Why it matters
Without AI, uneven germination, clumping in trays, canopy height variation, untimely harvests, and mould risk can escalate, driving crop loss potential and compromising safety and audits.
From design to safety
Design
Define growth sensing plan
Identify microgreens species, target canopy height, germination window, tray geometry, sensor mix, and data schema to inform layout, controls interfaces, and future calibrations. Specify required data latency, archival policy, and integration with the building management system for alarms.
- seed tray
- canopy height sensor
- mould detection camera
Build
Install sensing stack and calibration
Assemble trays and sensors, route cable harnesses, calibrate depth and moisture mappings, and run first crop trials to validate data integrity and repeatability. Install wall-mounted data hubs and PoE switches for reliable connectivity.
- seed tray
- mould detection camera
- environmental sensor module
Operation
Continuously monitor growth signals
Real-time tray status, density patterns, canopy rise, mould flags, harvest readiness, and data dashboards drive adaptive regimens and alerts. Enable automated drift corrections and push notifications to mobile devices.
- canopy height sensor
- harvest day timer
- environmental sensor module
Maintenance
Calibrate and clean sensors
Scheduled recalibration, tray sanitation, sensor inspections, and drift checks ensure long-term data quality and consistent growth outcomes across batches. Maintain logs of cleaning agents and contact times.
- canopy height sensor
- harvest day timer
- environmental sensor module
Safety
Food safety and worker safety checks
Record HACCP mould findings, electrical safety checks, PPE adherence, and incident logs linked to sensor-driven alerts and corrective actions. Include lockout/tagout procedure reviews and allergen traceability notes.
- mould detection camera
- environmental sensor module
- seed tray
Part by part
- 1
seed tray
- Watch for
- Signs of uneven germination or clumping across trays.
- AI action
- Trigger tray-level germination checks and adjust sowing map.
- 2
sowing density sensor
- Watch for
- Deviation from target density within tray cells.
- AI action
- Recalculate sowing plan and alert head grower.
- 3
canopy height sensor
- Watch for
- Uneven canopy height across rack tiers.
- AI action
- Adjust light recipe and irrigation to even growth.
- 4
harvest day timer
- Watch for
- Predicted harvest day drift beyond scheduled window.
- AI action
- Update harvest calendar and notify pack-out team.
- 5
mould detection camera
- Watch for
- Visible mould onset on microgreens leaves.
- AI action
- Trigger mould mitigation workflow and sanitation crew.
- 6
environmental sensor module
- Watch for
- RH/temperature spikes or dew point issues.
- AI action
- Adjust VPD setpoints and airflow to stabilize conditions.
Worked example
Monitoring a Microgreens Growth Stage in a 600 m2 vertical farm
Illustrative scenario with sample data.
A new batch of microgreens requires precise germination, uniform canopy, and timely harvest while preventing mould during propagation. The system tracks tray counts, density targets, and canopy trajectories across three rack tiers to sustain uniform leaf development and rapid sanitation readiness.
Signal in
trayCount=240, seedVariety=radish_microgreen, targetDensity=48 seeds/cm2, targetCanopyHeight=12 cm, harvestWindow=Day 7–9, RH=60%, Temp=22C, CO2=1000 ppm, DLI=12 mol/m2/day, batchId=MX-042, user=QA supervisor, germinationWindow=48h, traySpacing=5 cm, latency=2 s
System out
warning: uneven germination detected in 15 trays; action: adjust sowing map and resume germination checks; alert: harvest schedule recalibrated to Day 8; mould risk score elevated in Zone 3 trays; action: initiate sanitation protocol; new log entry created; pad of environmental drift flagged, triggering VPD recalibration.
Human sign-off: Head grower approves revised sowing map and harvest calendar; facility engineer verifies sensor calibration, run wiring integrity check, and updates QA records; safety officer signs off on sanitation plan and allergen traceability updates.
How to roll it out
Phase 1
Weeks 1 to 4: Baseline
Establish sensor readings, data flows, and initial calibration for trays and canopy height; document baselines in the design dossier.
Phase 2
Weeks 5 to 12: Calibration pass
Refine density targets, height thresholds, and harvest timing with iterative trials; lock validated configurations in the operations manual.
Phase 3
Weeks 13 onward: Full operation
Operate with integrated dashboards, automated interventions, and documented HACCP records; generate traceable records for audits and continuous improvement.
Where people stay in charge
- Decisions changing harvest windows must be signed by the head grower and facility manager.
- Mould intervention actions must require a fungi control plan approval from the safety officer.
- Sensor calibration changes must be reviewed and approved by the CEA engineer and QA lead.
Questions owners and crews ask
What sensors monitor germination and tray readiness in microgreens?
The system reads tray moisture, seed-to-soil contact, and early germination indicators; decisions adjust sowing patterns and monitor germination progress. It logs seed depth, seed spacing, and tray temperature history to confirm uniform germination across zones.
How is canopy height tracked and used to adjust growth?
Canopy height readings feed light and irrigation controls; adjustments are made to maintain uniform canopy across tiers. It correlates with DLI targets, air moisture, and substrate moisture to stabilize leaf thickness.
What happens if mould is detected on microgreens?
Mould alerts trigger sanitation workflows, pest and mould control checks, and temporary harvest adjustments to protect safety. The system flags affected trays, timelines sanitation actions, and updates HACCP records.
Who signs off on calibration changes to the sensing network?
Calibration changes require sign-off from the CEA engineer and head grower, with QA records updated.