Data-driven artificial intelligence in mineral processing: From sensor technology, intelligent algorithms to Genetic Mineral Processing Engineering

With the continuous advancement of sensor technologies, data-driven artificial intelligence (AI) has attracted increasing attention in the field of mineral processing. This review first examines the core advantages and application scenarios of three typical data-driven AI algorithms in mineral processing. Furthermore, the working principles and inherent application limitations of five novel AI-integrated sensors in mineral processing was summarized. Subsequently, the historical development and state-of-the-art research of AI technologies applied in key unit operations are systematically summarized from the perspectives of ore pre-concentration, grinding circuits, froth flotation, and magnetic separation. Finally, this review highlights the research achievements of Genetic Mineral Processing Engineering (GMPE) with respect to genetic characteristics of deposit genesis, ore properties, and mineral features. It also explores the theoretical evolution and technological roadmap of GMPE as a high-level intelligent system engineering. Given that AI is gradually transforming mineral processing from experience-based operations toward data-driven model assisted decision-making, this review proposes the critical directions in which AI-based mineral processing should achieve breakthroughs to enable reliable and stable operation under complex and adversarial industrial conditions.