AI Data Lake for Autonomous Driving R&D with Huawei

Title

AI Data Lake for Autonomous Driving R&D with Huawei

Content

GAC Group, one of China's largest automotive manufacturers, has built an AI data lake on Huawei OceanStor storage to power its autonomous driving R&D. The platform stores and processes massive volumes of sensor data including LiDAR point clouds, camera footage, and radar returns collected from test vehicles. GAC's autonomous driving research team uses this data to train perception and decision-making AI models. The Huawei OceanStor storage solution provides the high-throughput, low-latency performance required for AI training workloads. The system handles exabytes of data from thousands of test drives per year, enabling rapid iteration of autonomous driving algorithms. GAC's AI platform supports simulation-based validation, where scenarios from real-world drives are replayed in simulation to test algorithm improvements. The data lake architecture allows different research groups within GAC to access shared datasets while maintaining data security and access controls. GAC has reduced AI model training time by 40% since deploying the Huawei data lake infrastructure.

URL