Company ProfileTo realize a smart factory, the implementation of Industry 4.0 has to cover the following three dimensions: (1) vertical integration of manufacturing systems, (2) horizontal integration through value networks, and (3) end-to-end digital integration of engineering across the entire value chain.Due to the invention and application of computers, manufacturing execution system (MES) and enterprise resource planning (ERP) systems have been developed during the third industrial revolution. Because these two systems have different functions, they are usually managed by different departments in an organization. However, the objective of vertical integration is to integrate all information in a factory from machines to customers (e.g. PLC, SCADA, MES and ERP) and make decisions based on the integrated information, so that the optimal decision can be made in a timely manner. In order to integrate the two systems, we first need to clarify the logic of the upstream and downstream information flows and then make decisions by using mathematical models rather than human experience and intuition. The advanced planning and scheduling (APS) system plays a key role in this aspect.Since the 1990s, APS has been applied to solve demand planning and production scheduling problems. While demand planning is the brain of ERP, production scheduling is the brain of MES. Our solution focuses on scheduling problems in the high-tech industry. Since the problem can be commonly proved NP-hard, most of the existing solution providers have developed their own heuristics based on historical lead time. The solutions given by their approaches are commonly worse than humans, and have to be modified by experienced engineers.To achieve the vertical integration of manufacturing systems, we have enhanced the current APS system for Industry 4.0. Rather than using historical data, our new APS system (APS 4.0) can find a production schedule better and faster than humans by using advanced algorithms that can take advantage of strong computation power and using real-time data through IoT. Based on our powerful scheduling engine, we have further developed a Just-In-Time spare parts management system to achieve horizontal integration of supply chains. Our solution has been successfully deployed at UMC with positive feedback. We are the first one in the world who can generate production schedules better and faster than humans for semiconductor fabs.