Uber is fast-tracking its journey towards autonomous technology with an ambitious strategy to deploy 100,000 self-driving vehicles by 2027. The ride-hailing leader has formed a strategic alliance with Nvidia to develop and manage a substantial fleet of robotaxis, a development that has the potential to transform urban mobility experiences. This partnership was revealed at Nvidia’s GTC conference in Washington, DC, marking a significant move towards widespread automation in the ride-hailing sector. Nvidia’s state-of-the-art AI computing systems and self-driving vehicle technology will support Uber’s forthcoming fleet, facilitating the integration of robotaxis as a regular option within the platform.
This new venture builds on a previous agreement where Uber provided segments of its driving data to aid Nvidia in refining its AI and autonomous driving models. At the conference, Nvidia also introduced its latest technology, the Nvidia Drive AGX Hyperion 10 platform, aimed at equipping car manufacturers with the necessary hardware, sensors, and software for safe and efficient autonomous driving. As part of this collaboration, Stellantis NV, which produces brands like Jeep, Peugeot, and Chrysler, will deliver at least 5,000 Nvidia-enabled robotaxis to Uber. Foxconn will contribute to hardware integration and system assembly, with production anticipated to start by 2028. Once operational, Uber will oversee all functions, including remote assistance, charging, maintenance, cleaning, and customer service.
Although the extensive rollout is still a few years away, Uber is set to initiate pilot programs and testing shortly. These initial trials will progressively expand to achieve the targeted 100,000 autonomous vehicles by the decade’s end. The potential advantages for Uber could be significant; a fleet of self-driving cars is expected to lower operational expenses, enhance reliability, and offer a more uniform customer experience. The company’s long-term profitability has frequently been jeopardized by its reliance on millions of human drivers globally. With robotaxis, Uber could gain improved control and efficiency, assuming the technology is commercially viable and obtains regulatory approval.
Currently, Uber runs limited autonomous ride services in Austin and Atlanta through a collaboration with Alphabet’s Waymo and also works with WeRide in the Middle East. However, these initiatives involve only a few hundred vehicles. The partnership with Nvidia could greatly broaden Uber’s international presence in autonomous transportation, providing access to a wider network of AI developers and manufacturers. Besides the goal of 100,000 vehicles, Uber confirmed that the commitment includes the 20,000 Lucid Gravity and Nuro vehicles previously pledged under other contracts. Developers such as Avride, May Mobility, Momenta, Nuro, Pony.ai, Wayve Technologies, and WeRide are also anticipated to be essential in constructing this extensive autonomous network.
Furthermore, Uber and Nvidia are creating a “robotaxi data factory” — a center focused on gathering, labeling, and analyzing over three million hours of driving data. This extensive dataset will be utilized to train and enhance AI models, ensuring safer and more efficient driverless systems. According to Uber, the partnership aims to establish “a powerful data engine — encompassing ingestion, labeling, scenario mining, synthetic data generation, and large-scale training — that seeks to expedite the transition from pilot to profitable autonomy deployment.”
