Predicting Mineralogy and Geometallurgical Domains with ML
Predict geometallurgical domains by training gradient-boosted models on QEMSCAN, MLA or TIMA data, then carrying predictions into a Leapfrog block model.
Eigenform AI Geo Tooltips
Predicting Rock Strength (UCS) from Point Load Testing with AI
UCS prediction with ML: train XGBoost or Random Forest on paired point load index and UCS lab results, then estimate rock strength wherever you have PLT.
Proxy Models
Geometallurgical proxy models predict an expensive assay from a cheap measurement. Build one with Cancha for a turnkey path, or scikit-learn and XGBoost.
Quality Assurance and Quality Control
Automate geochemical QA/QC with ALS QCPro or ioGAS to track standards, blanks and duplicates, and let anomaly flagging catch lab drift sooner.
Quick Geological Logging
AI geological logging turns core photos into a first-pass lithology and structure log in hours. Datarock Core does it, and geologists review the draft.
Real Time Geometallurgy
Real-time geometallurgy with AI: pipe ore proxies such as hardness and liberation into a closed-loop platform like Imubit or ABB Ability to adapt the plant.
Real-Time Ore Grade Sensing at the Shovel
Shovel grade sensing with MineSense ShovelSense: XRF sensors on the bucket feed ML grade models that route each truck by measured ore grade in real time.
Recovery & Hardness Prediction with Machine Learning
Recovery and hardness prediction with ML: train XGBoost, LightGBM or Random Forest on mineralogy, geochemistry and geomet testwork, or use Cancha.
Refined Hydrological Response Model
Hydrological model calibration with PEST++: fit MODFLOW 6 or FEFLOW to monitoring data, while ML surrogates and ensemble methods emerge to speed it up.
Statistical Analysis and Domaining
Speed up geological domaining with clustering (K-means or GMM in scikit-learn) on assay and logging data, then validate the new domains in GeostatsPy.