The Future of Food
Embracing AI to monitor crop production and offset impacts of climate change
Climate change is real. No self-respecting farmer would deny it. We may argue about what’s causing it, but no one argues it’s happening.
Scientists around the world are exploring ways to combat climate change in the face of a growing global population and increasing stresses on food security. From agronomists to chemists to biologists and entomologists, the sharpest minds of agricultural sciences are engaged in an arms race to combat one of the most pressing challenges resulting from climate change: drought.
But is it enough?
Dr. Umesh Hodeghatta, Data Scientist and AI expert teaching graduate studies at CPS at the Portland campus.
Seen here with his wife, Rekha.
Umesh Hodeghatta is not an agricultural scientist. He is a data scientist and AI expert teaching graduate studies at Northeastern University’s College of Professional Studies. The first to tell you he doesn’t know the first thing about crop production or agriculture, Hodeghatta has been applying his expertise in data and applied machine intelligence to the study of drought tolerance on blueberry crops in Maine.
The result is a recently published report he co-authored and will present at this year’s International Conference on Precision Agriculture.
Based at Northeastern’s Roux Institute in Portland, Maine, Hodeghatta teaches graduate students about applied machine intelligence and analytics and recently partnered with fellow researchers at the University of Maine to explore how AI could augment data analysis of drought tolerance on blueberries using both manual and drone collected data.
The study was conducted on wild blueberries to assess how they responded to drought conditions.
Different genotypes were tested with biochar and drought treatments, and measurements were taken using a special sensor to collect data on water stress levels. Hodeghatta and his team used various data analysis techniques to find the best predictors for water stress and found tremendous potential in integrating hyperspectral sensing with machine learning for precise water stress monitoring. Their goal now is to secure additional funding to further refine and validate their models to facilitate practical implementation in real-world agricultural settings, fostering sustainable water management practices and enhancing crop security.
For more information about the study, please visit the International Society of Precision Agriculture.
