Physical AI · Concentrator autonomy

The concentrator runs itself. You decide how far.

Orebex is an agent operating system for mineral-processing plants. Specialist agents control crushing and grinding, flotation reagents, air and froth, classification and dewatering, watched by froth vision and online analyzers, coordinated by a whole-plant optimiser, rehearsed on an as-run twin, and bounded by a policy engine you control.

Book a circuit review See the platform

SUCCEEDED run_4b9e27 scenario complete10 steps · 1 approval gateLine 1 · SAG, ball and rougher bank 2

run_4b9e27 · agent graph · line 1 · scenario 10 nodes · 1 approval gate
01 · orchestrator ingest.shift ✓ SUCCEEDED 02 · oretwin twin.simulate ✓ SUCCEEDED 03 · crush_grind grind.setpoint ✓ SUCCEEDED 04 · grade_froth froth.read ✓ SUCCEEDED 05 · orchestrator approve.human ◆ APPROVAL 06 · float_reagent float.dose ✓ SUCCEEDED 07 · separate_classify classify.cyclone ✓ SUCCEEDED 08 · thicken_dewater dewater.thickener ✓ SUCCEEDED 09 · throughput_recovery optimise.targets ✓ SUCCEEDED 10 · quality_conf lot.release ✓ SUCCEEDED

Illustrative scenario: Line 1 · SAG and ball circuit, rougher bank 2 · copper porphyry blend, one shift, one approval gate

Design partners wanted · 2 to 4 concentrators · circuits we are looking for

Copper rougher banks SAG and ball circuits Gold flotation Lithium concentrators Nickel and zinc cleaners Iron-ore grinding and separation

Pre-launch. No customers, pilots or logos are claimed. The first cohort gets a paid pilot on one circuit against a baseline agreed in writing.

Run timeline · scenario

One scenario run, step by step

Every Orebex run is an ordered, inspectable sequence. This is run_4b9e27 on Line 1 · SAG and ball circuit, rougher bank 2 · copper porphyry blend. Scenario completes in 10 steps with one approval gate. P80 held at 148 µm as the blend hardened, froth stability recovered on rougher bank 2 after an air trim, a collector-dose change was approved by the shift metallurgist, predicted recovery held inside the plan, lot C-20931 released with full genealogy.

run_4b9e27 Line 1 · SAG and ball circuit, rougher bank 2 · copper porphyry blend illustrative scenario SUCCEEDED
  1. 01 ingest.shift SUCCEEDED 0.7 s

    Concentrator Orchestrator · Pulled the shift plan from the concentrator MES: copper porphyry blend B-14, P80 target 150 µm, concentrate grade at or above 25% Cu, recovery target per the geometallurgy model, energy and water budgets. Locked the operating envelope for the blend.

  2. 02 twin.simulate SUCCEEDED 58 s

    Concentrator Twin · Simulated 32 candidate grind, air and reagent schemes on the as-run twin for the hardening blend, ranked on predicted recovery, concentrate grade, kWh per tonne and water per tonne. Two candidates breached the thickener bed limit and were discarded.

  3. 03 grind.setpoint SUCCEEDED 3 m 40 s

    Crush-and-Grind · SAG-1 feed trimmed 3% and cyclone water raised as load and power signatures read a harder feed. P80 from the particle-size analyzer held at 148 µm against the 150 µm target; mill power stayed inside the energy budget.

  4. 04 froth.read SUCCEEDED continuous

    Grade-and-Froth · Twelve froth cameras and the online XRF on rougher bank 2 fused at the edge. Froth stability on cells 3 and 4 fell as the blend hardened; predicted recovery drifted 0.4 points below plan. Early warning published to the control agents.

  5. 05 approve.human APPROVAL 1 m 12 s

    Concentrator Orchestrator · Scheme #7 raises collector dose 6% on rougher bank 2, above the standing reagent scheme. Reagent-scheme changes stay supervised on this blend, so the step was held for the shift metallurgist and approved at 07:19:44.

  6. 06 float.dose SUCCEEDED 2 m 05 s

    Float-and-Reagent · Approved collector dose applied with an air trim of 4% down on cells 3 and 4 inside the envelope. Froth stability recovered within eight minutes; predicted recovery returned to plan with concentrate grade read at 25.6% Cu.

  7. 07 classify.cyclone SUCCEEDED 42 s

    Separate-and-Classify · One cyclone brought online in cluster 1 as feed density rose with the hardness change, holding overflow size and circulating load inside the plan.

