SAFETY
People and exposure
01MINING PLANT · AOI
Operational intelligence for a mining plant.
Connect safety, process, assets and product in one operational view. AOI brings visual evidence, plant data and context together for supervisors, metallurgy, maintenance and the control room.
Discuss a mining plant use case
AOI / PLANT INTELLIGENCE CONCENTRATION · LEACHING
RGB · THERMAL · 3D · DCS / PLC · LIMS
01 / AOI + AICEA SHARED OPERATIONAL VIEW
One shared view of the plant.
Four connected perspectives to understand operating conditions and decisions.
CONTINUITY
Assets and transfers
PROCESS
Stability and constraints
PRODUCT
Quality and traceability
AICE observes
Turns visual and sensor sources into observable entities, states and events.
AOI contextualizes
Relates events to assets, areas, shifts, lots and process conditions.
Predictive Intelligence anticipates
Looks for trends and deviations with models validated for the use case.
MINCORE recommends
Proposes actions with evidence, constraints and expected verification.
OBSERVE›CONTEXTUALIZE›ANTICIPATE›RECOMMEND
Local Edge capture and processing are defined by design. Integration with DCS, PLCs, historians, LIMS and CMMS depends on available interfaces.
02 / PLANT USE CASESAPPLICATION PORTFOLIO
Ten use cases for a mining plant.
Selection depends on the flowsheet, operational priority and available sources at each site. Concentration and leaching are different process routes and may not coexist.
AOI / SAFETY AND ACCESS01SAFETY AND ACCESS
People and equipment in one context
Identify exposure around conveyors, crossings, walkways and maintenance areas so attention is directed where it is needed.
WHAT IT OBSERVES AND RELATES
- Presence, visible PPE, intrusion and dwell time in configured zones.
- People-equipment interaction at crossings and maneuvers; calibrated distance where scoped.
- Access linked to credentials or permits when Identity and the required data are available.
DECISIONS IT SUPPORTS
- Notify the responsible person with location, time and event evidence.
- Check authorization and area conditions before an intervention.
- Review recurrence to refine segregation, routes and controls.
Validate coverage, occlusions and distances. Visual observation does not verify energy isolation.
AOI / CRUSHING, CONVEYORS AND CHUTES02CRUSHING, CONVEYORS AND CHUTES
Continuity from the first transfer
Make blockages, spillage and exposed objects visible before a local anomaly affects plant feed.
WHAT IT OBSERVES AND RELATES
- Buildup in bins and chutes; loading, discharge and stop sequences.
- Spillage, visible belt misalignment and exposed foreign objects.
- Thermal trends in motors, gearboxes and idlers alongside available load and speed.
DECISIONS IT SUPPORTS
- Locate the flow constraint and relate the event to its equipment section.
- Coordinate verification, removal or cleanup under plant procedures.
- Prioritize feed adjustments and inspections by recurrence and impact.
Vision identifies exposed objects and does not replace metal detectors. Automatic stops need integration and site-validated interlocks.
AOI / STOCKPILE AND FEED03STOCKPILE AND FEED
Measure the stockpile. Stabilize feed.
Connect stockpile geometry and inventory with particle size and reclaim flow to support consistent feed.
WHAT IT OBSERVES AND RELATES
- Stockpile geometry, volume and stacking or reclaim movements through 3D.
- Visual size distribution and oversize on the belt, linked to source and shift.
- Differences between geometric volume, scale weights and movements with validated density.
DECISIONS IT SUPPORTS
- Prioritize reclaim and review feeders when changes persist.
- Evaluate blending with grade, moisture and operating constraints provided by the plant.
- Reconcile inventory and adjust plans using movements, weights and sampling.
Volume is not mass without validated density. Surface particle-size estimates need calibration; grade and moisture require instruments or lab data.
AOI / GRINDING AND CLASSIFICATION04GRINDING AND CLASSIFICATION
Stabilize grinding and its product
Relate feed to circuit response to understand deviations in load, energy and classification.
WHAT IT OBSERVES AND RELATES
- Throughput, power, water, pressure, density and product size when instrumented.
- Feed changes, oscillations and circuit states; cyclone discharge anomalies when visible.
- External thermal condition of drives and visible leaks linked to the asset.
DECISIONS IT SUPPORTS
- Distinguish whether a constraint aligns with feed, grinding or classification.
- Suggest reviewing flow, feed or cyclone configuration within approved limits.
- Compare equivalent ore conditions to understand stable periods.
Pulp P80 and internal mill load need validated instruments or models. External cameras do not measure mill load.
AOI / COPPER FLOTATION05COPPER FLOTATION
Read froth with metallurgical context
Turn froth appearance and dynamics into cell- and bank-level information linked to process conditions and metallurgical results.
WHAT IT OBSERVES AND RELATES
- Surface speed, apparent bubble size, texture and froth stability.
- Cell and bank changes, collapse or irregular overflow with traceable images.
- Timing against available air, level, reagents, pH and assays, including their delays.
DECISIONS IT SUPPORTS
- Flag cells or banks that differ from the validated operating pattern.
- Suggest reviewing air, level or dosage with metallurgy and within approved limits.
- Compare the subsequent response with grade and recovery from the mass balance.
Froth is a process signal; grade and recovery require assays and a metallurgical balance.
AOI / MOLYBDENUM SEPARATION06MOLYBDENUM SEPARATION
Track molybdenum separation and cleaning
Relate froth, circuit chemistry and lab results to support stability and molybdenum concentrate quality.
