Conceptual cost estimation of road projects in Ethiopia using neural networks

Abstract

The rapid technological changes and advances in all business sectors strongly impose construction managers to facilitate their work through advanced software applications available to simplify different tasks. A research has shown that inaccurate cost estimation method is among the cause of cost overrun in the Ethiopian construction industry. This article presents conceptual cost estimation model developed using neural networks for federal road projects of Ethiopia. The conceptual cost estimation model developed has a mean absolute percentage error of 32.58% trained with only 48 exemplars. If the model is developed with the provision of enough data set to represent the road project with all-state of affairs, it is forecasted to improve the estimating capability of financers, employers and consultants. In addition, a friendly user interface is built to enable the utilization of the model developed easily and this article presents this interface with an example of actual road project in Ethiopia. The findings of this study indicate the prospect of application of neural network for cost estimation during early phase of the project development for Ethiopian road projects.