Using Earned Value Management with exponential smoothing technique to forecast project cost

Abstract

Earned value management (EVM) is developed for cost management and is widely used to evaluate and forecast project cost. Through earned value management, we can calculate the deviation between cost and schedule, as well as performance indicators such as cost performance index (CPI), schedule performance index (SPI) and duration performance index (DPI), and then predict project cost, schedule and duration. However, it is often not comprehensive to rely only on EVM for project prediction, because the project in the next testing period is not necessarily invariable and similar to the previous activities, but dynamic and flexible, and has some correlation with the previous project activities. Therefore, the research method of this paper is to combine the well-known earned Management (EVM) with the mature exponential smoothing technique to simulate the actual data of 32 simulation projects and predict the final cost of the project. After comparing the prediction results with the results of tens of thousands of simulation experiments, the final research results show that the combination of cost performance index (CPI) from earned Management (EVM) and simple exponential smoothing technique have promising performance for the final cost estimation of the project.