Modeling Uncertainty in Evaluating the Project Performance

Abstract

Monitoring and controlling project performance is a key to project success. Regular measurement of the project performance results in an early detection of the project deviation from the baseline plan and provides the opportunity to take corrective actions. Earned value management (EVM) is a well-known technique that assist project managers to evaluate and monitor the project performance, and to forecast the completion status of the project. EVM relies on three key variables of planned value, actual cost and earned value to evaluate the project status. While in reality the project performance data usually are from people's judgments, and hence, they carry levels of uncertainty, EVM variables are considered deterministic. Perceiving the uncertainty helps to measure the performance and progress of the project more realistically. In this paper, we propose a fuzzy-based EVM technique, which models the uncertainties in the project performance data. The new technique suits complex projects, particularly, where measuring the actual cost of the project activities are uncertain and inexact, or where the completion percentage of the activities are uncertain. By illustrating a small hypothetical project, we show applicability and usability of the new technique.