  8. 08 dewater.thickener SUCCEEDED continuous

    Thicken-and-Dewater · Tailings thickener TH-2 flocculant dose trimmed as bed pressure rose; underflow density held and overflow clarity stayed inside the water-recovery target.

  9. 09 optimise.targets SUCCEEDED 18 s

    Throughput-and-Recovery · Whole-plant targets recomputed for the hardened blend: feed rate, P80, air and reagent targets handed to the control agents with recovery, grade, energy and water inside the budget. The losing candidates and their scores are in the run record.

  10. 10 lot.release SUCCEEDED 1 m 30 s

    Quality-and-Conformance · Concentrate lot C-20931 released: genealogy written (blend, grind record, reagent and air moves, froth and grade reads, approvals, dewatering record, assays as they land) and dispositioned against the concentrate spec with cited evidence for the metallurgist's sign-off.

Tool calls · scenario

The agent shows its work

Every tool invocation, argument and result is written to an append-only, human-readable log, and every reasoning step is expandable. Nothing about a run is hidden from the plant.

tool-call stream · run_4b9e27
  1. 07:00:02mes.read_plan({ line: "L1", blend: "B-14", shift: "day" })
  2. 07:00:02ok envelope locked · P80 150 µm · grade ≥ 25% Cu · energy and water budgets
  3. 07:00:03twin.simulate({ candidates: 32, horizon: "1 shift" })
  4. 07:01:01ok best scheme #7 · recovery on plan · 2 candidates discarded (thickener bed)
  5. 07:01:02policy.evaluate({ scheme: 7, autonomy: "L2+" })
  6. 07:01:02ok 8 writes permitted · collector_dose requires human approval
  7. 07:01:03mill.set_feed({ tag: "SAG1.FEED_TPH", delta: "-3%" })
  8. 07:04:43ok P80 148 µm · power inside budget · hardness index up
  9. 07:04:44froth.read_bank({ bank: "R2", cams: 12, xrf: true })
  10. 07:18:30froth stability falling · R2 cells 3, 4 · predicted recovery −0.4 pt
  11. 07:18:31approval.request({ action: "collector_dose", bank: "R2", delta: "+6%" })
  12. 07:19:44ok approved by shift metallurgist · logged
  13. 07:19:45reagent.set_dose({ bank: "R2", scheme: 7 })
  14. 07:19:46air.set_rate({ cells: [3, 4], delta: "-4%", within: "envelope" })
  15. 07:27:50ok froth stability recovered · grade 25.6% Cu · recovery back on plan
  16. 07:27:51cyclone.set_count({ cluster: "CL-1", online: "+1" })
  17. 07:28:33ok overflow size held · circulating load inside plan
  18. 07:28:34flocculant.set_dose({ thickener: "TH-2", delta: "-2%" })
  19. 07:40:10ok bed pressure steady · overflow clear · water recovery on target
  20. 07:40:11cuopt.solve_targets({ line: "L1", objective: "recovery, grade, energy, water" })
  21. 07:40:29ok targets handed to 4 agents · all inside budget
  22. 07:40:30mes.log_lot({ lot: "C-20931", genealogy: "full" })
  23. 07:42:00ok scenario complete · 10 steps · lot released

Reasoning trace

Plan, thought, action, observation, exactly as the orchestrator would record it.

GoalRun the day shift on line 1 inside the plan for blend B-14: P8…

Run the day shift on line 1 inside the plan for blend B-14: P80 at 150 µm, concentrate grade at or above 25% Cu, recovery maximised, energy and water inside budget.

PlanSimulate grind, air and reagent schemes on the twin first. Lea…

Simulate grind, air and reagent schemes on the twin first. Lead with the grind so the liberation size is right before flotation sees it, then read the froth, then dose. Hold any reagent-scheme change for a human.

ThoughtThe blend is hardening. Holding feed rate would coarsen P80 an…

The blend is hardening. Holding feed rate would coarsen P80 and under-liberate; the twin predicts a 3% feed trim with more cyclone water holds P80 at the cost of a little throughput, and that protecting liberation is worth more than the tonnes.

Actiontwin.simulate(candidates=32) returns scheme #7: SAG-1 feed dow…

twin.simulate(candidates=32) returns scheme #7: SAG-1 feed down 3%, cyclone water up, collector dose up 6% on rougher bank 2, air down 4% on cells 3 and 4.

