Smart Manufacturing and Artificial Intelligence Enabling New Energy Vehicle Production

The new energy automobile industry faces the challenges of low manufacturing efficiency, high energy consumption and lack of flexibility. In this study, we propose the technical framework of “3D sensing + digital twin + edge computing”, which achieves 0.1mm inspection accuracy of highly reflective metal parts through the fusion of laser scanning and structured light imaging, and the defect leakage rate is reduced from 8% to 0.5%; reinforcement learning dynamically optimizes the energy consumption of coating process, and reduces the energy consumption of a single battery cell by 24%; spatiotemporal map convolutional network (STGCN) compresses the switching time of a vehicle to 18 minutes. Convolutional Network (ST-GCN) compresses model switching time to 18 minutes. Empirical studies show that Geely Xiangtan Base has reduced annual costs by 21 million RMB after applying the framework, which meets the carbon footprint traceability requirements of the European Union and pushes the industry towards the goal of “zero defects, zero carbon emissions and zero delays”.