Christopher Preece

Automated estimation of completion time for dam projects

An accurate estimation of construction duration for the successful completion of infrastructure projects is as important a task as keeping a project within the stipulated budget and quality. This research aims at developing an automated approach for estimating the duration of construction for dam projects by virtue of artificial neural network (ANN). Through literature searches and expert interviews, variables that critically influence the construction of dam projects were identified. A wide range of data on Iranian dam projects was used to develop seven ANN models. Different datasets were used to achieve the best performance by assessing the RMSE and R2 values as the reliability and validity indicators. Finally, a web-based automated prototype was developed and validated in order for stakeholders to estimate the duration of dam projects based on the ANN method and benefit from better infrastructure project management practices.