Drones and AI in Agriculture
From the Ag Information Network, I’m Bob Larson. Artificial intelligence is changing how we see the world and how we look at the crops in our fields.
While agriculture has long used light reflection to gauge basic plant health, new AI tools are taking this technology a step forward.
Associate Professor of computer sciences at the University of Illinois, Mohammed El-Kebir (Moe-HAHM-id El-kuh-beer) is using drones and AI to decode plant genetics from the sky, hunting for hidden, highly heritable traits that he says could build better, stronger crops for the future …
El-KEBIR … “So first we had to ensure that the vegetation indices that we find actually come from a function class that is differentiable in its parameters. So, this could be a linear model. It can be a neural network. It can be anything. So, it's a very mild restriction. And then the second thing that we had to do was the heritability computation itself also had to be differentiable. That one is a little bit tricky, but we managed to, you know, develop an algorithm that accomplishes both requirements. And because of that, we could actually optimize heritability and find new traits that maximize heritability.”
In short, light reflection, artificial intelligence, and drone-based data collection are used to identify highly heritable plant traits.
These traits can then be linked to DNA snippets for further study and potential manipulation to improve crops.
