Uber Reintroduces Self-Driving Cars, But Not for Ride-Hailing Services

Uber is reviving its autonomous vehicle initiative with the launch of its new AV Lab project, aimed at gathering crucial data for its numerous robotaxi partners. These vehicles will be equipped with the standard array of sensors found in self-driving cars, including cameras, lidar, and radar. However, it’s important to note that these cars won’t be operating as robotaxis; their primary role is to collect data for Uber’s partners in the robotaxi space.

This distinction is significant, especially considering Uber’s complicated past with autonomous vehicles. The company divested its AV division in 2020 following a tragic incident involving one of its self-driving cars in Tempe, Arizona. Since then, Uber has shifted its strategy, opting to collaborate with various AV startups rather than develop the technology in-house.

To support these startups, Uber is deploying a fleet of vehicles equipped for autonomous driving within its ride-hailing network. These manually driven cars will not only generate revenue by completing regular Uber trips but will also provide valuable data exposure to the diverse “edge” cases encountered daily, as the company fulfills around 40 million trips. Balaji Krishnamurthy, Uber’s CFO, emphasized the importance of this data collection.

Krishnamurthy suggested that AV operators typically require at least 10 million miles of data to prepare for their first public driverless launch. With this in mind, Uber’s new AV lab fleet aims to accumulate at least 2 million miles each month by the end of this year, with plans to expand further in 2027.

The project is kicking off with a single Hyundai Ioniq 5, although Uber executives have indicated they are open to using other models. The driving data collected will be offered at no cost to Uber’s partners, which include companies like Wayve, WeRide, Nuro, and Waabi. This approach recognizes that many AV developers may not have the financial resources of industry leaders like Waymo and Tesla, and could benefit from assistance in offsetting the costs of launching commercial services.

Earlier this year, Uber also introduced a new initiative called Uber Autonomous Solutions, designed to provide various services to its robotaxi partners. One key offering is training data from Uber’s extensive fleet of test vehicles, which are actively driving in multiple cities. This data is intended to help partners enhance their own autonomous vehicle technologies as they work to accelerate the growth of their robotaxi operations.