Uber has unveiled its ambitious plan to turn its vast network of drivers into rolling data-collecting machines, providing critical insights for autonomous vehicle companies. Chief Technology Officer Praveen Neppalli Naga revealed this vision at TechCrunch’s StrictlyVC event.

The goal is simple: harness the millions of rides provided by Uber drivers to gather real-world driving scenarios, which are essential for training AI models. However, regulatory hurdles and the need for clarity on sensor usage remain key challenges. For now, AV Labs operates a dedicated fleet but envisions a much broader transformation.

The bottleneck in autonomous vehicle development is no longer technology but data collection. Companies like Waymo require extensive datasets to train their models, but this often comes at a high cost. Uber’s approach could democratize access to such data, potentially altering the landscape of the entire AV industry.

By partnering with 25 AV companies and building an 'AV cloud' library, Uber positions itself not just as a ride-sharing service, but as a crucial data provider. The company hopes to invest more directly in partnerships to leverage this unique position.