Croptimistic Technology has reached a major milestone after securing a U.S. patent for its innovative SWAT CAM process, an autonomous in-field imaging and crop analytics solution powered by artificial intelligence and machine learning.
The patented technology, officially titled “Method and System for Automatically Capturing and Extracting Data From Images of Agricultural Field Crops and Weeds Using Machine Learning Processes,” recognises Croptimistic’s advancements in automated crop monitoring and precision agriculture.
The announcement comes as SWAT CAM surpasses one million acres mapped across western Canada during the 2025 growing season — only its third season in commercial operation. The company also confirmed that additional patent approvals, including Canada, are expected in the near future.
Revolutionising Crop Monitoring Through AI
SWAT CAM was developed to eliminate the time-consuming and labour-intensive process of manual plant stand counting traditionally carried out by agronomists and farmers.
Instead of physically visiting multiple locations within a field to count plant populations, the SWAT CAM system operates autonomously during normal spraying operations. Cameras mounted on a field sprayer capture images every 60 feet, generating thousands of field images in real time.
Using advanced machine learning models, the system analyses the images to determine crop and weed populations across entire fields. This provides farmers and agronomists with highly detailed insights into crop establishment variability throughout SWAT MAPS management zones — without requiring additional labour, vehicles, or field scouting.
The technology is helping growers improve decision-making and maximise profitability, particularly as seed costs continue to rise.
Improved Accuracy Across Multiple Crops
For the 2026 season, Croptimistic has significantly improved its cereal crop plant stand counting models for:
- Wheat
- Oats
- Barley
- Canary seed
The company has also introduced beta plant stand counting models for lentils and flax.
According to Croptimistic, the latest models can now predict plant populations within six percent of actual counts, representing a major advancement in precision agriculture accuracy.
“SWAT CAM is more than a single model; it’s a continuously learning system,” said Erfan Khalaji, AI and Data Scientist at Croptimistic.
“Every season’s data makes the next release sharper, more reliable, and more relevant to the growers who depend on it. Every year, more crop models and use cases are added.”
Designed for Practical Farm Operations
Croptimistic says the SWAT CAM hardware was specifically designed for seamless integration into everyday farm operations.
Hardware Director Mark Marianchuk explained that the compact camera system can easily mount onto the booms of virtually any sprayer model while capturing high-quality images throughout spraying operations from dawn until dusk.
The company is also field testing next-generation SWAT CAM hardware, including upgraded camera systems, improved mounting designs, and enhanced onboard processing capabilities aimed at further improving the user experience and agronomic insights.
Supporting Smarter Precision Farming
Beyond plant stand counting, SWAT CAM is also being used to generate variable-rate fungicide maps when satellite imagery is limited by cloud cover or wildfire smoke.
The technology integrates directly with SWAT MAPS zones, enabling farmers to make more precise input decisions while improving efficiency and reducing unnecessary costs.
Systems Engineering Manager Jason Harmon said the company’s goal remains clear:
“Our objective is to provide farmers with low-cost, high-density crop establishment variability mapping.”
In line with Croptimistic’s “Never Obsolete” philosophy, existing users will receive updated hardware systems at no additional cost once the next-generation units are released.
As precision agriculture continues gaining momentum globally, innovations such as SWAT CAM are demonstrating how AI-powered field analytics can help farmers improve productivity, reduce input waste, and make faster, data-driven management decisions.
