Why it matters
Without this, energy costs can drift due to inefficient lighting, HVAC cycles, and pump runtimes, leading to higher operating expenses, unpredictable energy bills, and potential power-related safety or reliability issues.
From design to safety
Design
Energy baseline and load modelling
AI reads current and projected lighting, climate, and pump loads, plus energy contracts, to set baseline targets and design compliant, efficient systems with defined guardrails.
- LED driver
- Sub-meter
- Air Handling Unit
- Water chiller
Build
Control integration and calibration
AI validates wiring, sensor placement, sub-meter data streams, and control loops; simulates load profiles to ensure real-world alignment with targets during commissioning and first trials.
- LED driver
- VFD
- Sub-meter
- Air Handling Unit
Operation
Real-time energy balancing
AI analyzes live kWh, PPFD/DLI, and VPD data, adjusting lighting, HVAC, and pump cycles to optimise energy use per kilo while maintaining yield and quality.
- LED driver
- VFD
- Sub-meter
- Pumps
Maintenance
Predictive energy system upkeep
AI monitors performance degradation in LEDs, fans, pumps, and chillers; schedules maintenance before efficiency falls below thresholds affecting kilo metrics and comfort.
- LED driver
- Pumps
- VFD
- Water chiller
Safety
Safety and compliance controls
AI flags electrical and water hazards, ensures safe energy operation, and evidences HACCP/GAP-compliant energy management for audits and certification.
- Sub-meter
- HVAC unit
- Pumps
- Water chiller
Part by part
- 1
LED driver
- Watch for
- Dimming anomalies or overheating that reduce PPFD consistency beyond target ranges over time, affecting uniformity across tiers.
- AI action
- Read real-time current, voltage, temperature; trigger corrective dimming or shutdown with override logging.
- 2
Sub-meter
- Watch for
- Unexpected kWh spikes or phantom loads indicating metering faults or unrecorded consumption across zones.
- AI action
- Compare meter streams to scheduled baselines; raise alert and auto-adjust thresholds with zone tagging.
- 3
Air Handling Unit
- Watch for
- Filter fouling or coil fouling increasing pressure drop and reducing airflow to multiple tiers.
- AI action
- Detect differential pressure changes; open damper sequences and schedule filter maintenance with air-quality notes.
- 4
Pumps
- Watch for
- Vibration or flow rate decline signaling wear or blockage impacting nutrient delivery and DWC/ NFT loops.
- AI action
- Alert maintenance; reduce cycle times to preserve kilo efficiency while maintaining dose integrity.
- 5
Variable Frequency Drive
- Watch for
- Motor current creep or heat indicating motor wear or incorrect fan/pump speed.
- AI action
- Normalize speed setpoints; trigger maintenance if deviation persists and log energy impact.
- 6
Water chiller
- Watch for
- Cooling capacity drift or refrigerant pressure anomalies affecting cooling efficiency across zones.
- AI action
- Rebalance cooling load; initiate service window for refrigerant and compressor checks with temperature trend.
Worked example
Optimising kilowatt per kilo in a compact grow rack
Illustrative scenario with sample data.
A medium-density vertical farm improves energy efficiency by coordinating lighting, thermal mass, and circulation to sustain kilo targets while preserving leaf quality and uniformity across six rack tiers.
Signal in
growArea: 1800 m2; rackTiers: 6; PPFD: 200-300; DLI: 16; targetKilo: 4.2 kg/kWh; subMeter: installed; HVAC: 2 AHUs; pumps: 4 circulators; chiller: 1 unit; VFDs: 6; energyContract: fixed-peak; baselineKWh: 5200/kWh yearly.
System out
Energy platform flags dusk-phase over-use in LED path; recommends dimming with staged ramp to maintain kilo within target; suggests auto-rotate demand-limiting HVAC cycles; generates weekly kilo-per-unit dashboard with multi-zone breakdown.
Human sign-off: Head grower and facility engineer review the suggested changes, approve the energy optimization plan, and sign off for quarterly audit reporting.
How to roll it out
Phase 1
Phase 1
Establish baseline energy map, connect sub-meters, and validate meter-plant alignment across zones.
Phase 2
Phase 2
Implement AI-driven control loops for lighting, HVAC, and pumps; calibrate with first crop data and constraints.
Phase 3
Phase 3
Operational optimization, predictive maintenance, and formal energy reporting for audits and certifications.
Where people stay in charge
- Decisions on plant-level energy targets must be approved by the head grower and facility engineer.
- Only qualified personnel may modify electrical and control system configurations; safety and certification proof required.
- Audits and energy contracts require sign-off from the food safety manager and facilities director.
Questions owners and crews ask
What sensors and meters are needed to monitor energy in a vertical farm?
A vertical farm needs sub-meters for lighting, HVAC, pumps, and chillers, plus environmental sensors for temperature, humidity, PPFD, and VPD to feed the energy AI.
How does AI decide when to dim lights or run pumps?
AI compares real-time kWh per kilo against targets, PPFD/DLI targets, and VPD; it adjusts lighting and pump cycles to stay within benchmarks.
What safety checks ensure electrical and water safety with AI control?
The system flags electrical faults, pump leaks, and coolant issues; all actions are logged for HACCP/GAP audits and signed by the safety officer.
Who signs off on operational energy optimizations?
The head grower and facility engineer review and approve optimization plans; audits and certifications require the food safety manager’s sign-off.