Uber Japan has named century-old taxi company Hinomaru Kotsu as the fleet operator for a Tokyo robotaxi pilot scheduled to begin in late 2026. The arrangement marks Uber’s first autonomous-vehicle deployment in Japan and assigns a licensed taxi company responsibility for depot operations, maintenance, charging, inspections, vehicle readiness, and regulatory compliance.
The partnership is structurally different from Uber’s conventional role as a ride-hailing intermediary. Under Japan’s regulatory framework, a commercial passenger service must be operated by an authorized taxi company holding the appropriate license. Wayve’s AI Driver will handle vehicle control, while the Uber app will provide ride matching and Hinomaru Kotsu will remain the legally responsible fleet operator.
Uber Japan Picks Hinomaru Kotsu to Run the Fleet
Japan’s Road Transport Vehicle Act requires commercial passenger transportation to be conducted through an authorized taxi operator. That rule prevents Uber from independently operating a commercial robotaxi fleet, even if its software supplies the booking interface and an outside company provides the autonomous driving system.
Hinomaru Kotsu will therefore carry the same legal role it performs in human-driven taxi services. Its responsibilities include managing depots, cleaning vehicles, arranging scheduled maintenance, completing safety inspections, charging batteries, and maintaining fleet availability. The autonomous system changes the driving technology, but not the regulatory category of the service: a robotaxi booked through Uber remains a licensed taxi service.
This constraint has already shaped Uber’s business in Japan. The company operates there through UberTaxi, which works with licensed taxi operators rather than relying on the personal-vehicle contracting model used in other markets. The robotaxi pilot extends that established structure, rather than creating an exception to it.
Operation and Technology
The pilot will use Nissan LEAF electric hatchbacks equipped with Wayve’s end-to-end AI Driver. During the initial phase, experienced Hinomaru Kotsu drivers will remain aboard as safety operators. A fully driverless stage, with no human inside the vehicle, will require additional approval under Japan’s Level 4 autonomous-driving framework.
Wayve’s system is designed to translate raw sensor information directly into steering, braking, and acceleration commands. Unlike a conventional modular pipeline, it does not require separate, hand-engineered interfaces between perception, prediction, planning, and control. The company has also emphasized self-supervised learning from real-world driving data and the ability to apply one model across different locations without rebuilding it for every city.
Tokyo will test that approach in a demanding environment. Its streets combine narrow corridors, dense pedestrian and cyclist activity, mixed traffic, kanji and trilingual signage, and stop-sign designs that may differ from those in markets where autonomous systems are primarily trained. Wayve has previously operated in Japan, but the commercial pilot will measure performance within a regulated fleet-service setting.
A Familiar Partner, Not a New Vendor
Hinomaru Kotsu has worked with Uber since 2023, when it began operating 100 Tesla Model Y vehicles for Uber Premium across Tokyo’s 23 wards. The existing charging infrastructure, operating systems, and working relationship reduce the need to establish a fleet operation from scratch.
The company also previously tested self-driving taxi technology with ZMP on Tokyo’s Otemachi-Roppongi corridor. Its president, Kazutaka Tomita, linked the robotaxi announcement to Japan’s driver shortage and said the pilot is intended to show how human-driven taxis and autonomous vehicles can operate alongside one another.
That labor context is significant. Japan’s taxi driver workforce fell from approximately 376,000 in 2009 to 231,938 in 2023, while the median age of a licensed taxi driver is now over 60. New overtime limits for transport workers have added further pressure. Against that background, the company presents autonomy as a complement to professional drivers rather than an immediate replacement.
Lessons for Foreign Autonomous-Vehicle Operators
The operating model illustrates a broader rule for foreign AV companies entering Japan: proprietary driving software does not replace the need for a licensed local operator. The platform may process bookings, and the technology supplier may provide the AI system, but an authorized taxi company must remain accountable for fleet management and compliance.
That arrangement differs from the company-to-consumer deployment model associated with some U.S. robotaxi services. Waymo has pursued local partnerships in Tokyo, working with Go Inc. and licensed taxi operator Nihon Kotsu. It also contrasts with Cruise’s attempt to build a U.S. robotaxi operation, which ended with General Motors ending funding and Cruise being integrated into the automaker’s broader autonomy efforts. The comparison underscores that deployment structures are shaped by each market’s regulatory and operational rules.
China’s Pony.ai offers another parallel: autonomous driving capability still depends on approvals, operating entities, vehicles, and local service frameworks. In each market, the ability to navigate licensing and fleet management can matter as much as the performance of the underlying AI. Japan’s requirement gives licensed taxi groups a central position in commercialization.
Uber’s Wider Autonomy Strategy
The Tokyo pilot is part of Uber’s effort to operate a multi-partner autonomous mobility network. Uber has invested in Wayve, while Wayve and Nissan plan to use the same AI Driver in Nissan’s next-generation ProPILOT driver-assistance system. Nissan is also developing a separate autonomous taxi project in Yokohama.
For Uber, the Hinomaru Kotsu agreement converts a previously announced collaboration framework into a named operating arrangement. It provides a route to test Wayve-powered vehicles in Tokyo while retaining the Uber app as the customer interface and an established taxi company as the regulated fleet manager.
Uber has not disclosed the pilot’s exact launch date, vehicle count, or geographic coverage. Late 2026 remains the stated window, and safety operators will be present initially. The final transition to driverless service therefore depends not only on technical performance, but also on approval from Japan’s transport authorities. In that sense, the Hinomaru Kotsu partnership shows that the future of Uber Japan robotaxis will be built on licensed taxi operations, not outside them.