Up-front cost estimating of change orders

Estimating the cost of change orders includes the evaluation of impact costs, in addition to the conventional direct and indirect costs. Impact costs can be divided into time related, and productivity related. This paper presents an automated system, developed to assist members of project teams in estimating productivity losses. The system incorporates a set of newly developed neural network models. These models were developed using actual cases from projects constructed in Canada and the US. Two numerical examples are analyzed in order to demonstrate the use of the developed software and to illustrate its capabilities.