The Federal Highway Administration requires State Transportation Agencies to have a formal method to estimate contract time for highway construction projects. To meet this requirement many states use an integrated scheduling system to estimate project durations based on assumed productivity rates and generic job logic. The current work investigated the accuracy of two of these systems found that both systems accuracy in predicting the duration of Kentucky Transportation Cabinet projects was greater than +200%. In response to this poor accuracy, a parametric project duration estimating tool was developed based on a multivariate regression analysis of bid item quantities and engineering and construction estimate. Five regression models were develop to estimate contract time for large projects (great than $1,000,000) based on key bid item quantities; limited access (+22% median error), open access (+35% median error), new route (+55% median error), bridge rehabilitation (+77% median error), and bridge replacement(+17% median error). It was not possible to develop a parametric estimating tool for predicting the duration of small projects (less than $1,000,000) as it appears that the duration of small projects is determined by factors outside of bid item quantities.
Digital Object Identifier
Taylor, Timothy R. B.; Goodrum, Paul M.; Brockman, Michael; Bishop, Barry; Shan, Yongwei; Sturgill, Roy E.; and Hout, Kevin, "Updating the Kentucky Contract Time Determination System" (2013). Kentucky Transportation Center Research Report. 320.