# Beneficiation Yield Optimizer

*/Opportunities/Beneficiation_Yield_Optimizer*

## Opportunity Overview

**Wedge**: The initial beachhead targets copper froth flotation circuits in mid-tier mines. Flotation is highly sensitive to reagent dosing and produces immediate, measurable yield improvements, generating fast proof of value. Expansion moves backward into comminution circuits to optimize particle size, and subsequently forward into leaching and tailings management.
**Timing**: Recent deployments of high-frequency IoT sensors like inline X-ray fluorescence (XRF) analyzers and edge AI inference hardware allow low-latency, localized control without relying on unstable cloud connections. Current models process multi-modal telemetry streams simultaneously to predict metallurgical outcomes in real time.
**Why This I C P**: Base metal and battery metal operators face rapidly declining ore grades and strict environmental constraints on reagent use, creating immediate financial pressure to extract maximum value from lower-quality rock.
**Size Of Prize**: There are roughly 2,500 active large-scale metal mines globally. Capturing a fraction of the yield improvement value supports an average contract of $500,000 per year per processing plant, creating a $1.25B annual addressable market.
**Gap Narrative**: Mining beneficiation circuits rely on static, human-configured setpoints that fail to adapt to minute-by-minute variations in ore hardness and grade. Operators leave valuable minerals in tailings because manual adjustments to reagent dosages and flow rates are reactive. This opportunity provides a closed-loop control system that continuously updates plant parameters based on live metallurgical telemetry.
**Defensibility**: Defensibility compounds through proprietary, plant-specific metallurgical models that map local ore variability to optimal setpoints over time. Direct read-write integration into the plant DCS creates deep workflow lock-in, as extracting the system immediately degrades plant yield.
**Why This Thesis**: A direct-integration software agent is structurally required because continuous, real-time control demands autonomous adjustments written directly to the Plant Control System (DCS), bypassing static dashboards that operators frequently ignore during high-stress shifts.

## Opportunity Linked Thesis

**Thesis**: [Software](/Theses/Software)

## Opportunity Linked I C P

**Icp**: [Mining Company](/CompanyTypes/Mining_Company)

## Opportunity Market Sizing

_Illustrative — target and order-of-magnitude estimate figures, not an achieved track record (this Thing is concept-stage)._

**S A M**: ~$750M - $1.2B targeting Tier 1 and mid-tier base and precious metal operations
**S O M**: ~$15M - $45M
**T A M**: ~20,000 global mineral processing facilities × ~$150k/yr ≈ ~$3B
**Growth Rate**: ~8-12%/yr, driven by globally declining ore grades requiring tighter recovery margins and rising reagent costs
**Paid Comparable Spend**: ~$200k - $500k/yr per site on legacy Advanced Process Control systems, third-party metallurgical consultants, and manual lab assay workflows

## Opportunity Incumbents

- [AVEVA Process Optimization](/Products/AVEVA_Process_Optimization) — Tool
- [FLSmidth Optimization Services](/Products/FLSmidth_Optimization_Services) — Service
- [IntelliSense Brain](/Products/IntelliSense_Brain) — Tool
- [In-House Excel Models](/Products/In-House_Excel_Models) — Spreadsheet
- [Custom MATLAB Scripts](/Products/Custom_MATLAB_Scripts) — DIY
- [SGS Metallurgical Services](/Products/SGS_Metallurgical_Services) — Service

## Opportunity Win Conditions

**Kill Thresholds**:
- Data ingestion and SCADA integration phase exceeds 45 days per site
- Operator recommendation acceptance rate drops below 40% after 14 days
- Demonstrated recovery yield improvement is under 0.5% against baseline after 30 days of active control
- Customer refuses to transition to a paid $150k per year contract after a 90-day pilot
**Leading Metrics**:
- Time-to-first-value (days from historian connection to first accepted setpoint change)
- Operator acceptance rate of recommended setpoint adjustments
- Reagent dosing variance against historical baseline averages
- Percentage of required sensor data streams successfully ingested per hour
- Frequency of manual system overrides per operator shift
**What Proves Right**: Control room operators accept the system's reagent dosing and flotation circuit recommendations at least 80% of the time instead of reverting to manual setpoints. Active sites demonstrate a 1% to 2% increase in baseline recovery yield or a 10% reduction in reagent consumption within the first 60 days of deployment. Customers convert from initial deployments to $150,000 annual software subscriptions without requiring continuous on-site implementation engineering.
**What Proves Wrong**: Plant metallurgists override the system consistently because the algorithms fail to account for upstream ore hardness variations or sensor drift. Data integration with legacy SCADA and historian servers requires more than 60 days of custom scripting per site, turning the product into a slow consulting service. The isolated recovery improvements fall below 0.5%, failing to justify the subscription cost over incumbent in-house Excel models.

