Skip to main content

SDSU research recognized at construction engineering and management conference

Construction site
Selecting a project delivery method for construction projects is one of the most important decisions of the planning phase. However, very little research exists on the decision-making process for selecting a delivery method. SDSU assistant professor Phuong Nguyen's research explores how probability theory can better inform these decisions in an award-winning paper.

How can probability theory help improve the planning process for highway construction projects?

This is the basis for South Dakota State University assistant professor Phuong Nguyen's research, which was recognized as "Best Paper" for the "Contract, Delivery and Legal Issues" track at the American Society of Civil Engineers' 2024 Construction Research Congress.

"Selecting an appropriate project delivery method for highway construction projects is a complex decision that typically involves assessing many variables and the relationships between them," Nguyen said. "One of the main challenges is to accommodate changes in the relationships among variables as the project moves forward. The objective of this paper was to develop a Bayesian decision support-support model for project delivery method selection."

During the planning phase of construction projects, decision-makers must select one of three project delivery methods: design-bid-build, design-build or construction manager/general contractor. Each method has pros and cons and are optimized for different types of projects that are dependent on budget size and scale, among other factors. 

Phuong Nguyen
Phuong Nguyen 

"Past research has shown the selection of an appropriate delivery method as one of the most important decisions in the project planning stage," Nguyen said.

Despite its importance, there is very little research on the decision-making process behind selecting a project delivery method. To fill in the literature gap, Nguyen proposed using a type of probability model — Bayesian networks — to help better inform these decisions for more effective projects.

For his proposed model, Nguyen collected data from 177 completed highways projects across the United States. He extracted various data points from each project, which were inputted into a decision-making model. The outcomes of the model returned true or false probability values for each delivery method.

"The proposed data-driven Bayesian model can construct a causal network to visualize the probabilistic relationships between variables in project delivery method selection and update the probabilities of the selection outcomes when more information becomes available," Nguyen said.

The key to the Bayesian model, as Nguyen notes, is its ability to consistently reflect the most up-to-date information. This allows for better decision-making throughout the planning phase.

The applications of Nguyen's work will assist transportation agencies in selecting appropriate project delivery methods for future highway construction projects. With historical data and new information for the given project, Nguyen's model will lead to effective decision-making in the planning phase of large-scale construction projects.