Voyah's Wuhan Plant Uses China Unicom 5G+MEC AI Visual Inspection to Catch 99%+ of Missing or Misfitted Parts

China Unicom's Hubei branch and Voyah, Dongfeng Motor's premium new-energy vehicle brand, have deployed an AI visual inspection system at six key stations of Voyah's final-assembly plant in Wuhan that detects more than 99% of missing or wrongly fitted body parts and has raised line efficiency by 20%, according to a July 2025 China Unicom release on trade site C114. It is Hubei Province's first 5G+AI industrial quality-inspection project. "Now Voyah can roll 600 complete vehicles off the line in a day, and the fastest delivery from order to customer takes only five days," said Hu Yong, director of Voyah's final-assembly plant. Voyah builds customized vehicles in different models, colors and configurations, and its 2024 sales grew about 70%. Hu said AI visual inspection, a fully 5G-connected factory and other new technology let one line produce SUVs, MPVs and sedans in mixed sequence, with a new vehicle every 118 seconds. The inspection system covers six stations including the instrument panel, doors and tires. It runs on a 5G private network with multi-access edge computing (MEC) and on small AI inspection models developed by China Unicom for the industry, combining motion control with deep-learning algorithms to recognize missing or misinstalled parts automatically. Processing is split across end devices, the 5G network, edge servers and cloud services, which also supports tracing problems back to their source. In a related project the partners separated hardware from software in production equipment, and engineers at a software-flashing center now write more than 50 vehicle software packages over 5G in real time. The project was listed among Hubei's typical AI application cases and Wuhan's first batch of AI-enabled manufacturing cases, and helped Voyah gain the 2024 Smart Manufacturing Demonstration Factory title from the Ministry of Industry and Information Technology. China Unicom Hubei plans to extend the cooperation to AI predictive maintenance and digital twins.