working on prevention measures to target harden and make sure train stations are a hostile environment for potential bike thieves.”
Hogg, the University of Leeds professor, recognises how tricky it can be for some public organisations to incorporate artificial intelligence into their working: “All of this involves automating the review of CCTV from a public space with the wider privacy issues that this raises.”
No word yet on whether letting a couple of professors give a quick talk to the police on mathematical theory would prove as controversial.
Binary Search is such a simple concept and various groups have been trying to get the police in Cambridge to do the same thing due to rampant bike theft, but it’s a battle. I’m not sure if it’s just institutional inertia, or just difficulty getting the message to the right people - probably both. Another challenge is the interface between the CCTV operators and the police - the requirement for a crime number and then GDPR processes seem to be used as an excuse to avoid taking immediate action, or to slow down the process so much that the videos get overwritten which is what happened when my partner’s bike was stolen in Cambridge station.
Cambridge police have become better at prosecuting bike theft, but only after years of intensive campaigning by Camcycle and other local groups - and this is in the UK’s cycling capital.
Whilst it’s easy to see why the police would rather tackle “proper” crime, I think they could gain an awful lot of public goodwill by taking bike and phone theft more seriously. It feels like such a violation when it happens and the police dismissing it by default really adds insult to injury.
Baidu and a few other Chinese firms were recently granted government incentives to trial Autonomous Electric Vehicles throughout the UAE; the aim is 25%, or 7000 vehicles by 2040. Waymo's name is absent from the big Gulf projects (Uber partnered with WeRide for Dubai), which leaves it casting about for the harder, less lucrative markets.
This would be us. Absent Gulf Incentives, infrastructure or regulatory encouragement, Waymo are likely expanding here because nobody else wanted to. The narrow, cohabited roads of our tottering metropole are not to be underestimated, and I imagine we will, no doubt, introduce their AI to the concept of frustration.
As with the Gulf initiatives, this also bodes poorly for the drivers currently slumming for Uber et al, without these bloody foreign robots coming here and taking their jobs.
Of course there is a downside to an AI taxi, and that's the membership tier. I imagine that at some point, if you don't pay the monthly stipend, Waymo will engage in the taxi equivalent of ads, and drive you past their sponsors. Pay your monthly fee, or find each trip involves a few minutes outside a branch of Dominos.
My compliments on another, excellent series of articles, and for the truly capital podcast on Prospect. Most illuminating, dear boy.