BIFPET methodology versus PERT in project management: fuzzy probability instead of the beta distribution

Abstract

Research throughout the years has centered on the use of the beta distribution to model variable activity times in the Program Evaluation and Review Technique (PERT). Justification for using a weighted average of optimistic, most likely and pessimistic times is based mostly on the beta distribution's ability to handle skewness and its ease of use for computing the mean activity times. This paper presents two variations of a fuzzy probability based model for project management. The Belief in Fuzzy Probability Estimations of Time (BIFPET) model uses human judgment instead of stochastic assumptions to determine project completion titties. Following a literature review of PERT critiques, background information is provided for BIFPET. Next, a foam block production machine project is described and solved based on three estimates of time for each activity. A variation of BIFPET that uses ranges on these time estimates is presented and the case is solved for fuzzy expected completion times. The results are compared to those derived by using PERT and benefits of the BIFPET approach are detailed. The paper concludes with a description of our ongoing research initiative in the area of fuzzy probability applications to project management.