Uber Tech partners with NVIDIA to build a fleet of 100,000 autonomous vehicles, planning to launch in 2027

Uber announced on Tuesday that it will expand a fleet of 100,000 autonomous vehicles powered by NVIDIA technology starting in 2027, stating that this plan is expected to reduce the cost of providing on-demand robotaxis to consumers.

The two companies established an early partnership in January this year. At that time, Uber agreed to provide NVIDIA with some driving data to help improve NVIDIA's artificial intelligence models and chip technology, enabling automakers to develop autonomous vehicles based on this data. On Tuesday, NVIDIA released a new technology platform—Nvidia Drive AGX Hyperion 10—that allows manufacturers to equip vehicles with hardware and sensors that work in conjunction with compatible autonomous driving software.

As part of this collaboration, Stellantis will be one of the first automakers to provide Uber with NVIDIA-powered autonomous taxis. According to a statement released by Uber on Tuesday, Stellantis will deliver at least 5,000 of these autonomous taxis for Uber to operate in the U.S. and international markets. Uber will be responsible for the end-to-end fleet management of these vehicles, including remote assistance, charging, cleaning, maintenance, and customer service.

Stellantis stated in a separate announcement that it will collaborate with Foxconn for hardware and system integration. Production is scheduled to begin in 2028, with initial operations starting in the U.S. The statement mentioned that pilot projects and testing programs will be launched in the coming years.

Media analysis suggests that these commitments will enable Uber to offer more autonomous taxis on its platform, thereby helping to reduce operational and commercialization costs. The company has established partnerships with more than a dozen autonomous driving technology developers and has committed to investing in some of these companies. Uber is betting on a future transportation ecosystem composed of both robotaxis and human drivers.

After the announcement, Uber's stock price initially rose by 1.2% but then turned to decline.

Currently, Uber has launched autonomous driving services in Austin, Texas, and Atlanta, Georgia, in collaboration with Waymo, a subsidiary of Google's parent company Alphabet, and is providing robotaxi services in Abu Dhabi and Saudi Arabia in partnership with WeRide, although the scale of these fleets remains limited. Uber has stated that it expects to expand the fleet with Waymo in Austin and Atlanta to "hundreds of vehicles" in the future; however, this is still negligible compared to Uber's millions of ride-hailing drivers and delivery personnel. This also makes it difficult for Uber to achieve substantial profits through its autonomous vehicle fleet. Currently, Uber still outsources daily charging, cleaning, and maintenance tasks to fleet operators.

However, the collaboration with NVIDIA will help increase the supply of these vehicles and gradually reduce related costs in the future. Uber's existing and future autonomous driving partners, including Avride, May Mobility, Momenta, Nuro Inc., Pony.ai, Wayve Technologies, and WeRide, can all utilize NVIDIA's technology to contribute to the deployment of this fleet of 100,000 vehicles on the Uber platform Among the target of 100,000 vehicles, there are also 20,000 Lucid Gravity and Nuro autonomous vehicles that Uber committed to jointly procure and operate with partners over the next six years in July this year.

Uber is also collaborating with NVIDIA to establish a "robotaxi data factory" for the research and development of autonomous vehicles. In this process, Uber will collect over 3 million hours of exclusive driving data from autonomous taxis to train and validate self-driving models. NVIDIA will provide processors, artificial intelligence models, and tools for data filtering, retrieval, and simulation for the project.

Uber stated in a press release:

"These capabilities together form a powerful data engine—covering data acquisition, annotation, scene mining, synthetic data generation, and large-scale training, aimed at shortening the path from pilot phase to profitable autonomous driving deployment."

The market has risks, and investment should be cautious. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investment based on this is at their own risk