An expert systems approach to cost estimation for steel structures using stochastic methods

Cost information is important to all parties in the steel fabrication industry: owners; engineers; and fabricators. The industry is driven by economic forces that encourage the efficient use of material, labour and equipment making cost estimation a vital tool in every decision. Unfortunately, these same forces hinder accurate estimation by imposing economic constraints that reduce the amount of time, effort and money that can be committed to the process; therefore, the development of new and improved cost estimation methods is encouraged. This thesis attempts to demonstrate the feasibility of applying expert systems technology as a means of estimating the fabrication cost of steel structures. It considers the historical factors limiting current methods of cost estimation and the strengths and weaknesses of computer technology in order to develop a new method of cost estimation that provides valuable information which is impractical to obtain using traditional methods. It introduces the concept of stochastic estimation which may be used to evaluate the variability of costs and assess accuracy. The new method is designed for efficient use of computer capabilities and is intended for development as a tool to aid engineers, estimators and fabricators. Current methods of cost estimation are unsuitable for application in an expert system because they rely heavily on subjective judgement and are bound by constraints imposed by the burden of manual calculation. These constraints are no longer valid now that the industry has access to powerful personal computers that are capable of performing sophisticated calculations and manipulating large amounts of data. The introduction of expert systems technology has provided an opportunity to examine traditional cost estimation methods and to propose new methods that are better suited for use in a computer-based environment. The development of the new estimation method and its implementation as an expert system promises many benefits to the steel design and fabrication industries. It focuses the power of computer technology on the estimation problem. It provides greater access to accurate estimation, thereby improving the basis of important economic decisions. It can also be an important tool for the training of new cost estimators and allows the efforts of experienced personnel to be directed into more profitable areas. Moreover, an estimate derived by an expert system is consistent and unbiased making it better suited for comparison than one compiled by human estimators who may be influenced by personal preferences and external factors.Applied Science, Faculty ofCivil Engineering, Department ofGraduat