Assessing the Potential of Machine Learning in Construction Safety: A Systematic Review

Many fatal and non-fatal occupational incidents have been reported in the construction industry globally. While many effective techniques have been developed in recent years to reduce the number of tragic accidents on the jobsite, there are still high rates of accidents. To improve management's decision-making processes, machine learning (ML) has drawn significant attention due to its ability to analyze large quantities of data to identify potential hazards on construction sites. Compared with traditional techniques, ML is able to handle large datasets and, through the use of different algorithms, can quickly analyze them to produce more accurate interpretations. Although machine learning has been identified as a useful statistical method for improving the decision-making process, little systematic research has been carried out on the correlation between machine learning and construction safety. To address this gap, this study was developed to explore a systematic review of the effect of machine learning on the safety of construction work sites. In examining and reviewing research two databases, it can be identified that ML techniques can be a powerful leverage for discovering useful knowledge from large datasets to perceive relationships, trends, and correlations. This study provides a contribution to the research area of ML applications in enhancing construction safety.