| Summary: | 碩士 === 國立中山大學 === 高階經營碩士班 === 104 === In this thesis it described in manufacturing how to pursuit low quality risk with high yield rate and controllable production cost, and how to utilize new technology to control the status of products quality in advance during manufacturing process in order to reduce the chance of insufficient delivery quantity or delivery delay due to defective quality, moreover, to provide necessary assistance to technicians to focus on the search of real cause for defective quality in order to reduce the time needed for corrections and to minimize the man power and resources required for the improvements. In additions, it is also expected with the results of this research to enhance new products in the industry and the capability of research development in production process. If the Enterprises can grasp highly controlled manufacturing capabilities and quality control capabilities not only can they sufficiently increase their compatibilities but it can also use the resources effectively to reduce waste and the impact of pollutions on the environment by avoiding the excessive use of resources.
The research method in this study utilized the machine equipment used in production process and the necessary parameters applied in mass production settings and the actual feedback. And from the Automatic Virtual Metrology (AVM) system that has been used in semiconductor and opto-electronic industry it retrieves equipment process parameters to analyze and predict product quality and the device status to see if the method can be applied to the R2R FPC industry operating profile. The AVM system contains Model Creation Server (MCS), AVM Server, VM Management Server (VMM Server), Database, and the user end (Client). On the base of collecting of the production/measurement data, it focused on the specific machine to first establish the prediction model groups (DQIX, DQIY, RI and GSI) to estimate the results of each batch production in order to confirm the key parameters of the machine operation during the FPC production. Hence it is able to effectively predict and control product quality, and to increase yield rate and unit counts. In additions, it provides of manager the ability to control the producing time and quantity for product supply in advance. After the data validation, AVM system-specific models of diffusion technique can be used to automatically transfer the forecast model to the other machines of the same type. It is expected to effectively decrease staff workload, reduce human operation errors, as well as to provide better monitoring and control of production data, and to aid and promote effectively on the production stability.
|
|---|
