Construction sites face persistent safety challenges, often resulting in severe injuries or fatalities. In the U.S., these concerns are heightened due to the high volume of construction projects and complex working conditions. This study analyzes 1,963 OSHA-reported construction incidents in the Southeastern U.S. using five Machine Learning (ML) techniques to classify fatal and non-fatal incidents. Random forest and decision trees achieved the highest accuracy, with random forest outperforming all models. Feature importance analysis identified age, height, occupation, and event type as key predictors of injury severity. These findings highlight the potential of ML models in providing predictive insights for proactive safety management. By identifying high-risk factors associated with severe injuries, this research contributes to data-driven safety interventions and policy improvements aimed at reducing incident rates and enhancing targeted risk assessment and preventive strategies on construction sites.
