Enhancing reliability of project duration forecast

Abstract

Reliability of Project duration forecast depends mainly on the accuracy of the activity duration estimates. During project implementation, many uncertain but predictable variables dynamically affect the activity duration. Currently, this impact is considered intuitively, the effectiveness of which depends on the skill of the scheduling engineer. This research identifies the need for a model which can simulate the project environment to incorporate the combined impact of the uncertainty variables in the activity duration estimates. -- Significant uncertainty variables are discussed along with random sampling procedures for quantifying their impact. A computer model “PRODUF” simulates and incorporates the combined impact in activity duration estimates. PRODUF computes a distribution of duration estimates for each activity. Monte Carlo simulation on such distributions gives a probability distribution for the tactical plan completion time. This is used at every progress review to develop project duration forecast from the strategic plan. -- The PRODUF model has many applications. It can be used to generate forecast of project duration with associated probability. The sensitivity of the activities to an uncertainty variable can be measured. As an extension of PRODUF, an improvised gaming model in project scheduling can be developed. Application of PRODUF to a hydro project highlights its practicality.