Conceptual estimate is often misleading because of the paucity of available information. An accurate range of preliminary cost estimates with desired confidence bounds is essential to determine the viability of a project. In particular, better prediction of future costs using historical data could permit better decisions and eliminate errors. This research aims to develop a system architecture for a preliminary cost estimation system that toggles project input information, predictive item-level quantity, and segregates unit price of highway projects. Item-level quantity models at project outset were developed using cost data that are stored in Texas Department of Transportation (TxDOT) Design and Construction Information System (DCIS). The basic estimating parameters were derived from statistical analyses and then documented so that the data with the system could be continuously updated. The statistical models are embedded in a Web-based relational database management system (RDBMS). These models are then used for computation and data storage purposes. The system has proved to reduce cost estimates variability and to eliminate the need of ad-hoc housekeeping by centralized maintenance. This application was developed by using open-source software including PHP (PHP Hypertext Preprocessor), APACHE server, and MySQL database server with the usage of browser interface that further reduces the cost of development and maintenance.
展开▼
