Building AI Soft-Sensor Models for Flotation Performance
Build a flotation soft sensor: keep JKSimFloat or METSIM for design, then train an ML model on live plant data to predict recovery, grade and throughput.
All
AI 129
Exploration & Target Generation 64
Geospatial Modeling 56
Machine Learning 44
AI Agents 39
Computer Vision 18
Anomaly Detection 17
Process Optimization 15
Hyperspectral Imaging 12
Drilling & Blasting Operations / Blasthole Analysis 11
Reinforcement Learning
Crushing
Crusher optimization with closed-loop AI: Imubit uses reinforcement learning to tune setpoints on live plant data, versus model-based ABB Ability.
Dispatch and Control Room
Mine dispatch optimization with AI mostly means adding an AI add-on, such as Modular Mining's Adaptive Config, to the fleet management system you run.
Flotation
AI flotation control feeds froth-camera video and plant sensor data into a closed-loop controller like Imubit that writes reagent, pH and airflow setpoints.
Thickening
AI thickener control applies model-predictive or reinforcement-learning control to flocculant dosing, through APC platforms like ABB, Metso or Imubit.