Search

AI learns to predict wheat growth patterns, including plant collapse

Boss, M.;Volpi, M.;Roth, L.

Crop Improvement

The wheat swaying in fields near you grows in messy, unpredictable pulses depending on weather and genetics, and better computer models of that messiness could help breeders spot which varieties are prone to toppling over before harvest.

Wheat plants don't grow at a steady pace, they speed up and slow down depending on weather, soil, and genetics, and sometimes they even fall over, a problem called lodging. Scientists trained a machine learning system to predict the whole range of possible growth patterns a wheat plant might follow, not just the average case, using simulated data tuned to match real Swiss weather stations and field trials. They then built new statistical tools to check whether the AI was correctly capturing rare but important outcomes, like plants that collapse, not just the typical growth curve.

Key Findings

1

The team used neural processes enhanced with normalizing flows to model full probability distributions of wheat height trajectories over time, not just single predicted values.

2

Synthetic training data was calibrated against real Swiss weather station records and the FIP1 wheat field trial dataset for realistic evaluation.

3

Two new evaluation metrics, Sig-MMD and CSig-MMD, were used to test whether models correctly captured rare outcomes like lodged (collapsed) wheat trajectories, not just typical growth patterns.

chevron_right Technical Summary

Researchers built an AI system that predicts how tall wheat plants will grow over the season, including rare events like plants collapsing under their own weight, by learning from patterns in weather and genetics data.

description

Abstract Preview

Original paper

Neural Processes with Normalizing Flows for Wheat Height Estimation

In this work, we investigate modeling plant traits over time using neural processes, a class of machine learning models that learn distributions over functions. Plant growth is an inherently stocha...

open_in_new Read full abstract

Abstract copyright held by the original publisher.

hub This connects to 9 other discoveries — Wheat crop-improvement, phenology, climate-adaptation 5 related articles

Species Mentioned

Was this useful?

mail Weekly plant science — one email, Saturdays.

Share: X/Twitter Reddit
arrow_forward Next Discovery

Gene editing removes 97% of celiac-triggering proteins from bread wheat

It could mean that people with celiac disease — roughly 1 in 100 worldwide — may one day safely eat bread made from real wheat, without sacrificing the taste...

eco Wheat
Species
Wheat

Wheat is a group of wild and domesticated grasses of the genus Triticum. As cereals, they are cultivated for their grains, which are staple foods around the world. Well-known wheat species and hybrids include the most widely grown common wheat, spelt, durum, emmer, einkorn, and Khorasan or Kamut....