Traditional factories operate with data distributed independently between different workshops, making interconnection impossible. Furthermore, the sheer number of machines and equipment lacks a unified control platform. Taiwan AI Cloud combines hybrid cloud and cloud-based AI training resources to help each workshop collect machine data. Through federated learning, models can be trained across workshops while maintaining data security. This allows for effective remote monitoring of equipment information and, through prediction and alerts, achieves goals such as process improvement, increased yield, and reduced quality control costs, ultimately realizing the vision of establishing a digital twin environment.