Background: Smart mining is rapidly evolving through the integration of automation, advanced sensing technologies, and Machine Learning (ML). Methods: This Systematic Literature Review (SLR), based on 99 peer-reviewed studies published between 2021 and 2025 and synthesized following PRISMA 2020 guidelines, analyzes the current role of ML algorithms in the mining sector, focusing on their applications, algorithmic prevalence, benefits, and challenges. Results: Machine Learning algorithms are primarily applied to equipment failure prediction, ore classification, grade and flotation control, transportation optimization, and environmental monitoring. The most frequently adopted algorithms include Decision Tree-based models, Artificial Neural Networks, Deep Learning architectures, Support Vector Machines, K-means clustering, and Gradient Boosting methods, reflecting different trade-offs between interpretability, computational complexity, and predictive performance. Reported benefits include improved operational efficiency, cost reduction, enhanced predictive maintenance, improved decision-making, and increased safety and environmental performance. However, widespread adoption remains constrained by limited availability of high-quality data, data heterogeneity, high implementation costs, shortages of specialized personnel, algorithm interpretability issues, and cybersecurity risks. Conclusions: Overall, ML algorithms emerge as key enablers of intelligent and sustainable mining. The review highlights the need for explainable and robust algorithms, improved multimodal data integration, and large-scale real-world validation to support the next generation of smart mining systems.

Mercury (Hg) is a heavy metal whose toxicity poses significant environmental and health risks. Utilising biochar prepared from biomass waste is a straightforward and effective method for removing mercury from water. This research centred on producing a thiol-functionalised biochar derived from olive-stone waste for the removal of mercury (II) from aqueous solutions. Characterisation analyses confirmed successful functionalisation. The biochar, despite having a limited specific surface area (4.14 m²/g) due to raw material nature and pyrolysis conditions, exhibited a notable ability for mercury (II) adsorption, primarily attributed to the thiol-modified surface. Adsorption was assessed using a 2³ factorial design, with the variables being adsorption time, biochar dose and initial mercury (II) concentration in the solution. Biochar dose emerged as the most influential factor, followed by adsorption time and, lastly, initial mercury (II) concentration. The peak removal efficiency of the model stood at 98.19%. The kinetics aligned with the pseudo-first-order and intraparticle diffusion models, suggesting a surface adsorption mechanism coupled with pore diffusion. This work accentuates the potential of olive-derived biochar, when thiol enhanced, in treating aqueous systems contaminated with mercury (II).