Author(s)

J. Christian & A.N. Pandeya

Abstract

An attempt has been made to predict the total ownership costs of two different types of facilities. The factors influencing these costs were critically examined and analyzed. The process of knowledge elicitation was divided into two phases, the compilation and analyses of annual historical operating and maintenance costs, and knowledge acquisition of various factors which affect these costs. Three methods were then used for the analysis and prediction of costs: neural networks, regression analyses, and random deviation detection. A prototype expert system was developed using the data and knowledge acquired. 1 Introduction Facility ownership is becoming increasingly more expensive due to the rising costs of design, construction, operation and maintenance of facilities.

Keywords