Artificial intelligence-driven construction: Leveraging machine learning for predictive modelling for commodities estimation

The integration of digital technologies like Artificial Intelligence (AI) and Machine Learning (ML), in construction practices can enhance productivity and reduce costs. ML is crucial in estimating construction costs, enhancing safety and reliability in projects like buildings, bridges, roads, airports, canals, and railroads. The price of commodities in construction includes all sorts of energy commodities and raw materials. This research analysed data from oil and gas and petrochemical projects in Egypt, Saudi Arabia, Qatar, UAE, and Oman from 2005-2017, focusing on daily reports and 460,172 records on piping erection, manpower, equipment, heat index, and delays. This research used predictive ML modelling to analyse daily piping erection rates in Egypt and Qatar under different conditions. Egypt's rate is high, but Qatar's implementation of HSE restrictions and high heat index significantly reduces it. The findings emphasised the importance of regulatory compliance, environmental conditions, and working conditions in optimizing production output in construction projects. The deep learning ANN model was found to be the most effective in predicting piping erection per day, outperforming other ensemble methods and suggesting better capture of the dataset's complexity and non-linearity.