AI App Reads Wheat Traits from Photos
South Dakota State researchers' WheatAI platform is in development and tentatively slated for free public release in 2028.
Updated on Nov. 13, 2026 in Botany

Live Poll
Do you think AI tools for farming will mostly benefit everyday food buyers?
South Dakota State University researchers are developing WheatAI, an AI-powered web application that analyzes wheat traits from photographs. The platform is in prototype development and is tentatively scheduled for public release in 2028 at no cost, so it is research-stage rather than available today.
Why it matters
Manual wheat phenotyping, the measurement of plant traits, is slow and subjective, and WheatAI aims to make it faster and more consistent so researchers can evaluate far more samples. Its image-based approach reflects a broader move in plant science toward using computer vision to replace labor-intensive field measurements.
The platform was built from smartphone and drone images of winter wheat taken five times between May and July at SDSU research farms near Pierre and Brookings and producer fields in Winner and Hayes. Its measured accuracy against manual methods has not been reported in the source, and the yield and disease estimates remain under development.
The players
South Dakota State University
A public land-grant university in Brookings with research farms across South Dakota, offering 230 majors, minors and specializations and reporting $86 million in fiscal 2026 research expenditures.
Sunish Sehgal
An SDSU professor who joined the university in 2014 after working as a research scientist at Kansas State and leads the WheatAI development.
Maitiniyazi Maimaitijiang
An SDSU researcher who co-leads the development of the WheatAI prototype with Sehgal.
The details
The system uses AI-powered image analysis to extract measurements of wheat traits directly from photographs, rather than having a person measure each plant by hand. Researchers collect images from smartphones and drones at repeated points in the growing season, which gives the software multiple views of the same crop over time. The platform is designed to estimate yield and assess diseases such as Fusarium, a fungal pathogen that damages wheat. Several graduate students worked with the lead researchers to build the tool.
Timeline
2022: Professor Sunish Sehgal began taking photos to develop the AI tool.
2025: The team began developing a working prototype with Maitiniyazi Maimaitijiang.
2028: WheatAI is tentatively scheduled for public release.
The Tech Race
The WheatAI effort places South Dakota's public university in the growing field of image-based crop phenotyping, where researchers are trying to replace slow hand measurements with computer vision. The source does not name a specific competing program or prior milestone for comparison.
The platform is free to use once released, and the team says it could help farmers make earlier and more accurate decisions about crop fate. It is not available yet, and the source gives no firm date beyond the tentative 2028 target.
The takeaway
The key trajectory is whether image-based phenotyping can match manual measurements in accuracy once validated. Watch for published accuracy results and the trait additions the team plans before the tentative 2028 public release at wheatai.net.
What happens next
The team plans to add 10 to 12 traits before the platform's tentatively scheduled public release in 2028.
Further reading
For more on how plant traits are studied, see Botany coverage on the site.
More information
Visit the WheatAI wheat phenotyping web application for details on the platform.
Live Poll
Do you think AI tools for farming will mostly benefit everyday food buyers?



