On uncertainty and sensitivity analyses in project duration based on dependency information
Abstract
The dependency structure matrix (DSM) shows the interdependency between activities, and it has been shown to be useful in the estimation of complex projects’ durations. The estimate of project durations is based on activity durations, their interrelationships, and the permitted level of overlapping, all of which are represented by DSMs. However, these variables have individual uncertainties that generate overall uncertainty in the project duration. The objective of this work is to show that uncertainty analysis and sensitivity analysis are essential parts of analyzing the uncertainty in project scheduling. Specifically, this work shows how to perform sensitivity analysis in project scheduling using DSM, how to reduce the number of input variables with uncertainty for sensitivity analysis, and how to identify input variables whose control of uncertainty reduces the uncertainty of the project duration. An example is used to explain the methodology, and a case study is used to show the usefulness of sensitivity analysis and uncertainty analysis.