Item-level quantity-based preliminary cost estimating system for highway structures and miscellaneous construction

The intent of this research was to improve the preliminary cost estimating procedures at the Texas Department of Transportation (TxDOT) and develop a computational tool that simplifies the process of preparing preliminary cost estimates. An item-level quantity-based approach was proposed for this research study. The primary motive for this approach was to segregate the impacts of unit prices from the estimates in order to improve accuracy and reduce the difficulty in updating the estimating models. Furthermore, this approach enables continuous tracking and control by initiating quantity estimates at the outset. This research study was focused on the development of statistical models for predicting item-level quantities based upon available basic project parameters. Major work items, defined as those accounting for 80 percent of the total cost, were identified for each major project type in TxDOT; and either a statistical prediction model or a fixed cost percentage was developed for each major work item based upon historical data. These quantity models and fixed cost percentages were integrated into a database management system, Preliminary Item-Level Cost Estimating System (PILCES), to estimate costs at both the item-level and the project-level. This dissertation discusses the statistical quantity models for structures-related and miscellaneous construction-related major work items. The design concept and implementation of PILCES are also described in this dissertation.