What does artificial intelligence on the farm mean?
Let’s go back a bit. In computer science, algorithms – mechanics in fact – solve well-defined problems: in a spreadsheet, you can thus calculate the sum of the columns. That’s easy. For more complicated issues, you have a set of parameters, like a bunch of little sliders. One or more settings provide the correct solution to a problem. Except that you have to find how to adjust all the little sliders. It’s very complicated. Artificial intelligence, often, these are techniques that will try to navigate in these spaces where there are many possible settings and move fairly quickly towards a good solution.
Can you illustrate it in the case of agriculture?
For example, when should a cow be inseminated? This is not obvious. Often the breeder does this by visual inspection. He checks if the cow has a typical heat behavior. Except that 30% show no heat. We will therefore have to look at other cursors: its temperature, its activity… A whole series of variables. Artificial intelligence, by looking at cows that we know and have followed, will learn and determine, based on these criteria, if there is a good chance that the cow is inseminable at a given time. .
Another example ?
In viticulture. The great fear of the winegrower is mildew (a disease that particularly affects the vine, editor’s note). As soon as I see an infected leaf, do I spread a product, even if it means putting on too much? Or am I waiting, at the risk of waiting too long and it’s over? Again, there are learning mechanisms for knowing when to intervene.
For some, the dependence on technology is an additional enslavement, even the beginning of the end of the peasants. What do you think ?
Before answering you, I invite those who are interested to consult the white paper on digital in agriculture prepared by Inrae and Inria (the two national research institutes, one on agronomy, the other on digital, editor’s note). The idea is to relieve the farmers, without deciding for them. But give them elements to decide better and waste less time on certain tasks that are perhaps not the most exciting. Rather than asking them to observe lots of curves, which is not necessarily their job, it’s going to be a question of saying “I think that your cow, you have to inseminate her because three days ago, she didn’t sleep well, his temperature was a bit high”. I tell you anything, it’s an example. AI is not there to replace the farmer or enslave him. She’s there to free up time for him to do what he loves and make more strategic decisions.
How much of these innovations can be freely accessed?
It is not won. When we take the example of the data that we all have. At Google or others. Basically, they belong to Google. They are not public. When farmers have connected tractors, milking robots, etc., it is the companies that sell this equipment that collect the data. They give back a bit in the form of visualizations. But they are the ones who have everyone’s data and can do much more interesting and strategic things. There have been efforts, an attempt at an agricultural data portal. I’m not sure it was a big hit.
The 6th edition of SmartAgri, dedicated to artificial intelligence, takes place on Thursday March 24, from 9 a.m. to 4 p.m., at the Pommerit high school, in La Roche-Jaudy. Full program and registration on smartagri.bzh. The white paper on digital technology in agriculture is available on the Inrae and Inria websites.
We would like to give thanks to the author of this short article for this remarkable content
Lannion – SmartAgri: “Artificial Intelligence is not there to replace the farmer or enslave him”
You can find our social media profiles here , as well as other pages related to them here.https://www.ai-magazine.com/related-pages/