Neural Network Method for Instrumentation and Control Cost Estimation of the EPC Companies Bidding Proposal

AbstractCost estimation of EPC companies bidding proposal plays an important role in determining eligibility for funding of a project. However, many variables especially indirect cost in cost estimation is subject to uncertainty. This level of uncertainty will arise because of increased imperfect knowledge about indirect cost variables. Consequently, the total cost offered maybe too high so the client will reject or maybe too low so the profit margin will decrease or even turn to a loss profit. This paper proposed artificial intelligence computation based on the Back Propagation method in Neural Networks. This method is used to model all of variables consist of direct cost and indirect cost based on previous typical projects to reduce uncertainty. This method is applied specifically to estimate the costs of instrumentation and control discipline. In the future, this method could be developed and applied for other discipline such as process, piping, mechanical, civil, electrical and others, so that the total uncertainty of project cost estimation could reduce significantly