<![CDATA[ In the mining business of Kogi State, the safety system has been weak, thus exposing the workforce to severe occupational hazards. In this research, Artificial intelligence (AI) was used to forecast work-related harms and aid in advance safety planning. Researchers compared data gathered from 1,200 miners and environmental sensors (PM2.5, CO, noise, temperature, and vibration) with institutional accident records from five years (2019-2024) using supervised models, including Random Forest, Support Vector Machine (SVM), Artificial Neural Network (ANN), and Decision Tree. Random Forest reached the highest accuracy of 91.3%, precision of 0.92, recall of 0.87, F1-score of 0.89, and AUC-ROC of 0.94. Important predictors included exposure to PM2.5 (0.118), use of PPE (0.105), noise (0.098), job role (0.093), and levels of CO (0.089). Excessive hazard levels: PM2.5: 109ug/m+ (WHO standard: 25ug/m3), noise: 89.2dB (OSHA standard: 85dB). There was the greatest risk to afternoon shifts and underground drillers. It is the first validated AI-based model of mining safety in Nigeria, which allows for making the risk forecast in real-time. The research suggests the mandatory environmental surveillance, the adoption of AI systems, and predictive analytics in occupational safety policy at the sub-Saharan Africa level.]]>
Copyrights © 2025
