ANALYSIS The global robotaxi map is being redrawn by platform partnerships rather than vertically integrated fleet owners. Uber formalized a Tokyo pilot with a licensed taxi operator2,3, WeRide reported surging overseas revenue and flagged Australia and Southeast Asia as next targets1,7, and Tesla's purpose-built robotaxi fleet remained stuck near 100 vehicles5,6 — a contrast that reveals diverging theories of how autonomous mobility scales internationally.
The three developments, taken together, suggest that the companies gaining ground in international robotaxi deployment are those assembling coalitions — local operators, OEM vehicle partners, and AI-stack providers — rather than attempting to own every layer of the value chain. Regulatory complexity in markets from Tokyo to Abu Dhabi is making local partnerships not optional but structurally necessary.
The big picture
Uber Japan signed an operational partnership with Hinomaru Kotsu Co., Ltd. for a robotaxi pilot in Tokyo scheduled to launch in late 2026. Under the arrangement, Hinomaru Kotsu will manage day-to-day fleet operations including depot management, vehicle cleaning, maintenance, inspections, charging, and uptime9. Uber will provide its ride-matching platform and operational support tools. The pilot will use Nissan LEAF vehicles equipped with Wayve's AI Driver technology, building on a memorandum of understanding signed on March 12, 2026 by Uber, Wayve, and Nissan Motor. In the initial phase, experienced Hinomaru Kotsu drivers will sit behind the wheel as safety operators10. Japanese law requires authorized taxi companies to operate passenger transportation services.
Uber global head of autonomous mobility and delivery Sarfraz Maredia called Tokyo "one of the most complex driving environments in the world," citing Hinomaru Kotsu's "extensive operational experience and rigorous safety standards, as well as Nissan's exceptional automotive engineering and Wayve's advanced AI technology" as vital to the effort.
Meanwhile, WeRide reported second-quarter 2026 revenue of 230 million yuan, an 82% year-on-year increase and a 103% surge from the prior quarter. Gross margin reached 37.5%, up 9.4 percentage points from the same period last year. Overseas revenue jumped 164% year-on-year in the second quarter and 169% quarter-over-quarter. The company's autonomous driving operations now span 13 countries and more than 60 cities. WeRide CEO Tony Han told Reuters the company is "talking with the regulatory authorities" in Australia about expansion, and named South Korea, Japan, and Southeast Asia as additional targets. Commercial deployments of right-hand-drive robotaxi vehicles are set to begin in Singapore and Hong Kong4. In the UAE, WeRide has achieved fully driverless operations in Abu Dhabi and Dubai with a fleet of roughly 400 vehicles.
Han emphasized a localization-first approach: "We want to hire locally and we want to work closely with the local government officials and the authorities and try to use as many local collabor[ators]".
ANALYSIS That language mirrors Uber's Tokyo playbook — both companies are treating regulatory and operational localization as the primary scaling mechanism, not raw fleet size.
Tesla presents a stark counterpoint. Future Fund Managing Partner Gary Black said Tesla's autonomous robotaxi fleet remains "stuck" at roughly 90 to 100 vehicles. Black questioned whether the company has the strategy to scale, arguing that Tesla "needs to invest in a long-term marketing strategy and branding in addition to bringing on engineering talent and building product superiority". He also questioned Tesla's roughly 195-times 2026 P/E in light of the fleet's stagnation.
ANALYSIS The structural difference is not just fleet count but partnership architecture. Uber's Tokyo pilot layers four entities — Uber (platform), Wayve (AI stack), Nissan (OEM), and Hinomaru Kotsu (licensed operator) — into a single deployment. WeRide is replicating a similar pattern across 13 countries, adapting to right-hand-drive markets and local regulatory regimes. Tesla, by contrast, is attempting vertical integration in a domain where regulatory fragmentation penalizes go-it-alone strategies. Australia illustrates the point: WeRide is already in regulatory discussions there, while Google's Waymo was warned it could face a seven-stage approval process taking several years.
WeRide's first-half revenue reached 350 million yuan, a 73% year-on-year increase. ◆ Combined with the 37.5% gross margin reported for Q2, these figures indicate that the partnership-driven model is beginning to generate commercially meaningful returns, not just pilot-stage activity.
What's next
WeRide plans to launch commercial robotaxi services with Uber in Madrid and Zurich later this year. Public operations in Denmark are slated for the first half of 2027. Uber's Tokyo pilot is scheduled for late 2026, initially with safety drivers, with any transition to fully driverless service subject to regulatory approval. Australian transport ministers have agreed to enable conditional deployment of robotaxis in some areas from 2027. ◆ The next twelve months will test whether the partnership-first model can convert regulatory access into sustained revenue growth — or whether the complexity of multi-party coordination becomes its own bottleneck.