Intelligent systems approach for predicting defects in housing construction

Abstract

Building defects are a phenomenon occurring through the normal wear and tear of building use or resulting from natural climatic exposure effects. These defects are commonly unavoidable. The majority of building defects, however, occur due to malfunction and hence are of a premature nature. These types of defects are avoidable and relate to design and construction problems, in addition to poor quality maintenance work and general negligence and abuse by users (C. Briffet and Leng Aug Aik, 1991). Such problems have been estimated to represent 95% of all defects, with an annual cost around #1000 million. It is therefore clear that the elimination or substantial reduction of such defects provides a strong financial incentive to improve design and building processes. The paper identifies current methods employed by the construction industry to overcome and reduce the number of defects occurring on sites and illustrates the pitfalls of the current techniques. In addition, identification of future intelligent systems alternatives to the current techniques in avoiding building defects are presented.