ObservationScheme #7 predicts recovery on plan, grade 25.5% Cu and energy…

Scheme #7 predicts recovery on plan, grade 25.5% Cu and energy inside budget. Two candidates with higher predicted recovery pushed the tailings thickener past its bed limit and were discarded.

ActionExecute scheme #7: eight bounded writes permitted on grind, ai…

Execute scheme #7: eight bounded writes permitted on grind, air, cyclones and dewatering; the collector-dose write routed to the shift metallurgist.

ObservationFroth stability fell on cells 3 and 4 at 07:18:30 with predict…

Froth stability fell on cells 3 and 4 at 07:18:30 with predicted recovery 0.4 points under plan. The approved dose and the air trim recovered it within eight minutes; grade read 25.6% Cu.

OutcomeScenario complete. P80 held, grade on spec, recovery on plan, …

Scenario complete. P80 held, grade on spec, recovery on plan, lot C-20931 released with full genealogy.

How it works

A shift plan becomes a run

Nothing about an Orebex run is implicit. The goal is declared, the plan is simulated on the twin, each step is typed and bounded, and the outcome is written to the lot's genealogy.

Active path

The orchestrator plans, the agents act

A run starts as a shift plan, becomes a scheme ranked on the twin, and executes as a graph of typed steps. The active path is always visible, always logged, always reversible.

RUNNING APPROVAL SUCCEEDED
<250 ms froth inference per camera stream at the edge · design target
<2 s control recommendation latency for low-risk setpoint moves · design target
8 to 24 froth cameras per flotation bank at 15 to 30 FPS each · design target
20 to 60 models served per site, with versioned canaries and rollback · design target
Human in the loop

Autonomy is a dial, not a switch

L1 shadow, L2 assist, L3 bounded autonomy on low-risk grinding and flotation loops, L4 supervised reagent-scheme adaptation for characterised ore types. Each plant sets the level per agent, per tag, per shift, and every write above the threshold waits for a named approver. The four figures above are design targets, not measured results.

Agent roster

Ten agents, a twin and an orchestrator

Each agent owns a step of the circuit, a bounded tool set and a metric the plant already measures. They negotiate for shared actuators through the orchestrator, never directly. Five named products ship them.

agent.grade_froth · shipped as Frothix

Grade-and-Froth

Froth vision, online analyzers, particle size, mineralogy

Fuses froth-camera vision, online XRF and analyzer readings and particle-size sensors into a live, per-cell read of grade, recovery, froth condition (bubble size, velocity, stability, load), P80 and mineralogy. Predicts grade and recovery seconds after a change rather than hours later from the lab, and flags froth collapse, feed upsets and mineralogy shifts before they cost recovery. It is the perception the rest of the loop acts on.

froth.read_bankanalyzer.read_xrfpsa.read_p80

agent.crush_grind · shipped as Grindex

Crush-and-Grind

Crushing, SAG and ball milling, liberation size, mill energy

Controls crushing and SAG and ball milling to hit target liberation size (P80) and mill throughput without over-grinding. Reads mill load, power, feed rate and water with Frothix particle-size reads, estimates ore hardness from load and power signatures, and trades power against liberation continuously, holding the size flotation actually needs at the least kilowatt-hours per tonne. Grind moves are validated in the twin before they apply.

mill.read_loadmill.set_feedcyclone.set_water

agent.float_reagent · shipped as Floatex

Float-and-Reagent

Flotation cells, reagent dosing, air, pulp level, froth

Controls flotation banks, collector, frother and depressant dosing, air and pulp level to maximise the grade and recovery trade-off, acting on live froth, grade and mineralogy reads and validating reagent and air moves in the twin before applying them. Holds each cell at its best operating point as feed drifts. Reagent-scheme changes stay supervised by the metallurgist until the ore type is characterised.

reagent.set_doseair.set_ratelevel.set_pulp

agent.separate_classify

Separate-and-Classify

Cyclones, classification, gravity, magnetic and dense-media separation

Controls cyclone clusters, classification and gravity, magnetic and dense-media separation so the size and density splits feeding flotation and the concentrate stream stay where the plan put them. Part of the platform, included in the Concentrator tier.

cyclone.read_pressurecyclone.set_countseparator.set_split

agent.thicken_dewater

Thicken-and-Dewater

Thickeners, filtration, flocculant, tailings

Controls thickeners, filtration, flocculant dosing and tailings handling for concentrate and tailings streams: underflow density, overflow clarity, filter cycle and water recovery. Part of the platform, included in the Concentrator tier.

