Drone Agronomists: Multispectral Survey for Precision Farming

There is one critical moment in Crimean viticulture and essential oil farming: when a problem is visible to the naked eye, the harvest is already lost. A drone with a multispectral camera sees a problem 5–7 days before an agronomist sees it.

What is NDVI and why is it needed

NDVI (Normalized Difference Vegetation Index) is a numerical assessment of plant health calculated through two channels: visible red and near infrared. A healthy plant strongly reflects NIR and absorbs red → NDVI is about 0.8. Sick — the opposite, → NDVI drops to 0.3–0.4.

Traditional field inspection — walking the field with a notebook — gives a delayed picture. The agronomist sees yellowed leaves when the bush is already severely damaged. By that time, the yield is reduced by 30–50%, and treatment costs are 3–4 times higher than if the problem had been detected in time.

A drone with a multispectral camera flies over a 50-hectare field in 15 minutes. The onboard processor builds an NDVI map with a resolution of 5 cm per pixel. The agronomist sees the map on the screen in 30 minutes after the flight — and immediately knows which bushes are sick, where watering is needed, and where there is a pest outbreak.

How it works technically

The standard kit:

• DJI Mavic 3 Multispectral — drone with 4 spectral channels (RGB, Red Edge, NIR, Green)
• RTK module — positioning accuracy ±2 cm (important for vineyards with narrow rows)
• Onboard processor — the map is built during the flight, the operator gets a finished result
• Cloud service — automatic upload of maps, comparison with previous flights, anomaly detection

Flight height: 50–80 m (for vineyards), 80–120 m (for open fields). One battery gives 25 minutes of flight and covers 1.5–2 hectares per pass. A full survey of a 50-hectare field therefore requires multiple flights, with operator supervision totalling roughly an hour.

What NDVI sees in Crimean conditions

In the 2025 season, we surveyed several vineyards in the Bakhchisaray and Simferopol districts and a small number of lavender plantations (specific farm names and exact counts are kept under NDA until our public case study is published). Results:

• Early detection of esca (black rot) — detected 7–10 days before the visual manifestation. Treatment at this stage preserves up to 80% of the harvest
• Watering problems — uneven watering identified by NDVI drop in individual sectors. One of the farms saved about 30% of water consumption after adjusting the irrigation system
• Soil salinity — a serious problem for Crimean soils. NDVI shows saline spots where the bush suffers from chlorosis
• Pest outbreaks — spider mites, phylloxera — detected in the early stages by the characteristic NDVI pattern

For lavender, the result was even more impressive: the drone detected the moment of mass flowering with an accuracy of 12 hours — this is critical for harvesting (lavender must be cut at the 50–70% bloom phase, the window is 10–15 days).

Cost and payback

The DJI Mavic 3 Multispectral kit — about 1.2 million rubles. Pilot service from Tavrida Robotics — 15,000–25,000 rubles per 50 ha. The service is more profitable for small farms that don’t need their own equipment.

Payback for a 50 ha farm: 1 season. Saved yield + reduced treatment costs + water saving = 2.5–4 million rubles per season. The drone pays for itself in the first year.

What’s next

In 2026, we are launching a service for the Crimean agricultural sector: monthly monitoring of vineyards and lavender plantations with automatic anomaly detection. The farmer receives a map and a list of recommended actions every 2 weeks.

If you have a vineyard, lavender, or essential oil crops — write to us via our contact form or leave a request through the contacts page. We’ll arrange a pilot flight and show you the state of your fields from a new perspective.

All numerical estimates in this article (drones per hectare, scanning speed, payback, market sizes) are approximate and based on publicly available industry data; verified numbers and case studies will be shared in the investor materials (available after NDA).