COMMINUTION & CRUSHING
Molycop’ s‘ digital-to-physical’ approach
With its OreVia solution, Molycop is looking to provide a holistic approach to process characterisation, defining the journey of ore through the processing circuit through three solutions: OreVia – Rock, OreVia – Slurry and OreVia – Froth. The improved visibility offered with these solutions provides clarity, insight and direction to optimise every stage of the milling-to-recovery process, it says.
Using machine learning, OreVia captures images across the entire mineral processing circuit, providing critical insights into process efficiencies, learning and improving over time like a human operator.
IM spoke to Edgar King- Manager- Machine Vision at Molycop to find out more.
IM: Can you talk to how long it takes for the AI algorithms to be trained before they start providing valuable insights? Are clients able to provide historic data to accelerate this process?
EK: OreVia was engineered to deliver impact from day one. Its AI models are pre-trained and context-aware, allowing operations to gain immediate process visibility upon commissioning. From there, performance is progressively enhanced through site-specific tailoring. Rather than relying on self-learning in production, OreVia Rock and OreVia Froth models are refined using site-captured footage to ensure accuracy under each operation’ s unique lighting, ore type and environmental conditions. This structured approach ensures stability, reliability and transparency in how insights are generated.
To accelerate deployment, Molycop offers off-site evaluation programs. Operations can provide existing video footage( for Rock and Froth) or slurry samples( for Slurry), enabling algorithm refinement before hardware installation. This significantly shortens the path from installation to measurable value.
For OreVia Slurry, finetuning to site process dynamics typically takes around four weeks, ensuring the system delivers robust, high-confidence particle size measurements aligned with real plant conditions.
IM: Could you identify some application cases where the data from the OreVia platform is helping operations with decision-making around how to more efficiently carry out downstream processing? Do you also see the opportunity to help operations reconcile their mine plans based on the data you are seeing?
EK: OreVia is enabling a step-change in grinding circuit control, visibility, decision making and mine-to-plant analysis.
At the front end of the plant, OreVia Rock provides real-time PSD analysis at stockpile reclaim feeders or SAG mill feed. This allows operators to identify ore segregation and adjust blending strategies before variability impacts throughput. PSD is critical to mill optimisation, throughput and the efficiency of the grinding process. Combining an optimum feed size with optimised media choice influences all the downstream liberation and recovery processes. PSD measurement via OreVia is a critical new tool in Molycop’ s circuit analysis toolkit. The result is a more stable grinding circuit, reduced media and liner wear and improved energy efficiency.
Further downstream, OreVia Slurry delivers continuous P80 measurement in the cyclone overflow stream. Instead of relying on delayed laboratory results, operators can make immediate adjustments to density, feed rate or classification parameters – improving liberation and flotation performance in real time.
Visibility of P80 measurement allows for the balancing of recirculating loads with classification systems. This facilitates the optimum choice of Molycop reagents and chemicals.
In terms of mine plan reconciliation, OreVia introduces powerful new feedback loops. While the system is not currently deployed at the shovel for direct blast fragmentation measurement, downstream installations – such as beneath the primary crusher – provide valuable indicators of blasting performance, particularly in finer fractions. This data bridges the traditional gap between mining and processing, enabling operations to better align blast design assumptions with actual plant performance.
OreVia Rock provides real-time PSD analysis at stockpile reclaim feeders or SAG mill feed, allowing operators to identify ore segregation and adjust blending strategies before variability impacts throughput
IM: Where have you trialled / deployed the solution to date? What sort of insights were gained?
EK: OreVia has been successfully deployed across multiple global mining operations and is fully integrated with the StarCS control systems and the MillROC / VIP Platform. Installations span major gold and copper operations in the Americas and Africa, as well as copper, gold and lithium sites worldwide.
Across deployments, key insights have included early detection of flotation instability through OreVia Froth, identification of adverse pulping conditions and direct measurement of fine particle fragmentation( 1-100 µ m) within slurry pipelines – eliminating reliance on manual sampling and reducing data lag.
Collectively, these implementations demonstrate how OreVia is transforming process measurement from a periodic check into a continuous, intelligent decision-support system. By converting real-time particle data into actionable control strategies, Molycop is helping operations move toward a more predictive, datadriven future. This adds real time decision making to the existing Molycop Tools analytical toolkit.
IM: Anything else to add on the topic of process characterisation within the flowsheet?
EK: It is important to view OreVia not as a standalone sensor, but as a holistic‘ mapping’ of the ore’ s journey. Process characterisation isn’ t just about digital data; it’ s about merging that data with physical metallurgical expertise. We combine the real-time AI insights from OreVia with our laboratory capabilities – like the Bond Ball Mill Work Index and Abrasion Index testing – to provide a complete picture of ore hardness and grindability. We combine this with Molycop’ s mill analysis, media and flotation chemical expertise to give a full grinding-to-recovery solution. This‘ digital-to-physical’ approach allows us to forecast metal production and plant performance with a level of accuracy that traditional sampling simply cannot match.
50 International Mining | MARCH 2026