thickener.read_bedflocculant.set_dosefilter.set_cycle

agent.throughput_recovery · shipped as Recovex

Throughput-and-Recovery

Grade, recovery, throughput, energy, water, reagents, as one objective

Takes the live state from Frothix and the control headroom from Grindex and Floatex and solves the grade, recovery and throughput trade-off against energy, water and reagent constraints for the whole concentrator, then hands each agent its targets. Four competing local goals become one plant policy, recomputed as conditions change.

cuopt.solve_targetsmes.read_planenergy.budget

agent.circuit_takt · shipped as Recovex

Circuit-and-Takt

Circuit balance, feed variability, stabilisation

Balances the moving circuit and stabilises it against feed-grade, hardness and mineralogy variability so the control agents are not fighting an upset they could have seen coming. Shipped with Recovex.

feed.read_variabilitycircuit.balancesurge.set_level

agent.robot_handling

Robot-and-Handling

Robotic sampling, materials and reagent handling, inspection rovers

Drives robotic sampling routes, reagent tote inspection and handling, and inspection rovers across crushing, grinding and flotation areas. Skills are validated in Isaac Sim before they run on site, and robotics stays assistive until the site safety case is proven. Part of the platform.

isaac.plan_routesampler.take_samplesafety.check_zone

agent.quality_conf

Quality-and-Conformance

Concentrate spec, non-conformance, lot genealogy, environmental traceability

Builds the concentrate lot genealogy, dispositions against the concentrate specification with cited evidence, records non-conformance and keeps the environmental and water record every action touches. Part of the platform.

mes.log_lotdisposition.writeaudit.write

agent.metallurgy_knowledge

Metallurgy-Knowledge

Reagent schemes, flowsheets, geometallurgy, standards, operator Q&A

Answers process, metallurgy and geometallurgy questions from reagent schemes, flowsheet and geometallurgy specs, mining and environmental standards and recovery and genealogy procedures, with every answer cited to its source and scoped to the tenant. Part of the platform.

rag.retrievestandards.citeprocedure.lookup

twin.oretwin · shipped as Oretwin

Concentrator Twin

As-run concentrator simulation

Simulates crushing, grinding, flotation, classification and dewatering against the site's real telemetry so grade, recovery and throughput can be predicted, and control moves validated, before the shift. Grindex, Floatex and Recovex test every candidate move here first. Rare froth-collapse, feed-upset and mineralogy-shift cases that never appear in normal logs are generated for it.

twin.simulatetwin.compare_asrunplan.optimise

agent.orchestrator

Concentrator Orchestrator

Planning, arbitration, human approval

Plans the shift, arbitrates between agents competing for the same actuator, enforces the autonomy level and routes anything above the risk threshold, reagent-scheme changes first among them, to a named approver.

plan.composepolicy.evaluateapproval.request
Where it pays

Four loops, four metrics

Orebex is not a dashboard. Each agent closes a loop a concentrator already measures in recovery points, kilowatt-hours per tonne, reagent per tonne and water per tonne. The tables show what each loop watches and what it is allowed to move.

Grade and recovery

See the froth before the assay does

Froth cameras, online XRF and particle-size sensors are fused at the edge into a live estimate of grade, recovery, froth condition and P80 per cell. Frothix predicts where recovery is heading seconds after a change and publishes early warnings the control agents act on, instead of a lab result hours later.

  • Grade and recovery predicted continuously, labelled later by the assay
  • Froth collapse and mineralogy shifts flagged before they cost recovery
  • Every read tied to the grind, reagent and air state behind it
agent.grade_froth SUCCEEDED
SignalWatchesMoves
Froth conditionCamera banks, 8 to 24 per bankPublished to Floatex and Recovex
Grade and recoveryOnline XRF plus froth modelEarly warning, targets
P80Particle-size analyzerPublished to Grindex
Grind and energy

Grind to the size flotation needs, not the curve

Mill load, power, feed rate and water are read with Frothix's P80. Grindex estimates ore hardness from the load and power signatures and moves feed, water and cyclone settings so liberation size holds as the blend changes, at the least kilowatt-hours per tonne, with every move validated on the twin first.

  • Over-grinding soft ore and under-liberating hard ore both caught
  • Energy per tonne priced on every move
  • Grind moves rehearsed on the twin before they apply
agent.crush_grind SUCCEEDED
SignalWatchesMoves
P80Particle-size analyzerFeed rate, cyclone water
Mill load and powerMill telemetryFeed rate, speed where available
Ore hardnessEstimated from load and powerScheme candidates ranked on the twin
Reagent, air and froth

Hold each cell at its best point as the feed drifts

Collector, frother and depressant dosing, air and pulp level are controlled per cell against the live froth and grade read. Floatex holds the grade and recovery trade-off the plan asked for, trims air and level inside the envelope, and routes any reagent-scheme change to the metallurgist until the ore type is characterised.

