Identification of Factors Significantly Influencing the Labour Productivity of Formwork Systems Using Artificial Neural Network

Abstract

Predicting labour productivity accurately in critical activities like formwork erection would enable management interventions to improve the site situations especially in the context of high rise building construction. In this study, Artificial Neural Networks (ANNs) were employed to model and predict three categories of formwork erection activities - aluminium formwork, horizontal formwork and vertical formwork. 16 input factors were identified and a total of 19,344 data points from 42 construction sites all over India were used to train and validate the ANN models. The developed models show a high degree of accuracy in predicting the productivity on sites. The models also give major insights into the factors affecting the productivity of formwork related activities. The adverse effects of some factors like the number of workers on the site were also discussed. The study indicates the usefulness of data-driven techniques for prediction of labour productivity of formwork activities on Indian construction sites. Health hazards related to formwork assembly/preparation are identified.