Use of DM techniques in earthworks management: a case study

Abstract

In most transportation infrastructure projects, earthworks are generally associated with the highest percentage costs and durations. Because these tasks are reliant on heavy machinery and repetitive tasks, they are strongly susceptible to optimization. This paper presents a study based on the application of artificial neural networks (ANN) on data originated from an actual earthwork construction project. Results show a good adjustment to the data while emphasizing the importance of optimal equipment allocation throughout the construction site. Finally, the architecture of an intelligent earthwork optimization system is presented, combining both data mining (DM) and modern optimization technologies, among others, to support equipment distribution optimization in earthwork projects.