No Scheduling Problem but Rather a Rescheduling Problem: Measuring Robustness of Schedule Network in Construction Projects

Abstract

The available slack-based methods for measuring the robustness of schedule networks do not pay sufficient attention to the structural characteristics of such networks nor do they provide a comparison point for measuring the sensitivity to changes. This research fills these gaps by addressing a core research question: How can an integrative measure of schedule robustness be developed? To answer this question, this paper performs a centrality analysis to capture the structural characteristics of the schedule network. Further, it draws on the concept of perfect uniformity to establish a baseline for measuring the degree of schedule sensitivity to disruptions. Finally, it uses entropy theory to develop an integrative measure of schedule robustness by combining two classes of robustness measures: quality and solution robustness. An experimental analysis is undertaken to evaluate the statistical performance of the new measure. The validation results show a high inverse correlation between the proposed measure and a cost-based indicator of schedule performance; in turn, proving the better performance of the new measure compared to the two widely discussed and quoted slack-based measures of robustness. This can be ascribed to the ability of the new measure to simultaneously capture various attributes of a schedule network. The results of this research can assist decision makers in making informed decisions in the appraisal and monitoring phases of construction projects. This research provides the opportunity for future studies to extend the proposed method by incorporating resource-constrained and cost-related attributes of project activities into the formulation of schedule robustness.