Mobileye Is Finally Doing Their Own Robotaxis
Plus: Around the Nurburgring, autonomously.
This is the Ride AI newsletter: The most comprehensive weekly digest of news and intelligence at the intersection of technology and transportation.
ICYMI, early bird tickets for Ride AI 2027 are available. If you’d like to apply for an early spot, please do so here.
Top Story
Mobileye is launching its own robotaxi business, moving beyond selling self-driving systems to owning and operating an autonomous ride-hailing service in a U.S. city starting in 2027. The plan starts with about 100 vehicles in a major U.S. metro, phased through 2027, then scales to roughly 17,000 over the following five years. Mobileye will pair its Mobileye Drive system with Moovit, its mapping and trip-planning subsidiary, to run the consumer app, fleet management, and teleoperation in-house.
Founder Amnon Shashua framed it as additive, not a replacement for selling Mobileye Drive to automakers, but he argued the industry has grown “increasingly dependent on a small number of technology providers and business models.” A subtle swipe at the Waymo monopoly, perhaps, but it’s a move that’s long been foreshadowed, where nearly every major AV player eventually decides it wants to own the full stack rather than just supply a piece of it. More details are coming at a Capital Markets Day before year-end.
Domestic News
Stellantis, Wayve, and Uber have partnered to jointly develop and deploy Level 4 robotaxis at a global scale, pairing Stellantis’ L4-Ready Platforms, Wayve’s end-to-end Driver, and Uber’s mobility network. Stellantis designs and manufactures the vehicles with embedded sensors, redundancy, and the safety needed for high-utilization driverless use, Wayve supplies the autonomy, and Uber brings the network.
You might remember Stellantis launched a similar partnership a week ago with Pony.ai, based on a minivan and a similar L4-Ready Platform promise. However, Wayve told Ride_AI that the L4-Ready Platform isn’t tied to a single model, but more of an architecture that multiple vehicles will be built on, and the companies aren’t naming which vehicles yet. (I reached out to Stellantis for the specifics but hadn’t heard back by publication.) The deal builds on the partners’ existing relationships, Stellantis and Wayve’s L2++ agreement from last month, and Wayve and Uber’s plan to launch autonomous rides in London, Tokyo, and ten other cities starting this year.
Uber, Nuro, and Lucid have picked Houston as the second market for their robotaxi service, targeting a mid-2027 launch. The fleet uses Lucid Gravity SUVs running Nuro’s Level 4 stack, available exclusively on Uber, following the San Francisco Bay Area launch expected later this year. Nuro is already testing around the clock with safety operators in both Houston and the Bay Area, Uber has leased a 50,000-square-foot Houston depot, and Lucid has started building validation units in Arizona.
The broader deal targets at least 35,000 vehicles over six years across dozens of markets.
Rivian CEO RJ Scaringe says the company will ship supervised, point-to-point self-driving later this year, comparing it directly to Tesla’s FSD. Scaringe laid out a three-step roadmap: supervised point-to-point driving in 2026, eyes-off unsupervised driving in 2027, and a commercial robotaxi service with Uber in 2028. It’s a big jump from Rivian’s current Universal Hands-Free system, which holds the lane and speed across roughly 3.5 million miles of marked roads but can’t handle turns, traffic lights, roundabouts, or parking lots.
Having tried UHF, I have to say I’m a bit skeptical. The experience was incredibly rough, and extremely behind on capabilities and smoothness. Like Tesla’s FSD Supervised, this would be a Level 2 driver-assistance feature with the driver fully responsible. It runs on Rivian’s new in-house compute and what the company calls its large driving model.
International News
WeRide and Uber have announced plans to launch commercial robotaxi service in the Greater Zurich region, their second European market in two weeks after Madrid. Service is expected later this year via the Uber app, in coordination with Switzerland’s Federal Roads Office, with local operator Rydera managing the fleet day-to-day. The company already holds a Swiss driverless permit, secured last November for Zurich’s Furttal region.
Zurich is the fifth of 15 cities under the existing WeRide-Uber agreement. It’s the same strategy behind the Madrid launch I covered last issue, and the third of this issue’s three Uber marketplace tie-ups.
Japanese ride-hailing giant Go raised 88.6 billion yen ($553 million) in the country’s biggest IPO of 2026, earmarking the proceeds for robotaxis and acquisitions. Go runs Japan’s largest taxi app, with 35 million downloads, around 80% of the market by usage, and coverage of 46 of the country’s 47 prefectures. The push into robotaxis is rooted in a very unfortunately human problem: Japan’s taxi-driver count has fallen about 20% in recent years, and an aging population means it’s unlikely to recover.
Go has partnered with Waymo and taxi operator Nihon Kotsu (handling the strategic coordination) and will seemingly follow Uber’s playbook. The company says it won’t build its own autonomy stack, and it hasn’t set a timeline for fully driverless operations. Go isn’t the only transportation company eying autonomy in Japan, however. Uber, Wayve, and Nissan are separately planning a Tokyo robotaxi pilot late this year.
Komatsu has become the first manufacturer to commission 1,000 autonomous ultra-class haul trucks, with the milestone machine now hauling ore at a Nevada gold mine. Its FrontRunner system has run since 2008 and has moved more than 11.5 billion tonnes of material, and the 930E electric-drive truck (the 1,000th was a 290-tonne 930E-5AT) is the most-deployed model with over 500 autonomous units. We’re usually so enamored by the more consumer-facing robotaxis, but this news is a good reminder that off-road autonomy is already routine and profitable in mining, even as robotaxis soak up the headlines.
Unmanned Speed
Xiaomi has set the first officially-timed autonomous lap of the Nurburgring, sending its YU7 GT around the 12-mile circuit with nobody in the driver's seat in 10:29.483. That's about three minutes off the human record (Xiaomi's own YU7 GT holds the outright SUV record at 7:22.755), and both a 1939 Mercedes racecar and a Ford Transit van have gone quicker, but the car did hit 130 mph on the back straight, presumably on custom software since its road-going system isn't built for that, and it was a wet lap. Xiaomi acknowledged the time was relatively slow, but calls the record a starting point, not an end point.
Would you go around the Nurburgring in an AV? Let me know.
Alright, that’s it from me… until next week. If you enjoy this newsletter, share it with your friend, colleague, or boss. Thank you for reading; Sophia out!
I don’t know-- 130 mph with no driver behind the wheel?? Actually, no matter how odd the ride would be, I certainly would take a lap around Nürburgring in an AV. I say odd because a customary ride would have a human driver and a real racing engine. The absence of a human driver and hum (whine!?) of an AV would be the surreal experience I’d remember for a long time. I think taking and reporting about that ride should be on your Ride_AI To-Do list. Obsessed Honkers would enjoy that story...
So great to read about the Komatsu electric-drive truck success! Thank you for again bringing us all sides!