Project Makespan Estimation: Computational Load of Interval and Point Estimates

TL;DR

This paper reviews the computational load of different project completion time estimation procedures, highlighting the influence of activity probability models and comparing the efficiency of interval and point estimates.

Contribution

It introduces a multiple polynomial regression model for interval estimation procedures and analyzes the computational tradeoffs involved.

Findings

Accuracy of activity probability models significantly affects estimation quality.

Monte Carlo simulation has higher computational time compared to other methods.

Computational time is generally manageable, allowing for simpler estimation procedures.

Abstract

The estimation of project completion time is to be repeated several times in the project planning phase to reach the optimal tradeoff between time, cost, and quality. Estimation procedures provide either an interval or a point estimate. The computational load of several estimation procedures is reviewed. A multiple polynomial regression model is provided for major interval estimation procedures and shows that the accuracy in the probability model for activities is the most influential factor. The computational time does not appear to be an impeding factor, though it is larger for MonteCarlo simulation, so that the computational time can be traded off in search of a simpler estimation procedure.