Predictive maintenance systems that can prevent costly asset downtime by catching component failures have become a reality in recent years as connected sensors are integrated into mineral processing facilities. But some companies are already looking to the next big evolutionary step in asset health monitoring – a system that does not just identify failing components, but also uses AI and machine learning to recommend a long-term solution.