Archive — US — techcrunch
Uber to put 500 data-collection vehicles on the road this year
Uber plans to deploy 500 customized Hyundai Ioniq 5 vehicles this year to collect real-world driving data for its autonomous vehicle partners. This represents a comeback for Uber’s AV initiatives since selling its division in 2020, enabling the collection of 2 million miles of data per month, enhancing training datasets for robotaxis.
Key points
- Uber is rolling out 500 modified Hyundai Ioniq 5 vehicles to collect autonomous driving data.
- This marks the first self-assembled vehicle by Uber since its autonomous division was sold in 2020.
- The fleet will collect about 2 million miles of data per month, enhancing the training datasets for partners like Waymo and Avride.
- Vehicles equipped with 14 cameras, eight lidar sensors, and nine radars will be retrofitted by Roush Performance.
- Data will be processed through Nvidia’s Dual Drive Thor computer, with ongoing updates expected.
- Uber aims to achieve a diverse geographic dataset to train self-driving software.
- The initiative builds on previous data collected from vehicles in numerous cities and captured data from Lucid Air vehicles over the past two years.
Why it matters
- This move aligns with the growing trend in 2025 towards robust data collection strategies for AI and autonomous technologies.
- The development is significant in the autonomous vehicle sector, particularly after notable acquisitions and partnerships.
- Increased investment in autonomous vehicles is seen globally, reflecting a shift in transportation technology in the past few years.
Uber's initiative not only marks a significant step in its autonomous vehicle strategy but also showcases the industry's push for extensive, diverse datasets to train AI, highlighting an ongoing trend in the tech landscape as we progress towards 2025.