  • Air and level moved inside the envelope
  • Reagent-scheme changes supervised by a named metallurgist
  • Reagent per tonne logged per decision
agent.float_reagent SUCCEEDED
SignalWatchesMoves
Froth stabilityFrothix per cellAir rate, pulp level
Concentrate gradeOnline XRFCollector and depressant dose, under approval
Tailings gradeOnline XRFFrother dose, air, under approval
Whole plant

One objective for recovery, throughput, energy and water

Grinding chases power, flotation chases recovery and sustainability chases water. Recovex solves them as one problem against the plan's constraints and hands each control agent its targets, recomputed as the feed, the constraints and the metal price change.

  • Four local goals, one plant policy
  • Targets handed to agents, never actuators directly
  • Losing candidates and their scores kept in the run record
agent.throughput_recovery SUCCEEDED
SignalWatchesMoves
Recovery and gradeFrothix, assaysTargets to Floatex
Energy per tonneMill and plant metersTargets to Grindex
Water per tonneThickener and plant metersTargets to Thicken-and-Dewater
Multi-agent handoff

Agents negotiate, they do not collide

Two agents will want the same actuator. The orchestrator arbitrates on the run goal, not on who asked first, and the handoff is logged like any other step.

Contested move: rougher bank 2 air rate

  1. 01throughput_recovery.request(air +5%) SUCCEEDED0.2 s

    Throughput-and-Recovery wants more air on bank 2 to lift recovery against the shift plan.

  2. 02grade_froth.request(air −4%) SUCCEEDED0.2 s

    Grade-and-Froth reports froth stability falling on cells 3 and 4 and asks for less air to protect concentrate grade.

  3. 03orchestrator.arbitrate SUCCEEDED0.4 s

    Grade at or above 25% Cu is a hard constraint in the run goal; recovery is the objective. Grade-and-Froth wins the actuator for 20 minutes, inside the envelope.

  4. 04throughput_recovery.handoff(returned) SUCCEEDED20 min

    Actuator returned. Throughput-and-Recovery pursues recovery through the approved collector dose instead. Grade holds.

Illustrative scenario.

Arbitration rules

  • Run goal first. Every request is scored against the declared goal, not the requesting agent's local objective.
  • Constraints outrank objectives. A concentrate grade floor, a water budget or a tailings limit is never traded for a recovery point.
  • The envelope is never contested. Both requests must sit inside the approved operating envelope; anything outside goes to a human.
  • Time-boxed ownership. An agent holds a contested actuator for a bounded time, then must re-justify.
  • Everything is logged. The losing request, the score and the reason all appear in the run record.
  • Deadlocks escalate to a human rather than resolving by timeout.

See the agent roster

Autonomy

Four levels, set per agent and per tag

A concentrator does not go from manual to unattended in one step. Orebex makes the level explicit, auditable and reversible at any time. The first release runs shadow mode, then assist mode, then bounded autonomy on low-risk grinding and flotation loops. Reagent-scheme changes stay supervised until the ore type is characterised.

Autonomy levels and the human role at each
LevelWhat the agent doesWhat the human doesWhere it sits in a pilot
L1 · ShadowReads froth, analyzer, particle, mill, flotation, dewatering, robot and MES signals; predicts and recommends; writes nothingRuns the circuit as today; compares predictions to assays and metallurgist outcomesShadow mode, baseline measured first
L2 · AssistProposes grind, reagent, air, froth and dewatering moves with evidence, confidence, envelope and rollback pathApproves each action in the review consoleAssist mode, the remainder of a 90 to 120 day pilot
L3 · BoundedExecutes pre-approved low-risk grinding and flotation loop moves: feed, water, air, level, flocculant, cyclones online, inside APC, DCS and envelope limitsApproves reagent-scheme changes and lot releaseGraduated autonomy, after audited accuracy and ROI
L4 · Constrained auto-controlSupervised reagent-scheme adaptation for characterised ore types, across circuits and sitesSets the envelope; reviews the shift record and the evaluation resultsPlanned; after golden-dataset evaluation, twin validation and metallurgist sign-off
Targets

Design targets, stated as targets

These are the figures in our compute plan and our first-release plan. None is a measured result. Every pilot measures against the concentrator's own baseline, reproduced from the plant's own assays, meters and lot records, not from our telemetry.

