Graph-Based AI Model Predicts Potato Crop Water Needs with High Accuracy in Canada
Bioengineer.org · October 24, 2026
The report in brief
Researchers at the Canadian Center for Climate Change and Adaptation in St. Peter’s Bay developed an AI model to predict potato crop water needs using data from three Prince Edward Island fields over the 2023 and 2024 growing seasons. The model achieved accuracy scores of 0.984 to 0.997, but was not best on every measure and used more computing memory than the other models tested. The researchers said further testing is needed across more years and potato varieties.