Why it matters
Without AI guidance, growers face unpredictable mildew pressure, delayed sprays, and inconsistent recordkeeping, leading to suboptimal disease control, higher input costs, and potential regulatory or quality issues.
From design to safety
Design
Design AI inputs and controls for mildew risk
Define data sources, sensor types, thresholds, and decision logic for mildew risk forecasting; specify interfaces with weather data, leaf wetness, growth stage, and historical infection events.
- powdery mildew risk index
- leaf wetness sensor
- weather station
Build
Build data pipelines and calibration framework
Install sensors, connect controllers, calibrate sprayers and risk thresholds; validate data quality, sensor placement, and spray mapping before first vintage, including label constraints.
- powdery mildew risk index
- leaf wetness sensor
- sprayer calibration
Operation
Operate continuous risk forecast and spray logic
Ingest weather, leaf wetness, and canopy data in real time; run risk models; trigger spray windows, dosage guidance, pre-harvest checks, and generate spray records.
- powdery mildew risk index
- sprayer calibration
- spray records
Maintenance
Maintain sensors and rotation plans with auditability
Service sensors periodically, verify calibration, refresh data quality checks, and adjust fungicide rotation to counter resistance and label changes; keep thorough spray logs and mapping.
- powdery mildew risk index
- leaf wetness sensor
- fungicide rotation
Safety
Ensure safety, labeling, and compliance controls
Document CO2 and confined-space considerations when applying sprays, ensure label compliance and appellation requirements, and maintain HACCP-like sanitation and handling records.
- spray records
- fungicide rotation
- weather station
Part by part
- 1
powdery mildew risk index
- Watch for
- Inaccurate risk spikes from missing leaf wetness data or miscalibrated weather inputs.
- AI action
- AI detects exceeding risk thresholds and schedules spray windows.
- 2
sprayer calibration
- Watch for
- Inconsistent spray pattern or under- or over-application at set speeds.
- AI action
- AI compares flow rate and droplet size against target map and issues calibration tweaks.
- 3
fungicide rotation
- Watch for
- Resistance risk from repeated fungicide within same class.
- AI action
- AI recommends rotation plan and re-sequencing by efficacy and label constraints.
- 4
spray records
- Watch for
- Missing entries or incorrect application timing and product lot data.
- AI action
- AI auto-generates compliant spray events with product, rate, date, code and operator.
- 5
leaf wetness sensor
- Watch for
- Sensor downtime or readings outside calibrated range.
- AI action
- AI flags gaps and triggers maintenance order for sensor replacement.
- 6
weather station
- Watch for
- Missing or unreliable ambient conditions data affecting risk model.
- AI action
- AI recalibrates risk forecasts and adjusts spray windows accordingly.
Worked example
Catching a mildew risk spike before a spray window
Illustrative scenario with sample data.
A boutique estate detects rising evening humidity and leaf wetness increases, signaling a powdery mildew risk spike during a warm spell. AI triggers a preemptive spray window and rotation check.
Signal in
data_source: weather_station; leaf_wetness: 8.6|7.2|8.9; temp: 22.5C; humidity: 72%; block: North Block; growth_stage: veraison; product: sulfur-based; rate: 1.5 L/ha; wind: calm; last_spray: 7 days; Brix: 22; YAN: low; canopy_index: moderate
System out
alert: mildew-risk spike; spray_window: 2026-11-10 14:00; product: Myclozanid; rate: 1.8 L/ha; target_area: blocks A and B; calibration_needed: nozzle_mismatch; required PPE: gloves, goggles; notes: verify label limits and applicator training.
Human sign-off: Winemaker approves the spray window, confirms product choice and PPE, and authorizes an interim rotation adjustment for block A before application.
How to roll it out
Phase 1
Weeks 1 to 2: Baseline
Install sensors, establish baseline risk model, validate data feeds, implement basic alerts, and document integration with the ERP.
Phase 2
Weeks 3 to 6: Calibration
Calibrate sprayers, align rotation plan with label guidance, test reporting templates, and verify spray mapping against block boundaries and ullage records.
Phase 3
Weeks 7 to 10: Operational
Run full risk forecasts during growing season, adapt spray windows and keep spray records updated with lot numbers and operator initials.
Where people stay in charge
- All pesticide and rotation decisions stay with the licensed vineyard manager or winemaker, with label constraints and regulatory approvals recorded.
- No changes to safety protocols or spray timing may occur without documented sign-off from the vineyard manager or winemaker.
- All records must be auditable by regulatory bodies and linked to specific blocks, dates, and product lots.
Questions owners and crews ask
What triggers a mildew risk forecast and a spray window?
The system reads leaf wetness, temperature, humidity, and canopy data to compute risk; when thresholds are crossed, it suggests timing and product choice for the spray window and logs the rationale.
How is sprayer calibration maintained over the season?
Calibration data is collected, compared to target spray map, and adjustments are issued; maintenance records document nozzle performance, flow rate, droplet size, and calibration history.
What records are kept for compliance?
Spray events include date, product, rate, block, lot numbers, operator, weather, and calibration IDs; these feed label and appellation compliance audits and traceability reports.
Who signs off on changes to rotation plans?
The vineyard manager or winemaker signs off; regulatory documentation for rotation and label compliance is filed by the compliance officer or QA lead.