On the Estimation of Mean Remaining Activity Duration for an Ongoing Project Activity in PERT

Abstract

Studies on the probability distribution of Project Activity Duration have been widely discussed in the context of PERT (Project Evaluation and Review Technique) in situations where activities have not yet commenced. The activity duration is then estimated based on the expected value of various assumed probability distributions, predominantly, the Beta distribution. As the project proceeds, an activity in progress will affect its expected completion time (duration) through the change of probabilities. Failing to recognize this often leads to the misconception that the remaining activity completion time is the original estimated activity duration minus the elapsed activity duration. This paper develops the conditional partial expected value which can adequately estimate the mean remaining activity duration as the project proceeds. Mathematical formulae for these estimates are derived from Uniform distribution, Triangular distribution, Beta distribution, and Normal distributions. These formulae can be easily incorporated (programmed) into existing project management software.