“The app has artificial intelligence and machine learning algorithms that help identify weeds and disease in the field,” said David Gray, xarvio’s commercial and business development manager in a recent press release. “When you’re in the app you snap a picture of the weed and it tells you with a percentage of confidence what you’re looking at. Over 90% confidence you should consider that accurate.”
BASF acquired xarvio from Bayer as a condition of Bayer’s Monsanto acquisition. The app was developed by Bayer but will be marketed by new owner BASF.
The system was launched in 2017 for the European market and is already used in 90 countries with nearly 60,000 users. With its artificial intelligence platform, as each picture is added to the system, the system gets smarter.
It works by using a remote server that contains more than 150,000 weed and disease images in a massive database for comparison. When you take a picture it quickly compares it to all other images to find a match, which is why adding more pictures increases the accuracy of the app.
“We’ve used several thousand images to train the app and the algorithm gets smarter and more refined with each new image,” Gray said. “The same weed in North Dakota may be slightly different than the Illinois version.”
The disease ID is no slouch either. “The app works at the very early stages of infection when diseases can be harder to identify,” Gray said. “We’ve focused on gray leaf spot and Northern corn leaf blight and we can identify each disease at the early stages when lesions are small and it’s hard to distinguish.”
When you send pictures you’ll get a faster response where internet is more reliable—speed also depends on the number of pictures you send. The app is free and can be found by searching for xarvio in the App Store or Google Play. Greenbook- Farming Tech.
