本研究以面板製造業的製造資訊部門導入ITIL及ISO20000為個案,探討如何利用精實六標準差的工具及DMAIC方法,分析服務台的事故管理流程以及應用程式管理的問題管理流程。文中詳述DMAIC各個階段實作,其中定義階段以狩野分析、SIPOC、VSM(Value Stream Map)以及力托法則(Litte's Law)定義問題;衡量階段以統計軟體衡量所實際的資料並建立基線;分析階段以PFMEA(Process Failure Modes and Effects Analysis)進行流程失效分析並以因果圖、多重投票方式縮小範圍;改善階段,思考如何以IT系統輔助改善;控制階段,擬定控制計畫並以防錯法以及統計流程控制SPC(statistical process control)進行監控有效降低服務管理流程失效的變異與風險。 服務台在事故管理流程失效主要分佈在事故升級階段,原因可分為1. 服務人員不熟相關系統2. 系統異常資訊不足導致升級失效3. 系統數量與種類繁多;並以此提出解決方案1. 系統本身應主動的偵測、收集、告警2. 系統主動告警可透過SLAM、E-Mail與簡訊3. 顧客可隨時查看事故的分析與事故處理進度以及提出服務升級請求。 問題管理流程失效主要分佈在問題分析階段,主因可分為1. 專業或背景知識不足,害怕分析出錯2. 系統龐大且複雜,導致SOP失效;改善的方案為1. 快速建立協同分析、編輯與審查平台2. 建立SOP執行測試與防錯機制。
Case study of IT department of the manufacture partially adopted the ITIL or ISO20000 as service management standard and use of Lean Six Sigma tools and DMAIC methods to analyze and improve the Incident management processes in Help desk and Problem management processes of applications' developing. The content described in detail how Lean Six Sigma tools in DMAIC various stages of implementation. In the Definition stage, define the problems by Kano analysis, SIPOC, VSM(Value Stream Map) and the Little’s Law, Measurement stage used the statistical software to measure the actual data help to establish a baseline, next in the Analysis phase adopted the PFMEA(Process Failure Modes and Effects Analysis)、cause-and-effect and Multi-voting for the process of failure analysis, the phase of Improving ,thinking about how to improve the process with IT systems supporting .Finally ,the Control phase, develop control plan which use mistake proof and SPC(statistical process control) to prevent and monitor the failure issues to reduce variation and risk of service management processes . Help desk's incident management process mainly failure by upgrade phase when incident occur, and the reasons can be divided into: 1. Service personnel unfamiliar with related systems. 2. System abnormalities upgrade failure due to insufficient information .3. The number and variety of the systems .The solutions list like below: 1. The system itself should take the initiative to detect, collect, alert.2.Alarm System initiatives through SLAM, E-Mail and SMS.3. Customers can keep track of incidents and accidents, and the progress made to upgrade service request. Problem management process failure mainly in the problem analysis stage, the main reason can be divided into 1.Professional or insufficient background knowledge, fear of error. 2. The system large and complex, leading to failure of SOP; to improve the program by 1. Quickly create synergy analysis, editing and review platform 2. To establish SOP testing and error prevention mechanism.