Digitization and Digitalization of Cost and Duration Estimation in Philippine Public School Building Projects Using Artificial Neural Network

Digital technology and innovation integration in the construction sector have demonstrated their effectiveness in boosting workplace productivity. The construction business has benefited from the use of computer-aided design (CAD) software, which significantly decreased project cost and length while maximizing efficiency. Estimating cost and duration (CD) during the pre-construction phase of a project is critical due to the limited available data during this phase. Utilization of machine learning (ML) methods, such as artificial neural network (ANN), are now evident to solve complex problems even with minimal data available. At the pre-construction stage of a construction project, this study created a two-phase ANN model that can forecast the project cost and duration for public school buildings in the Philippines based on the data provided. The ANN model was initially composed of multiple sub-models that will form a final comprehensive model, wherein the output of the sub-models will become the input of the final model. The sub-models were based on the determined Work Breakdown Structure (WBS) of the project, namely earthworks, structural, finishing, plumbing and sanitary, electrical, and mechanical. Based on the excellent performance of the input parameters (IP), its structure gave a completer and more understandable ANN model. The contribution of the input parameters available at the preconstruction stage of a construction project to the overall project cost and overall project duration was also determined using sensitivity analysis (SA).