The Florida Department of Transportation (FDOT) recently conducted a study to develop agency cost factors for maintenance, rehabilitation, and repair (MR&R) of bridges, for the state's implementation of the AASHTOWare Pontis Bridge Management System. The study analyzed three information systems on maintenance and construction history available at the FDOT. Pontis defines many combinations of bridge element, environment, condition, and feasible action, resulting in more unit cost values than can be estimated from the available data. To simplify cost determination, a classification scheme (MR&R Action Subcategory) was devised to group the Pontis combinations. Statistical analysis was performed on the raw cost data to extract useful numbers from the available data sources. The results were supplemented with the judgment of an FDOT panel of experts. Although the panel found the analysis of historical data to be useful for 70% of the actions considered, only 50% of the actions had a statistically significant sample, and only 15% were usable directly from a statistical analysis. The main issues were differences in data collection units and differences or uncertainties in the scope of recorded actions. The study recommended improvements in the automated systems to enhance the ability to use the data in future updates of the agency cost model.
Abstract