Implementing Bayesian Networks in Delivery Selection for Highway Construction Projects

Selecting an appropriate project delivery method (PDM) is a complex decision that typically involves assessing many variables and the relationships between them. One of the main challenges in PDM selection is to accommodate changes in the relationships among variables when more information becomes available as the project is involved during the project development process. The objective of this paper is to develop a Bayesian decision-support model for PDM selection in highway construction by using cost performance data of 177 highway projects delivered by design-bid-build, design-build, and construction manager/general contractor. The conditionally probabilistic inferences associated with these delivery methods were evaluated to determine an appropriate PDM. The model was validated using the k-fold cross-validation and case projects. This paper contributes to the body of knowledge by demonstrating the implementation of Bayesian networks as a data-driven decision-making tool in the construction industry and supporting transportation agencies in their PDM selection process.