統計製程管制應用於管控系統變異之研究

Abstract

隨著生產自動化程度推進，系統使用程度也跟著生產模式越來越密切，更多的新生產方式與生產產品也迫使系統需不斷地演進來適應環境。系統演進帶來了更多的企業利潤，但也使得系統維護成本上升。 系統在長時間維護資源匱乏下，產生了不穩定因子，這不穩定因子等同系統的地雷，監控系統狀況依賴維護人員的過去經驗，使得監控上錯失發現異常的良機，往往等到系統發生異常時才可能會被發掘，而系統發生異常對於企業來說等同是個災難。 統計製程管制(Statistical Process Control，SPC)在製造業已被廣泛使用在管控產品品質，但製造業在軟體發展流程管控的推行並非能夠推展很順利。本研究建立一套流程與模型，以製造業觀點輔助軟體維護人員使用SPC來管控系統，對於系統開發方面也提供一套流程，協助再利用系統資源以減少開發成本。系統經過SPC方式檢視狀況，可以提早發現系統異常點加以改善，透過改善的過程中也可釋放系統資源形成一個良性的循環。

The more automated production process, the more closely related to the degree of use of the control system and the production model. More new production methods and new products makes the system needs to evolve continuously so that it can adapt to the new environment. Although the system evolution might bring more enterprise profit, it causes the system maintenance cost rise. Lacking of maintenance of the system for a long time will generate instability factor. This factor may make the system unstable or even crash it at any time. Most monitoring systems are designed based on past experience of system maintenance staff. Those systems are often unable to detect system abnormalities in the earlier stages and thus it may cause a fatal exception ultimately, which is a disaster for an enterprise. SPC(Statistical Process Control) has been widely used in the quality control in manufacturing. However, the promotion of using Software Development Process control for the manufacturing industry is not quite be recognized. In this study, we established a set of processes and models, in the manufacturing point of view, to assist the software maintenance personnel to use SPC control system. We also provide a process for system development so that system resources can be reused to reduce the development costs. Inspecting the system status through SPC, the abnormalities can be detected earlier so that we can improve the system in time. Through the process of improving the system, some resources can be released to form a benign cycle.