Plant health, scored continuously — not checked once a day.
Real, defensible computer vision. One calibrated profile per crop type, trained and continuously improved on your operation's own accumulating labeled data.
Z-01
Butterhead
Z-02
Basil
Z-03
Romaine
N-deficiency trend +4d
Z-04
Kale
Z-05
Arugula
Z-06
Mint
Z-07
Chard
Z-08
Cilantro
Six diagnostics running on every frame.
Per-zone health scoring
Every zone and plant cluster gets a continuously updated health score, so you glance at one number instead of interpreting raw footage.
Leaf color, curl & growth-rate
Tracks color shifts, curl, and growth-rate trends over time — the subtle early signals a once-a-day walkthrough tends to miss.
Early pest & disease markers
Flags visual markers of pests and disease at the earliest detectable stage, before they spread across a rack or room.
Nutrient-deficiency patterns
Recognizes the visual signatures of common nutrient deficiencies and surfaces them as an actionable pattern, not a vague warning.
Growth-stage classification
Automatically identifies seedling, vegetative, flowering and fruiting stages so diagnostic thresholds adapt across the crop cycle.
Multi-crop model support
Leafy greens, herbs, fruiting vegetables, ornamentals and more each get a calibrated diagnostic profile — not one generic model.
Live feeds run continuously. Photos run on schedule.
Where a zone has a connected camera, analysis runs continuously — day and night. Where it doesn't, staff capture scheduled or ad-hoc photos in the mobile app, and both routes flow into exactly the same vision pipeline.
▸ zone Z-03 · source: rtsp://cam-2 · mode: continuous
health 94 → 88 → 82 → 71 (48h window)
▸ zone Z-11 · source: mobile · mode: scheduled 2×/day
growth-stage: vegetative → flowering
▸ threshold auto-adjusted for stage change