A NEW PARADIGM OF BIG DATA–BASED RISK ASSESSMENT: EVIDENCE FROM A METAL MINING COMPANY

Abstract

In recent years, mining companies have been increasingly exposed to a wide range of environmental, social, economic, and technological risks, which have adversely affected operational sustainability. Conventional risk assessment approaches have limited capacity to incorporate real-time data and to adapt to dynamic operating conditions. In contrast, artificial intelligence and machine learning methods offer new opportunities to address these shortcomings. This study applies GRU, BiLSTM, XGBoost, and Random Forest models to three primary data sources: a copper price series covering 1960 to 2024, more than 700,000 hours of industrial process data, and over 188,000 recorded occupational accident cases. Overall, the findings demonstrate that AI and ML-based approaches can transform mining risk management from a reactive framework into a proactive, real-time, and data-driven integrated system.