Why it matters
Without this, mis-timed harvests, inconsistent phase separation, and equipment drift can cause inferior oil quality, unplanned downtime, and missed certification targets. These gaps complicate throughput, cost control, and brand integrity.
From design to safety
Design
AI informs grove and mill layout decisions
AI models evaluate decanter configurations, feeding paths, water dosing plans, and energy requirements; they simulate oil yield, emulsion stability, and waste handling across two-phase and three-phase systems; outputs guide capex, plant layout, and utility integration across irrigation headers and pump rooms.
- two-phase decanter
- three-phase decanter
- wash water reservoir
Build
Installation standards and calibration
Sensor networks, load cells, flow meters, PLCs, and SCADA interfaces are installed; calibration baselines establish bowl speed, differential speed, water dosing, feed pump alignment, and verification tests with standard reference oils.
- two-phase decanter
- three-phase decanter
- decanter feed pump
Operation
Real-time optimisation during extraction
AI reads phase separation metrics, feed rate, water addition, motor torque, emulsion index, oil temperature, viscosity, and solids content; it tunes speeds and dosing to maintain target free acidity and K232/K270 while maximizing throughput and minimizing energy consumption.
- centrifuge motor
- PLC controller
- wash water reservoir
Maintenance
Predictive maintenance cues and records
Vibration, temperature, torque, bearing wear, seal integrity, and gear mesh analysis trigger inspection windows; service orders reference batch IDs, calibration history, grease cycles, and parts replacement in lifecycle logs with replacement intervals.
- centrifuge motor
- two-phase decanter
- three-phase decanter
Safety
Safety-critical monitoring and approvals
Control room dashboards, fault logs, PPE compliance, lockout-tagout procedures, and batch traceability records feed into safety reports; calibration and maintenance sign-offs rest with qualified personnel according to IOC trade standards.
- PLC controller
- two-phase decanter
- three-phase decanter
Part by part
- 1
two-phase decanter
- Watch for
- inconsistent bowl balance causing solids carryover
- AI action
- read real-time bowl speed and torque; adjust differential speed accordingly
- 2
three-phase decanter
- Watch for
- phase separation drift reduces oil clarity
- AI action
- monitor phase split metrics and modify feed rate in real time
- 3
decanter feed pump
- Watch for
- pulsed flow causing surge
- AI action
- stabilize pump speed to smooth flow into decanter entry port
- 4
wash water reservoir
- Watch for
- water quality drift affecting emulsion
- AI action
- adjust water addition based on emulsion readings in real time
- 5
centrifuge motor
- Watch for
- bearing wear or overheating
- AI action
- read motor temp and vibration; alert for service needs soon
- 6
PLC controller
- Watch for
- control loop drift or faults
- AI action
- log faults and auto-tune control parameters for maintenance today
Worked example
Catching a parameter drift in decanter balance
Illustrative scenario with sample data.
A mid-harvest batch shows rising solids carryover despite stable setpoints, suggesting evolving feed quality and balance. AI cross-references temperatures, torque, and phase metrics to re-balance operations.
Signal in
systemStudy: extractionLine; dataPoints: bowlSpeed=1900 rpm, differential=120 rpm, feedFlow=1800 L/h, waterAddition=5%, oilTemp=32 C, emulsionIndex=0.52, viscosity=1.8 mPa·s, torque=120 Nm, bearingTemp=68 C, K232=0.18, K270=0.04, freeAcidity=0.15%, batchID=B-1048
System out
alert: balance drift detected; adjust differential speed to 160 rpm; revise feed rate to 1750 L/h; log for QA review; recommended maintenance window scheduled; batchID B-1048 noted.
Human sign-off: Mill master reviews the suggested changes and approves the re-tuning plan, documenting the decision for traceability.
How to roll it out
Phase 1
Weeks 1 to 4
Baseline readings collected; AI learns target ranges for two-phase and three-phase decanters; initial control rules stored.
Phase 2
Weeks 5 to 8
Pilot adjustments tested on controlled lots; performance metrics captured; confidence thresholds updated for production use.
Phase 3
Weeks 9 to 12
Full integration with batching and QA; dashboards deliver continuous monitoring and alerting for operators and management.
Where people stay in charge
- Decisions on critical process changes must be signed off by the mill master or estate manager after QA review.
- All safety-related adjustments require mandatory validation by the qualified person and compliance with IOC trade standards.
- All data and changes must be captured in the production records and certification logs to ensure traceability.
Questions owners and crews ask
How does AI optimise decanter settings during harvest campaigns?
AI analyzes real-time sensor data from bowl speed, torque, feed rate, water addition, and emulsion quality to adjust speeds and dosing for stable separation and target oil quality.
What data should I collect to start implementing this?
Capture bowl rpm, differential rpm, feed flow, water addition percentage, emulsion index, K232, K270, free acidity, oil temp, viscosity, torque, bearing temp, batch IDs.
Who signs off the critical adjustments in production?
The mill master and estate manager review the proposed changes, ensuring alignment with quality targets, regulatory requirements, and certification scope.
How do I prove compliance for PDO or IOC standards with this system?
Maintain audit trails of control decisions, calibration records, laboratory results, and chain-of-custody; ensure IOC protocols and PDO specifications during inspections.