Challenges Hindering the Promotion of Machine-Learning Techniques in the Construction Industry

Machine learning (ML) is transforming how we design, build, run, and manage buildings and infrastructure as it slowly gains traction in the built world. Machine-learning approaches have the potential to significantly improve operations, increase energy efficiency, raise occupant comfort, and boost sustainability in general by utilizing the power of data and sophisticated algorithms. However, integrating machine-learning techniques with already-in-use technologies and processes in the construction industry can be challenging. A smooth integration may encounter difficulties due to interoperability problems, fragmented data sources, and legacy systems. Therefore, it is crucial to assess the challenges preventing the use of machine-learning techniques in the built environment. The study has a survey-style design. One hundred and ninety (190) of the two hundred and fifty (250) questionnaires that were sent to stakeholders and professionals in the construction industry were returned and declared appropriate for the study. Percentages, mean item scores, standard deviation, and Kruskal–Wallis were used to examine the data that had been gathered. The findings show that the absence of industry-academic collaboration, a lack of tools to assist the application of ML, and model flexibility are the main challenges preventing the development of machine learning. Collaboration between academic institutions, technology suppliers, and industry partners is necessary to overcome these challenges. The study came to the conclusion that initiatives should concentrate on fostering industry–academic partnerships, advancing interdisciplinary education, improving the interpretability and transparency of machine-learning models, proving the value proposition, and creating ethical frameworks.