關鍵鏈專案管理執行計劃之研究

Abstract

關鍵鏈專案管理(Critical Chain Project Management, CCPM)是1997年由Eliyahu M. Goldratt所提出的限制理論(Theory of Constraints, TOC)管理方法之一。Goldratt延續其限制理論的主張，認為專案環境中，唯有專案如期達交才能使企業獲得產出(throughput)而達到良好的績效表現。因此，造成專案無法如期達交而績效不佳的限制因素，便是專案環境的瓶頸。由於專案是由一系列的相依任務和資源所組成，因此專案交期受到任務相依和資源相依的最長路徑─關鍵鏈─的影響，故由關鍵鏈長度所決定的專案時程便是專案管理的瓶頸，即企業改善專案管理績效的槓桿點。 近十年來，關鍵鏈專案管理所闡揚的管理思維在實績的佐證下，已陸續證實其有效性。然而，現有相關研究偏重關鍵鏈排程及緩衝管理之技術議題探討、針對關鍵鏈持正反兩方的辯論、或關鍵鏈與要徑法的比較，對於如何執行甚少著墨。 2006年Goldratt提出策略與戰術圖，將其概念層層展開，發展為架構更明確的改善方案，其中針對專案交期績效的改善提出了六個改善方案。不過，由於此六個改善方案仍屬於較高層次的中繼目標，距企業實際導入時賴以依循的步驟仍有差距，本研究以策略與戰術圖為主軸，進一步發展與研究出適用於導入的執行任務和執行計畫，讓有意導入關鍵鏈專案管理的企業有參考的依據。

Critical Chain Project Management is one of the applications of TOC (Theory of Constraints) proposed by Eliyahu M. Goldratt in 1997. Continuing his assertion, Goldratt claims enterprises can gain throughputs and achieve good performance only when they can deliver projects on time for project environments. Therefore, the restricted factor causing projects delay is the constraint of project environments. Projects are composed of series of tasks and resources. Accordingly, the project due date is affected by the longest path, called “Critical Chain”, composed of dependent tasks and resources. The project schedule determined by the length of Critical Chain is consequently the constraint of project management which is the leverage point for enterprises to improve their project performances. In this decade, CCPM has proved its practicality by increasing successful implementations. However, the tendencies of present researches are technical issues of critical chain scheduling and buffer calculations, debates between supporters and antagonists, and comparisons between CCPM and CPM (Critical Path Management). There are seldom discussions about execution of CCPM. In 2006, Goldratt proposed “Strategy and Tactic Tree (S&T Tree)” to deploy his concepts into executable scheme. There are six sub-topics regarding high project due date performance. The author considers these six sub-topics are intermediate objectives which deserves further studies. This thesis deploys executable tasks and constructs the implementation plan based on S&T Tree for those who feel interested.