Predictive Analysis
Plantix
AI can be anywhere these days—even on your phone. Plantix, created by German company PEAT, uses deep learning and image recognition to detect plant species and diseases, providing its users with simple solutions. It also provides a forum for farmers to discuss challenges with each other.
Using imaging recognition to diagnose plants is nothing new, but Plantix is the first to be successful with both accurate image recognition technology and a significant user base. It currently has 1 million monthly users, with 3-5 thousand from the US. It’s also been covered by news sources such as the BBC and The Better India.
“We will try our best to democratize knowledge about farming, suitable products, good agricultural practice with our app and its features,” says Korbinian Hartberger, localisation and communication for PEAT.
Plantix may be the only app of its kind, but there are others that do similar work. These apps can be helpful for farmers looking to incorporate more kinds of affordable or free technology.
Plantvillage is part of a research and development unit at Penn State University that focuses on smallholder farmers to provide information through a user moderated Q and A forum. Prospera collects, digitizes and analyzes data from farms to create plans that optimize production and growth of crops. WeFarm is a farmer to farmer platform used primarily in Africa that allows smallholder farmers to communicate with one another about crops and business without internet. Instead, it uses SMS, that categorizes the comments made through machine learning technology.
To learn more about Plantix, visit the website here.
Senthold Asseng, Ph.D. discusses
the use of AI and predictive analysis.
Mathews L. Paret, Ph.D., an associate professor of plant pathology, focuses on vegetables and ornamentals. His research specifically looks at etiology and epidemiology of plant diseases. He explained the process farmers use to look for pathogens and other pest problems.
Farms need to be monitored to prevent an outbreak of diseases. Traditionally, small farmers will monitor their crops themselves and large farmers will hire scouting agencies to do it for them. Rather than going through every acre of land, scouts use random sampling to visit certain areas for monitoring. Early detection of these diseases is key to prevent crop loss, and AI can help.
Paret describes how the scouting process could be improved with drones by equipping them with multispectral sensors. This process would be able to detect abnormalities/irregularities anywhere in the field so any further outbreak could be stopped.
With drones, Paret says, one of the main detractors is the price. The drone, the sensors and the cameras are all expensive. It has to be cost effective for a farmer to use them. This technology also requires FAA (Federal Aviation Administration) approval and license to be flown, someone who knows how to fly them and another person who is able to understand the data collected.
“The technology is there, and what is possible is amazing,” Paret says. “But making it actually practical is harder.”
To learn more about Mathews L. Paret and his research, visit his IFAS page and his lab website.
