A new app from the University of Maryland makes diagnosing soybean diseases quick, easy and reliable
Image Credit: Edwin Remsberg
What if farmers could diagnose crop diseases early with just a cellphone?
Well, now, soybean farmers can, with the help of SoyAI, a new, artificial intelligence-powered diagnostic tool from the University of Maryland's Digital and Precision Agriculture Lab. The team has released the first phase of the AI-enabled diagnostic tool for field testing. This initial version supports the identification of several key soybean health conditions, including Bacterial Blight, Cercospora Leaf Blight, Downy Mildew, Frogeye Leaf Spot, Healthy, Potassium Deficiency, Soybean Rust, and Target Spot.
With SoyAI, if a farmer sees a plant that doesn’t look quite right, they can snap a photo, upload it to the app, and immediately know what kind of disease they’re dealing with. No more sending samples to a lab and waiting for the results. The diagnosis is in the palm of their hands.
“We are excited to release this first version of SoyAI, which is browser-based, and we will be releasing it as an app for Android and iPhone soon,” said Precision Agriculture Specialist Hemendra Kumar, director of the lab and leader of the effort to develop SoyAI. “We want to encourage users to explore the platform, share it with colleagues and collaborators, and provide feedback that will help improve the app and guide future enhancements.”
A fast, in-field diagnosis provided by SoyAI will help farmers manage emerging issues with their crops early, reducing the amount of chemicals required to treat disease, which can save money and promote more sustainable, productive crop production.
The app uses advanced AI and deep learning technologies to analyze RGB images of soybean plants and identify eight major soybean diseases and disorders. Once a photo is input, the app displays the name of one or more diseases that may be present, along with a percentage representing the algorithm’s confidence level in its prediction.
Because it is the first app of its kind, thousands of photos of each soybean disease did not exist for training the AI algorithm, which means there is room for improvement in its prediction accuracy. Field evaluations will help validate and further refine the system before its broader public deployment. As more users upload images, Kumar intends to increase the confidence level of every diagnosis by having agricultural experts verify each image and predicted disease.
The next phase of the project will focus on expanding the SoyAI platform by collecting additional field data to improve the accuracy of existing disease classifications and incorporating new soybean disease classes. Looking ahead, Kumar envisions integrating SoyAI with drone technology and edge computing, enabling farmers to detect, identify, and monitor soybean diseases remotely in real time. This advancement will provide rapid, field-scale diagnostics and support timely, data-driven crop management decisions.
Kumar is also building other diagnostic tools, like one for identifying diseases in corn, including the emerging disease tar spot, which is becoming a significant challenge for farmers.
SoyAI is available at https://soyai.life/login, and will be available in App stores very soon.