2,000 to 8,000 GPU-hours per commodity model family, once partner data exists · design target
50k to 250k synthetic froth-collapse, feed-upset and mineralogy-shift scenarios per commodity and circuit class · design target
2 to 4 design-partner concentrators in the first cohort, each on a paid 90 to 120 day circuit pilot · plan
1 circuit to start: grinding or flotation, with a baseline agreed in writing · plan
Commodities

Where concentrators start

Copper, gold, lithium, nickel and zinc, iron ore and rare earths. Most plants pick one circuit and one wedge, prove the number, then expand.

CopperBeachhead, high throughput

Copper

Porphyry copper concentrators with instrumented SAG and ball circuits and rougher banks. Frothix and Floatex on the rougher bank where a recovery point is already counted in metal.

GoldRecovery value

Gold

Gold flotation and gravity circuits where a recovery point is worth the most per tonne. Frothix on the flotation bank first, Grindex on the mill as the energy case lands.

LithiumNew plants, new ore

Lithium

Spodumene concentrators ramping on variable feed. Oretwin to rehearse blends and schemes before the shift, Floatex on reagent control.

Nickel and zincReagent and grade

Nickel and zinc

Sulphide flotation with tight concentrate specifications and depressant chemistry. Floatex and Frothix on the cleaner bank with the metallurgist's approval gates.

Iron ore and rare earthsEnergy and separation

Iron ore and rare earths

Grinding, classification and magnetic or gravity separation where energy per tonne dominates. Grindex first, then the classification and dewatering agents in the Concentrator tier.

Accelerated computing

Physical AI, at the bank and the mill

Froth changes in seconds and a shift of feed is gone before the assay lands. Perception has to be local, deterministic and fast, so Orebex runs GPU inference at the plant edge and keeps training, simulation and optimisation in the cloud or on-prem.

Jetson Orin · DeepStream · Holoscan · TensorRT

Edge perception

8 to 24 froth cameras per bank at 15 to 30 FPS plus online XRF, particle-size, pump, level, air, reagent, density and energy streams fused at the plant edge. Sub-250 ms froth inference and local fail-safe through a network outage are the design targets. Jetson Thor is planned.

Triton · NIM

Model serving

Froth, particle and mineralogy, recovery prediction, time-series anomaly and metallurgy retrieval models behind one orchestrator at the plant edge, 20 to 60 per site as a design target, with versioned canaries and rollback.

DGX or HGX · NeMo

Training

Froth segmentation, bubble velocity and stability, grade and recovery prediction, ore hardness, grind-size response, reagent response and metallurgy assistants trained on froth video, analyzer readings, telemetry and metallurgist corrections. 2,000 to 8,000 GPU-hours per commodity model family is the design target once partner data exists.

Omniverse · OVX

Concentrator twin

Grinding, flotation, classification and dewatering simulated as-run against as-designed before the shift. GPU comminution and DEM, flotation and geometallurgy simulation is the workload; physics-informed surrogates with Modulus are planned so schemes rank at production cadence.

Cosmos · Omniverse Replicator

Synthetic data

Rare froth-collapse, feed-upset, mineralogy-shift, equipment-drift and camera-fouling cases generated and validated against real assay distributions before promotion. 50k to 250k scenarios per commodity and circuit class is the design target.

cuOpt

Optimisation

Throughput, energy, water and reagent allocation and the grade and recovery trade-off solved before the shift and during upset recovery, and handed to the control agents as targets.

Isaac Sim · Isaac ROS

Robot sampling and handling

Sampling routes, reagent tote inspection and inspection rovers trained and validated in simulation before they run on site. Assistive until the safety case is proven.

RAPIDS

GPU ETL

Daily features from froth, analyzer, mill, flotation, dewatering and genealogy telemetry for retraining, drift detection and recovery, energy, water and reagent benchmarks.

Developers

Define an agent, bound it, run it

The Orebex SDK is typed Python. Tools are declared with schemas, limits and the operating envelope for the blend, and the policy engine enforces them at call time, not in a review meeting. This is a design preview; the API will change before release.

