IT Project Planning Method Based On Markov Decision Process

With the rapid growth in domestic IT industry, IT project management is becoming the most important problem in the research and development of IT projects. One critical issue in IT project management is the progress planning, which is also the Achilles' heel in the IT project deploying practice. Currently, there lacks formal and efficient progress planning techniques, and the project managers often fail to have an in-depth understanding of the key factors that affect the progress. Moreover, the IT project itself has some characteristics such as dynamics, randomness, complexity and etc. Therefore, there are a lot of IT projects which cannot be finished in time.Recently, there is extensive research work addressing problems in IT project management. Among these approaches, quantitative methods have attracted much attention. Particularly, for the progress planning, statistical methods have been proven powerful because of their capability to describe the dynamic random process. The statistical modeling, simulation and planning methods are being widely applied in practical IT project management.In this paper, we focus on the progress planning of IT project using statistical modeling methods. We surveyed current progress planning methods in IT project management and compared their strengths and weaknesses. A multiple-phase optimization method for IT project progress planning is presented to meet the uncertainty of IT project process and dynamic of project resources. The IT project process is divided into a series of phases according to the critical path. By modeling each phase as an independent Markov decision process, we use dynamic programming method to solve optimizing planning problem and minimize cost in each phase. The proposed method is efficient and useful for the progress planning of IT project.