This paper presents a study based on the application of cascade ANN prediction models to an earthwork construction database. Results show not only a good adjustment to the data, but also the influence of each earthwork construction process on the work rate of the global production line. Furthermore, the obtained results emphasize the importance of optimal resource allocation and management throughout earthwork construction phases. Following this framework, the study concludes with a demonstration of how the developed models can be integrated into a more complex system in order to pursue that purpose.