Healthy
Confidence · 99%The model's reference state. Uniform green tissue and regular vein structure — this is what every other classification is measured against.
Home / Pests
Active classifications
Pests detectable through our platform.
We started with the vineyard — one of Portugal's most valuable and most vulnerable crops. Our computer-vision model distinguishes healthy leaves from four major threats, each with its own visual signature.
ESCA
Confidence · 96%A destructive trunk disease. Its tell-tale tiger-stripe discolouration between the veins is exactly the kind of pattern our model is trained to catch early — before whole vines collapse.
Black Rot
Confidence · 91%A fungal disease that leaves dark circular lesions on leaves and can mummify entire grape clusters. Early lesions are small and easy to miss on foot — and easy to spot from above.
Leaf Blight
Confidence · 88%Browning and necrosis that creep in from the leaf edges, cutting the plant's ability to photosynthesise. Detecting the first affected rows keeps it from spreading across the parcel.
Phylloxera
Confidence · 93%The insect pest that devastated Europe's vineyards in the 19th century. It forms distinctive galls on the leaves — a visual signature the platform picks up row by row.
?
New additions
Coming soon…
Why the vineyard first?
Wine is one of Portugal's flagship crops — high value per hectare, and highly exposed to disease. It's where early detection pays back fastest, and where our model can prove itself before expanding to other crops.
How confidence works?
Every detection comes with a confidence score — the model's certainty about what it's seeing. High-confidence detections trigger alerts; borderline cases are flagged for a closer look, so you stay in control.