Parametric Cost Estimating Models for Resurfacing Projects

State DOTs are constantly faced with the challenge of producing accurate cost estimates for highway maintenance projects. Inaccurate estimates can result in financial losses, delays, disputes, and even cancellation of planned projects because of insufficient funding or waste of available funding if it is not used. If a project is approved with an erroneous estimate, budget constraints during execution may cause disputes, delays, quality compromises, and the need for additional funding that strains public resources. Legal disputes over budget overruns can further inflate costs and damage trust among stakeholders. Therefore, accurate methodologies and tools for cost estimation are essential. The goal of this study is to develop parametric models for estimating maintenance project costs with a minimal set of information available at the early phase of project development. To achieve this goal, the study analyzed 11 years of maintenance cost data from the Tennessee Department of Transportation (TDOT) that included various information, such as number of lanes, type of work, treatment type, and location (e.g., county and region). Eleven state-level, eleven region, and nine county-level regression models for maintenance projects were developed from the analysis. The R2 values were calculated to assess the performance of the models. Moreover, the model was validated using an independent dataset, demonstrating reasonable predictive accuracy with an overall mean absolute percentage error (MAPE) of 14.6%. A spreadsheet-based tool was developed to implement the models. The methodology and tool are expected to aid TDOT engineers in estimating maintenance project costs quickly and accurately at the early phase of project development with minimal information.