## Opportunity Build Profile

**Hardest Part**: Modeling the non-linear, time-delayed interactions in the flotation circuit using highly noisy and frequently miscalibrated physical sensor inputs to predict precise recovery outcomes.
**Min Viable Scope**: Deliver an open-loop recommendation dashboard solely for the froth flotation circuit of base metal concentrators, outputting specific reagent dosage and airflow adjustments for metallurgists. Deliberately exclude closed-loop PLC control, upstream comminution circuit optimization, and complex polymetallic ores.
**Cold Start Problem**: The system requires massive volumes of site-specific historical SCADA data to establish baseline physics before recommending safe optimizations. Break this by partnering with a single mid-tier mine operator to extract two years of OSIsoft PI historian data for offline backtesting.
**Time To First Value**: 3-4 weeks to ingest historical historian data and prove theoretical yield improvements via backtesting against known production shifts
**Data Moat Available**: true
**Technical Difficulty**: High

## Neighborhood

### Where the gap lives

- [Coal Mining](/Industries/Coal_Mining) — latent gap · Industries

### Incumbent in

- [SGS Metallurgical Services](/Products/SGS_Metallurgical_Services) — incumbent in · Products
- [In-House Excel Models](/Products/In-House_Excel_Models) — incumbent in · Products
- [IntelliSense Brain](/Products/IntelliSense_Brain) — incumbent in · Products
- [AVEVA Process Optimization](/Products/AVEVA_Process_Optimization) — incumbent in · Products
- [Custom MATLAB Scripts](/Products/Custom_MATLAB_Scripts) — incumbent in · Products
- [FLSmidth Optimization Services](/Products/FLSmidth_Optimization_Services) — incumbent in · Products

### Applies thesis

- [Mining Company](/CompanyTypes/Mining_Company) — applies thesis · CompanyTypes

### Embodies

- [Software](/Theses/Software) — embodies · Theses

### Similar Opportunities

- [Ore Beneficiation Controller](/Opportunities/Ore_Beneficiation_Controller) — similar · Opportunities
- [Extracta Yield Engine](/Industries/Mining,_Quarrying,_and_Oil_and_Gas_Extraction/Opportunities/Extracta_Yield_Engine) — similar · Opportunities
- [Beneficiation Control Node](/Opportunities/Beneficiation_Control_Node) — similar · Opportunities
- [Automated Ore Blending](/Opportunities/Automated_Ore_Blending) — similar · Opportunities
- [Algorithmic Mineral Blending](/Opportunities/Algorithmic_Mineral_Blending) — similar · Opportunities
- [Predictive Mining Site Dispatch](/Opportunities/Predictive_Mining_Site_Dispatch) — similar · Opportunities
- [Autonomous Routing for Mine Operations](/Opportunities/Autonomous_Routing_for_Mine_Operations) — similar · Opportunities
- [Coal Blending API](/CompanyTypes/Integrated_Coal_Preparation_Operators/Opportunities/Coal_Blending_API) — similar · Opportunities
- [Autonomous Loop Tuning](/Opportunities/Autonomous_Loop_Tuning) — similar · Opportunities
- [Refinery Dynamic Deadband Adjustment](/Opportunities/Refinery_Dynamic_Deadband_Adjustment) — similar · Opportunities
- [Geospatial Quarry Load Balancing](/Opportunities/Geospatial_Quarry_Load_Balancing) — similar · Opportunities
- [Hydrological Forecasting for Mines](/Opportunities/Hydrological_Forecasting_for_Mines) — similar · Opportunities
- [Thermal Energy Telemetry for Mills](/Opportunities/Thermal_Energy_Telemetry_for_Mills) — similar · Opportunities
- [Feedstock Blending Optimizer](/Opportunities/Feedstock_Blending_Optimizer) — similar · Opportunities
- [Yield Setpoint API](/Opportunities/Yield_Setpoint_API) — similar · Opportunities
- [Predictive Peroxide Dosing for BCTMP](/Opportunities/Predictive_Peroxide_Dosing_for_BCTMP) — similar · Opportunities
- [Real-Time Kiln Optimization](/Opportunities/Real-Time_Kiln_Optimization) — similar · Opportunities
- [Combustion Optimization API](/Opportunities/Combustion_Optimization_API) — similar · Opportunities
- [Autonomous Cane Feed Control](/Opportunities/Autonomous_Cane_Feed_Control) — similar · Opportunities
- [Freeness Control Agent](/Opportunities/Freeness_Control_Agent) — similar · Opportunities