What you get

  • Typed tool definitions with unit-aware ranges, rate limits and operating envelopes
  • Deterministic replay of any historical run against a new model
  • Local twin harness so a scheme is simulated before it is shipped
  • Autonomy policy as code, versioned and reviewed like any other change

Read the docs Developer guide

l1_grind_agent.py 
# Design preview. Bound the grind agent to SAG mill SAG-1.
from orebex import Agent, Tool, Limit, Autonomy

feed = Tool(
    name="mill.set_feed",
    tags=["SAG1.FEED_TPH", "CL1.WATER_M3H"],
    limits=[Limit(max_step="3%", per="300s", envelope="blend-B14")],
)

grind = Agent(
    id="agent.crush_grind",
    goal="P80 150 um +/- 10, least kWh per tonne, throughput inside plan",
    tools=[feed, Tool("psa.read_p80", read_only=True)],
    # bounded writes on feed and water; reagent scheme stays with a human
    autonomy=Autonomy.L3,
    # simulate on the twin before every write
    verify="twin",
)

run = grind.start(line="L1", blend="B-14")
for step in run.stream():
    print(step.name, step.status, step.duration)
CLI

Start a run from anywhere

The same run engine, the same policy checks, the same audit trail, from the terminal, the review console or the SDK. Design preview.

orebex · cli · scenario 
$ orebex run "day shift, line 1, blend B-14" --autonomy L2+

→ plan composed          10 steps · 1 approval gate
→ twin.simulate          32 candidates · best #7 · recovery on plan
→ policy.evaluate        8 writes permitted · collector_dose held for human
→ executing              grind.setpoint ... ok   3m40s
→ executing              froth.read ....... ok   continuous
! approval required      collector_dose R2 +6%
→ approved               shift metallurgist · 07:19:44
→ executing              float.dose ....... ok   2m05s
→ executing              optimise.targets . ok   18s
→ scenario complete      grade 25.6% Cu · lot C-20931 released

$ orebex runs show run_4b9e27 --format genealogy
Pricing

Priced per circuit, per plant, per group

Land on one workflow with a measured baseline. Expand by circuit, area, concentrator and autonomy level. Outcome components are available on recovery, energy, water and reagents. Annual prepay carries a 15 to 20% discount.

Circuit $20,000 per grinding or flotation circuit / month · billed monthly

Comminution or flotation control, or grade sensing, for one circuit. For concentrators validating ROI on the wedge.

  • One circuit, one agent module
  • Plant-edge inference node included
  • APC, DCS and sensor reads, guarded writes
  • Run log, audit trail, lot genealogy
  • Business-hours support
Start a pilot
Recommended Concentrator $130,000 per plant / month · billed monthly

The whole plant: crush and grind, float and reagent, separate and classify, thicken and dewater, grade and froth, throughput and recovery, circuit and takt, robotic handling and the twin.

  • All control agents on every circuit
  • Froth-vision and analyzer fusion on every bank
  • Whole-plant optimisation and circuit stabilisation
  • Robotic sampling and handling
  • As-run concentrator twin
  • 24/7 support, quarterly model review
Talk to us
Enterprise Custom land $900k to $12M ACV

Multi-site mining groups standardising on Orebex, with custom ore, mineralogy and recovery models.

  • Multi-site fleet management
  • Custom ore, mineralogy and recovery models
  • Plant edge fleet with on-prem option
  • Group autonomy policy and approval chains
  • SLAs, SSO and RBAC, dedicated engineer
Contact us
Who it is for

Built for the people who run the concentrator

We have no customer quotes to show yet and will not invent any. These are the three roles in a concentrator we build for, and the problems they bring to the first meeting.

Concentrator or Processing Manager

Owns recovery, throughput, grade, energy and water, and cost per tonne.

What this person is dealing with today

  • Fixed setpoints and periodic assays that let recovery drift and metal walk out in the tailings
  • Enormous energy, water and reagent consumption eroding cost and licence to operate
  • A severe, remote, retiring metallurgist and operator workforce that cannot continuously optimise

Senior or Chief Metallurgist

Owns grind, flotation, reagent and geometallurgy strategy, process windows and recovery.

What this person is dealing with today

  • Grind-size drift that over-grinds or under-liberates
  • Reagent, air and froth errors that sacrifice grade or recovery
  • Feed-grade, hardness and mineralogy variability that destabilise the circuit

Sustainability, Environmental or Energy Lead

Owns energy, water, tailings and environmental compliance and licence to operate.

What this person is dealing with today

  • Grinding as the site's largest electricity load, run on fixed curves
  • Water per tonne and tailings governance under tightening constraints
  • No shared record tying every control move to its energy, water and reagent effect
Integrations

It speaks concentrator, not cloud

Orebex reads and writes through the systems already in the plant. No rip and replace, no parallel historian, no new HMI to learn. Vendor names are examples of systems found in concentrators, not certified integrations; connector availability is confirmed per site.

