Page 26 - AWA Vol.41-No.8

HORTICULTURE

CultiWise autonomous crop monitoring system

“ScoutAI” recognized as future-farming innovation

CultiWise ScoutAI, an autonomous multi-sensor crop
monitoring system, has been selected as finalist for the
DLG Agrifuture Concept Winner Award. Developed by
Czech startup Skymaps, the AI-based system automatically
integrates satellite, drone, sprayer and field data— giving
farmers earlier warnings of crop issues while simplifying
workflows and reducing unnecessary inputs. The award
recognizes forward-looking innovations that are intended
for market launch in 5-10 years.
By integrating what were previously standalone sensor
tools, ScoutAI is automating not only crop surveillance but
also key field decisions. Traditional field scouting methods
rely on manual observation and take time to reveal emerg-
ing issues. Whereas ScoutAI identifies problems earlier, sensor systems, the farmer is presented with clear over-
enabling proactive, targeted interventions before they es- view and machine-ready prescription maps, which can be
calate.
transmitted directly to compatible field equipment.
Precise, needs-based applications deliver substantial re- “For the farmer, it’s like having a team of scouts working
source efficiencies, yielding measurable benefits, includ- around the clock, flagging issues before they spread, and
ing up to 50 percent reduction in fertilizer, herbicide and pinpointing exactly where and how to treat. This frees up
water use, and more than 60 percent savings in pesticide valuable time for farmers to focus on higher-value deci-
volumes. At the same time, automated detection and ma- sions,” says Kornel Cziria, Chief Technology Officer, Sky-
chine-ready prescriptions reduce scouting time and labor maps.
giving farmers virtual eyes across every hectare and the
peace of mind that every field is under constant watch and Development and Global Adaptability
control. With a potential launch in five years, ScoutAI is currently
From stand-alone sensors to fully integrated field intel- under development by a team of 30 software and hardware
ligence engineers. More than 15 different sensor systems from 10
common brands are being tested, including automatic bat-
ScoutAI connects tools that were never designed to work tery changes for drones and smart drone docking stations.
together. Unlike most digital tools — such as drone data,
sprayer logs, and weather apps—which often operate in Several large-scale farms are taking part in the pilot
isolation and therefore lack interoperability, decision-mak- phase. “The initial feedback is promising. Farmers clearly
ing platforms typically rely on single data sources and are value the ability to understand their field conditions from
tied to specific brands. Farmers therefore contend with a multiple angles and compare them in real time. That pro-
mix of incompatible tools. cess is time-consuming and relies on the farmer’s experi-
ence,” adds Cziria.
ScoutAI works similar to a smart security system for
the farm—only instead of guarding buildings, it protects ScoutAI’s intelligence is built on globally trained AI mod-
crops. Trained to detect suspicious changes, like bare els that learn from satellite, drone, and sprayer imagery
patches and weed growth, AI models continuously scan across diverse regions and crop types, making it suitable
daily satellite images of the crop. To verify initial findings, for farms of all sizes.
ScoutAI automatically triggers farm drones — stationed in As a modular platform, CultiWise ScoutAI can be tailored
smart docking hubs around the farm — to take off and in- to any production system worldwide or equipment brand
vestigate up close. — eliminating vendor lock-in.
The drones capture high-resolution images, even zoom- “ScoutAI is a textbook example of sensor fusion. The
ing in on individual leaves, and use onboard AI to analyze process collects and processes data from multiple sensor
the footage in real time during flight. In turn, cameras systems that observe the same field conditions but from
mounted on the sprayer boom provide live in-field con- different angles. This fusion of diverse data streams deliv-
text, with onboard processors checking crop conditions on ers a far more reliable and nuanced understanding of crop
the go. Meanwhile, in-crop sensors contribute by monitor- health than any single sensor could ever provide,” adds
ing temperature, humidity, and pest outbreaks. Cziria.
Data collected from the diverse sources is combined by ScoutAI builds on the CultiWise cloud platform, including
the cloud-based CultiWise platform, which then recom- Zoneye, which was launched in September this year — an
mends to the farmer corrective options, including pre- AI model that can detect 37 weed species with centime-
scription maps for variable and targeted treatments like tre-level precision, even in green-on-green conditions.
spot spraying. The overall result is presented within min- Zoneye and ScoutAI will be presented at Agritechnica
utes in a single, easy-to-read dashboard. 2025, 9 –15 November, at the CultiWise stand in Hall 9,
“Instead of manually piecing together data from multiple Stand H35. Circle 24 on enquiry card
24 Vol. 41 No. 8