The robotaxis have reached Manchester, quietly. Uber announced on October 14 that testing has begun in the north-west England city, with self-driving cars from Pony.ai mapping local streets, a safety driver at the wheel and no passengers on board. It is the unglamorous first step every serious robotaxi service takes, and Manchester is now on that path.
No fares. No launch date. Just cars learning a city, street by street.
Three partners, three jobs
The division of labour is clean. Pony.ai supplies the driving system, in its seventh-generation vehicles. Uber supplies the booking platform that riders already have on their phones. Verne supplies operating know-how, earned the hard way: the three partners already run a commercial autonomous taxi service in Zagreb, where this week Zagreb began testing fully driverless robotaxis with nobody on board at all.
That Zagreb connection is Manchester’s real credential. This is not a team sketching an idea. It is a team that has carried paying passengers in one European city, is now removing the safety operator there, and is applying the same playbook in a second country. An agreement from August between Uber and Pony.ai targets more than 2,000 robotaxis across Europe and adds four cities beyond Zagreb. Manchester looks very much like one of the next names on that list.
Mapping before passengers
Why map first? Because a self-driving car does not really drive on the road you see. It drives on a detailed digital model of that road – lane markings, kerb lines, junction geometry, the odd local one-way system – and checks the real world against the model dozens of times a second. Manchester’s streets, with their tram tracks, Victorian terraces and rain, need to be learned properly before anyone sells a ticket.
The test cars run under a government permit covering the Driver and Vehicle Standards Agency and the Department for Transport. No operator has yet applied to run a passenger service in Manchester, and the local transport authority says it has seen interest but received no application. That honesty is refreshing, and it sets expectations correctly: this is a data-gathering stage, and passenger rides are a later decision, taken in public.
The wider UK framework is moving in parallel. A national pilot application process for taxi-style and bus-style driverless services opened in May 2026 under the Automated Vehicles Act 2024, and officials say commercial pilots could start before the full regime is in place. London is further along: Uber began supervised passenger trips there with Wayve on September 3, with a licensed driver on board, and has since announced plans to add Pony.ai cars to the London fleet.
Why Manchester matters
London gets the headlines, but Manchester may be the better test of whether robotaxis work as ordinary British transport. It is a big, wet, busy city with strong local government, a large student population, late-night demand and suburbs that buses serve thinly after midnight. A service that works there works in most of urban Britain.
Scale is the other prize. Uber expects robotaxi trips in up to 15 cities worldwide by the end of 2026, most of them in the United States and the Middle East. Europe is the contested ground, and every city where its partnership is already mapping is a city where a rival has to catch up. The race is no longer about who has the best demonstration. It is about who has the most streets learned.
What it means for travellers, cities and investors
For Manchester travellers, nothing changes this month, and that is the honest summary. You may spot the test cars; you cannot ride in one yet. What has changed is the direction of travel: a named operator, a named technology partner and a permit are now facts, not rumours.
For the city, the mapping phase is a chance to shape the service before it exists. Local leaders can press for clear answers on pick-up points, accessibility and how driverless cars share streets with trams and buses, while the fleet is still small and the operator is still listening. Cities that ask early tend to get better services than cities that argue after launch day.
For investors, watch the sequence rather than the slogans. Zagreb carries passengers. London carries supervised passengers. Manchester is mapping. That is a pipeline, and pipelines are how a technology becomes a network. The next milestone that matters is a passenger application in Manchester – filed, debated and decided in the open.
There is also a local knowledge problem that only miles can solve. Manchester’s Metrolink trams share road space with cars in the centre, its football and concert crowds can flood whole districts in an hour, and its bus lanes and loading bays follow local rules that no generic model guesses correctly. Mapping runs through all of that, in all weathers, are what turn a competent self-driving system into a Manchester driver.
Every robotaxi city starts the same way: with a car driving slowly around the block, learning where everything is. Manchester’s turn began this week.





































Be First to Comment