Comminution and mill control

Mill control and APC (Metso, FLSmidth, ABB, Weir class systems)

Load, power, feed and water reads, guarded writes

Flotation APC and expert systems

Flotation APC and expert optimisers (ABB Expert Optimizer, Metso ACT, Honeywell, Andritz class systems)

Air, level and reagent reads; supervised reagent writes

Froth vision

Froth cameras and froth-vision systems (Metso VisioFroth, Stone Three, JKTech class tools)

Camera streams fused by Frothix

Online analyzers

Online XRF and particle-size analyzers (Malvern Panalytical, Thermo Fisher and CIDRA, Metso, FLSmidth PSI class systems)

Grade and P80 reads per stream

Reagent dosing

Collector, frother, depressant and flocculant dosing systems

Dose reads; supervised and bounded writes

Classification and dewatering

Cyclone clusters, separators, thickeners, filters

Pressure, density, bed and cycle reads, guarded writes

DCS and historians

Plant DCS tags and time-series historians

Backfill, replay, time alignment

Lab assays

LIMS and assay results

Delayed labels for every prediction

See all integrations

Guardrails

An agent that can move a reagent pump needs a leash

Orebex writes to production equipment. Every capability is scoped, every write is policy-checked against the operating envelope, and every action is written to an append-only audit log the plant owns.

  • Bounded action space. Each agent writes only to an explicit tag allow-list, inside per-tag rate and magnitude limits and the operating envelope for the blend.
  • Policy engine before every write. Autonomy level, shift, ore type, interlock state, water and tailings constraints and metallurgist policy are evaluated before a setpoint moves.
  • Human-in-the-loop gates. Reagent-scheme changes and lot release wait for a named approver. Robotics stays assistive until the site safety case is proven.
  • Immutable audit log. Append-only and exportable. Every action records source evidence, the envelope applied, confidence, estimated recovery, grade, energy, water and reagent impact, rollback path, approver and lot scope.
  • Hard fallback. Loss of the edge node, the network or a model returns control to the APC or DCS last known-good state. Interlocks are never in Orebex's path, and the edge runs fail-safe through a network outage.
  • Tenant and ore-body isolation. Ore-body data, geometallurgy models, reagent schemes and recovery records never cross a customer boundary. On-prem and edge deployment is available for remote sites.

Compliance posture

ItemScopeStatus
SOC 2 Type ICloud control plane QUEUED Planned, first 6 months
SOC 2 Type IICloud control plane QUEUED Planned, 6 to 12 months
SSO and RBACConsole and approvals QUEUED Planned, 6 to 12 months
Environmental and water assurance packagesRecovery, energy, water and tailings evidence QUEUED Planned, 6 to 12 months; not a certification claim
Standards-cited reasoningISO, ICMM, tailings and water governance documents RUNNING In the design, every answer cites its source
Per-tenant and ore-body isolationOre data, schemes, models, memory RUNNING In the design from day one
On-prem and plant-edge deploymentRemote and IP-sensitive sites RUNNING Design option

Read the security overview

Enterprise

Standardise autonomy across the group

One policy model, one audit trail, one benchmark across every circuit in every concentrator, with the ore, mineralogy and recovery models kept private to each site.

Talk to us Enterprise details

  • SSO and role-based access down to the tag level, planned for the 6 to 12 month release
  • Group-wide autonomy policy with per-site override and approval chains
  • Cross-plant benchmarking on recovery, grade, energy, water and reagents
  • On-prem and plant-edge deployment for remote and IP-sensitive sites
  • Edge autonomy that survives control-plane loss by returning to the DCS last known-good state
  • Dedicated deployment engineer and quarterly model review per concentrator
FAQ

Straight answers

The questions processing managers, metallurgists and sustainability leads ask in the first meeting.

Yes, but only within an explicit tag allow-list with per-tag rate and magnitude limits, inside the operating envelope for the blend, and only at the autonomy level your plant has set. Design partners start in shadow mode, where Orebex writes nothing. Reagent-scheme changes stay supervised by a named metallurgist until the ore type is characterised.

Get started

Bring autonomy to your concentrator

Start with one grinding or flotation circuit and one measurable baseline. A paid pilot on froth-vision recovery control, grind-size energy or reagent dosing shows the number before you commit further.

Book a circuit review See pricing