Uber is deploying 500 modified Hyundai Ioniq 5 vehicles across the United States this year to fuel its autonomous vehicle ambitions. Each vehicle carries a suite of sensors designed to collect real-world driving data for the ride-hailing giant's newly established AV Labs division.
The data collection effort targets one of self-driving development's hardest problems: understanding edge cases and rare scenarios that machine learning models need to handle safely. Uber's fleet will capture footage and sensor readings from diverse driving conditions, traffic patterns, and road infrastructure across multiple cities.
Uber has been rebuilding its autonomous vehicle program since 2020, when the company sold its ATG (Advanced Technologies Group) to Aurora Innovation. That move marked a strategic pivot away from full self-driving development. The new AV Labs division represents a more focused approach, emphasizing sensor technology and data pipeline infrastructure rather than end-to-end autonomous taxi services.
The Ioniq 5 choice reflects a practical strategy. Hyundai's electric SUV offers sufficient cargo space for sensor arrays and thermal management systems that multi-camera and LiDAR setups require. The 500-vehicle deployment also signals Uber's commitment to scale. At that volume, the company can capture data across geographies and weather patterns in a year rather than over multiple years.
This data will inform Uber's broader autonomous vision, potentially supporting future integrations with partners building self-driving technology. Rather than competing directly against established players like Waymo or Aurora, Uber positions itself as a data and logistics platform that autonomous vehicle companies can leverage.
The AV Labs announcement also carries competitive weight. Waymo operates a smaller commercial fleet in select cities. Tesla collects autonomous driving data through its existing vehicle fleet. By putting 500 sensor-laden vehicles on roads nationwide, Uber establishes itself as a serious data collector in the autonomous