
Uber aims to expand its AV Labs program by outfitting its millions of human-driven cars with sensors to collect real-world data for autonomous vehicle (AV) companies and AI model training. This approach could provide vast, diverse data sets that AV developers currently lack due to high costs of deploying their own fleets. Uber is building an "AV cloud" to share labeled sensor data with partners and allow them to test models in "shadow mode" on real trips. The company sees this as a way to democratize AV data and maintain relevance in the evolving AV industry, despite having abandoned its own self-driving car ambitions.