ABSTRACT
Attempts at modelling the cost of buildings have hitherto attempted to explain cost as (simple) functions of different measurements of the finished building, element or component, and therefore by implication seeing building as a single discrete step. Such an approach ignores the fact that construction is a process consisting of separate but dependent physical activities occurring over a period of time and subject to uncertainty in terms of, inter alia, cost and duration. It does not explain when and how costs are incurred (inexplicability), disregards interrelatedness (unrelatedness), and ignores uncertainty (determinism). To overcome the above weaknesses, a process-based modelling approach is suggested. PERT-like networks can be used in a structured manner to model cost by representing expenditure as it occurs during construction activities. The proposed system will use A.I. (artificial intelligence) based on Nebula Theory to link subnetworks representing the construction of different elements (e.g., foundations, basement, etc.) to create a complete, representative network for the entire project. For each element, different sub-networks representing different designs (different in type, and not size or similar parameters) will be available, enabling the user to explore different design/construction alternatives. Being computer-based, the proposed system will use a data-base containing cost information, sets of sub-networks, different default values, and definitive functions.
