Uber Technologies has announced plans to deploy 100,000 autonomous vehicles powered by Nvidia technology, marking a significant expansion that could reduce robotaxi operating costs and accelerate commercialization of the expensive technology.
The companies expect the rollout to begin in 2027, Nvidia announced Tuesday during its GTC conference in Washington, DC. The collaboration builds on a January partnership where Uber agreed to share driving data to improve Nvidia’s AI models and chip technology for autonomous vehicle development.
Technology Platform and Manufacturing Partners
Nvidia unveiled its Drive AGX Hyperion 10 platform, enabling manufacturers to equip vehicles with hardware and sensors compatible with autonomous-driving software. Stellantis will be among the first automakers delivering at least 5,000 Nvidia-powered robotaxis for Uber’s U.S. and international operations, with Uber managing fleet operations including remote assistance, charging, cleaning, maintenance, and customer support.
Stellantis will collaborate with Foxconn on hardware and systems integration, targeting production start in 2028 with initial U.S. operations. Pilot programs and testing are expected to expand over coming years.
Current Operations and Expansion Strategy
Uber currently offers autonomous rides in Austin and Atlanta through Alphabet’s Waymo, and in Abu Dhabi and Saudi Arabia via China’s WeRide. However, existing fleet sizes remain limited—Waymo’s Austin and Atlanta presence is expected to reach only “hundreds” of vehicles, minimal compared to Uber’s millions of human drivers and couriers globally.
The limited scale makes profit margin optimization difficult, as Uber currently outsources robotaxi fleet operations, including daily charging, cleaning, and maintenance.
Data Factory and Development
Uber and Nvidia are building a “robotaxi data factory” for autonomous vehicle advancement. Uber will collect over three million hours of robotaxi-specific driving data for model training and validation, while Nvidia provides processors, AI models, and associated tools for data curation, search, and simulation.
“Together, these capabilities form a powerful data engine – spanning ingestion, labeling, scenario mining, synthetic data generation and large-scale training – that aims to shorten the path from pilot to profitable autonomy deployment,” Uber stated.
The 100,000-vehicle target includes 20,000 Lucid Gravity and Nuro vehicles Uber committed to purchasing in July. Existing and future robotaxi partners, including Avride, May Mobility, Momenta, Pony.ai, Wayve Technologies, and WeRide, can utilize Nvidia technology to contribute to the fleet expansion.