WHAT IT OBSERVES AND RELATES
- Froth patterns and changes between separation and cleaning stages.
- pH, ORP, flow and dosage when instrumented.
- Mo and impurity assays, feed and recirculation with residence times considered.
DECISIONS IT SUPPORTS
- Distinguish a visual change from a deviation confirmed by process variables and assays.
- Prioritize pH, ORP or reagent review against the validated pattern.
- Suggest sampling or stage review when selectivity deteriorates.
Cameras do not measure selectivity or detect H₂S. Gas management needs sensors, engineering and circuit-specific procedures.
AOI / LEACHING, IRRIGATION AND CONTAINMENT07LEACHING, IRRIGATION AND CONTAINMENT
See distribution. Verify irrigation.
Locate unusual surface conditions and compare them with irrigation, flows and balances to prioritize inspections.
WHAT IT OBSERVES AND RELATES
- Ponding, runoff or other surface appearance outside the expected pattern.
- Visible damaged lines or obstructions in channels, ponds and transfers.
- Flow, pressure by sector, irrigation cycles, weather and available solution chemistry.
DECISIONS IT SUPPORTS
- Geolocate sectors for inspection and check the irrigation network.
- Prioritize distribution, containment and drainage reviews when anomalies persist.
- Relate events to solution consumption and campaign performance.
Surface appearance does not prove internal moisture, permeability or recovery. Check instruments, balances and field tests.
AOI / CRITICAL ASSET CONDITION08CRITICAL ASSET CONDITION
See deterioration. Prioritize intervention.
Combine Thermal Analytics and Predictive Intelligence to recognize persistent deviations under comparable conditions.
WHAT IT OBSERVES AND RELATES
- Surface temperatures, asymmetries and trends in defined asset regions.
- Visible leaks or vibration alongside available instrumented vibration and electrical data.
- Evolution against load, ambient conditions, equivalent assets and maintenance history.
DECISIONS IT SUPPORTS
- Prioritize inspections by criticality, persistence and rate of change.
- Link evidence to work orders and verify condition after intervention.
- Tune thresholds by asset with maintenance and confirmed patterns.
An anomaly does not confirm failure. Account for emissivity, reflections, distance and load; remaining life needs asset-specific history and validation.
AOI / THICKENING, FILTRATION AND INVENTORY09THICKENING, FILTRATION AND INVENTORY
Connect recovered water, product and inventory
Follow thickening and filtration continuity, and reconcile concentrate movement with its condition, lot and inventory.
WHAT IT OBSERVES AND RELATES
- Overflow, spills, filter sequences and visible cake formation.
- Flow, torque, density, solids and moisture where reference measurements exist.
- 3D volume, scale weights, movements and lot assignment in storage and dispatch.
DECISIONS IT SUPPORTS
- Review deviations between feed, discharge and recovered water.
- Identify unusual filtration cycles and prioritize review or sampling.
- Reconcile stock, weights and dispatch; investigate differences by lot.
RGB does not measure moisture or grade. Converting volume to mass needs validated density and comparable reconciliation bases.
AOI / AI RECOMMENDERS AND OPTIMIZERS10AI RECOMMENDERS AND OPTIMIZERS
Recommend within plant constraints
AI Process Operator and MINCORE combine AOI evidence with process models to evaluate alternatives within approved operating windows.
WHAT IT OBSERVES AND RELATES
- Safety, stockpile, grinding, flotation and asset events on one timeline.
- Treatment, energy, water, quality and availability objectives and limits.
- Response to previous actions, delays and data quality for scenario comparison.
DECISIONS IT SUPPORTS
- Recommend an action with an owner, evidence and expected verification.
- Evaluate feed, blending, air or level alternatives using validated models.
- Record approval and measure the recommendation’s actual effect.
Advisor mode with human approval. Optimization and closed-loop control need validated models, constraints, permissions and interlocks.
03 / AOI ECOSYSTEMAOI ECOSYSTEM
Eight modules. One operation.
AICE provides observation; AOI organizes context and traceability to connect signals with a concrete decision.
AI Safety
People, PPE and exposure in defined zones.
Process Intelligence
States, sequences and relationships across stages.
Thermal Analytics
Surface thermal condition and trends.
Predictive Intelligence
Anomalies, persistence and validated anticipation.
Proximity Awareness
People-equipment interaction and calibrated zones.
Identity
Access traceability by credential or ID.
Quality & Inventory
Particle size, product inspection and reconciliation.
AI Process Operator / MINCORE
Recommendations, constraints and action outcomes.
04REFERENCE ARCHITECTURE
SOURCESRGB / thermal / 3D · DCS / PLC / sensors · LIMS / CMMS / ERP
AICE + AOILocal capture / data quality / events / context
DECISIONAlerts / dashboard / MINCORE recommendation / verification
Edge capture and processing are defined by design. Integration depends on available interfaces, permissions and data quality.
05SCOPE AND VALIDATION
Every signal needs a valid reference.
Applications are configured and validated for each site. Images and video do not replace instruments, assays or safety procedures: particle size, mass, moisture, grade, internal mill load, selectivity and equipment failure need specific references. Brochure indicators are proposed validation measures, not measured results.
06NEXT STEP
Choose a high-value decision.
Define the plant’s first control point from a loss, its available sources and the decision to improve.
01WHERE
Priority area and loss
02WITH WHAT
Real sources and coverage
03FOR WHAT
Decision and indicator