Predicting the preliminary engineering duration of bridge projects

Preliminary engineering (PE) is a critical preconstruction phase of infrastructure projects, encompassing the initial planning and design processes necessary for successful commencement and completion of a project. The duration of PE influences both the cost and schedule of any civil infrastructure project. Numerous studies targeted at construction duration estimation exist, but few address strategies to estimate PE duration. This study sought to identify the project factors that influenced the PE duration of bridge projects and to determine if regression modeling with those factors would provide a feasible tool for PE duration estimation. Some 416 bridge projects conducted by the North Carolina Department of Transportation (NCDOT) served as the analysis sample. Among the project factors investigated, four descriptive factors significantly influenced PE duration: geographical area of the state, classification of the required National Environmental Policy Act (NEPA) document, bridge construction scope, and responsible party preparing the NEPA document. For validation purposes, the authors randomly extracted 60 of the 416 bridge projects before regression modeling. Regression models achieved a mean absolute percentage error (MAPE) of 21% when applied to the validation set. Comparatively, using the mean PE duration of 66 months as a single point estimator, the MAPE was 22% for the same validation set. Based on these results, the project team recommended that the NCDOT can use the regression equation to estimate PE duration. It is also clear that this is a difficult problem that deserves further research attention.