Skip to content
AI-Daily-Builder

2026-06-07

WeRide, Uber and AVOMO bring robotaxis to Madrid — Spain's first commercial AV service

views

On June 2, 2026, WeRide, Uber and fleet operator AVOMO announced Spain's first commercial robotaxi pilot in the Region of Madrid — WeRide GXR vehicles hailed through the Uber app, launching later this year with safety operators before scaling to hundreds of cars and eventual

Spain joins the robotaxi map

On June 2, 2026, WeRide, Uber and fleet operator AVOMO (a Moove Cars Group company) announced plans to launch Spain’s first commercial robotaxi service in the Region of Madrid, working alongside the regional government (Comunidad de Madrid). Riders will hail a WeRide vehicle through the Uber app, with service expected to begin later in 2026. It is the first time the WeRide-Uber pairing has entered the European market, and it makes Spain Uber’s fifth European robotaxi market.

The hardware is WeRide’s GXR robotaxi, running on the company’s “WeRide One” autonomous-driving stack. As is now standard for a first-city rollout, the early phase keeps a trained safety operator behind the wheel; the partners say they intend to scale to hundreds of vehicles as performance milestones are met, then expand to fully driverless commercial service across Madrid’s core urban areas.

How this fits the broader deal

Madrid is not a one-off pilot. It is the fourth city in a 15-city global framework WeRide and Uber laid out, with the remainder targeted through 2030. The two firms first paired up in 2024 and already run fully driverless commercial robotaxi service together in Abu Dhabi and Dubai, with Riyadh slated next. WeRide says it now operates autonomous vehicles across more than 40 cities in 12 countries, and Madrid becomes its 12th city for robotaxi operations specifically.

The AVOMO role is the part worth underlining. AVOMO already serves as Uber’s AV fleet-operations partner in Atlanta and Austin in the US, where it runs on the order of 400 autonomous vehicles. Fleet operations — depot management, cleaning, charging, maintenance, remote assistance, incident handling — are the unglamorous layer that determines whether an AV service is actually a business or just a demo. Putting an experienced operator into the Madrid launch from day one signals the partners are treating this as an operational deployment, not a science project.

ElementMadrid deployment
AnnouncedJune 2, 2026
Tech / vehicleWeRide One stack, WeRide GXR robotaxi
BookingUber app
LaunchLater in 2026, safety operator on board initially
PartnersWeRide (AV tech), Uber (platform), AVOMO/Moove (fleet ops), Comunidad de Madrid (regulator)
Framework4th of 15 WeRide-Uber cities through 2030

Why a robotaxi story is a physical-AI story

It is tempting to file robotaxis separately from humanoids and manipulation, but they are the same embodied-AI problem under different packaging: a learned perception-and-control policy that has to act safely in an open, adversarial physical world it cannot fully model in advance. A Madrid street — narrow lanes, scooters, pedestrians who jaywalk by local convention, roundabouts, and Spanish-specific signage and right-of-way norms — is a genuine domain shift from the Gulf cities where this stack has its driverless track record. The interesting technical question is not whether the car can drive, but how much of the Abu Dhabi/Dubai policy transfers and how much has to be re-collected and re-tuned for European urban behavior. The safety-operator phase is, in practice, the data-collection-and-validation phase.

Practitioner note

Treat “launch” language carefully. What was announced is a permitted pilot with safety operators and a staged path to driverless, not paid driverless rides on day one — the same sequence Waymo, Zoox and others follow. If you are tracking AV progress as a leading indicator, the milestones that matter are (1) removal of the safety operator on defined routes, (2) the move from a fixed pilot zone to “core urban areas,” and (3) fleet count actually reaching the promised hundreds. Each is a separate regulatory and operational gate, and the gap between announcement and any one of them is typically measured in quarters.

Under-considered angle

The quietly significant detail is the division of labor: a Chinese AV developer (WeRide), a US ride-hailing platform (Uber), and a Europe-facing fleet operator (Moove/AVOMO), stitched together under a local regulator. This decoupling — algorithm vendor, demand aggregator, and asset/fleet operator as three separate companies — is a different industrial model from the vertically integrated approach (build the car, the stack, the app and the depot in-house) that dominates US robotaxi coverage. If the WeRide-Uber framework actually delivers cities on schedule, the takeaway for embodied AI broadly may be that the winning structure is modular and partner-led rather than fully owned, precisely because the hard, expensive, locale-specific work lives in fleet operations and regulatory approval rather than in the model weights. Watch whether European regulators are comfortable with a non-domestic AV stack as the long-run dependency.


Sources

Tip