Agriculture and AI: Farming with Smart Data

Michaela Meyer did not grow up in the countryside, nor did she ever sit on a tractor in her life before her current job. Nevertheless, the 32-year-old is now he

Agriculture and AI: Farming with Smart Data

Michaela Meyer did not grow up in the countryside, nor did she ever sit on a tractor in her life before her current job. Nevertheless, the 32-year-old is now helping to shape the agriculture of the future: as a smart data technologist at the American agricultural machinery manufacturer John Deere, for example, she is currently helping farmers distribute manure efficiently in the field - with the help of artificial intelligence (AI). Together with her team, the mathematician has developed algorithms for sensors that accurately detect the ingredients in the organic fertilizer in real time. "If you know what's in the manure, you can plan pretty precisely with how much you have to spread at the different points of the field," says Meyer.

However, it is highly complex to find out exactly what the slurry is made of. For farmers, the AI-controlled sensors from Meyer's laboratory are therefore an important tool on the way to a sustainable and economic future. "Agriculture often has to deal with the accusation that it pollutes the environment and is responsible for residues in food," says a company spokesman from John Deere Germany. "That is why it is important to reduce the use of fertilizers without reducing yields."

Intelligent technology also helps farmers to automatically detect weeds in fields and to use plant protection products specifically only in these places. The work of software developers, mathematicians, statisticians and engineers saves money and plant poison in the long run. "Our team includes people from many different scientific disciplines," says Michaela Meyer. "This is important because every employee has different skills.“ The challenge when working with AI: Often when experimenting, something comes out in the end that you didn't expect. "Then it's good if you have people who have different ideas about what could be the reason," says Meyer. That's why she hopes that even more people from areas other than software development will be interested in careers with AI. From their point of view, one thing is especially important: that applicants can be enthusiastic about statistics. "For the most part, artificial intelligence is statistics," says Meyer.

She also considers specific courses of study such as "Data Science", which you can study at the Technical University of Dortmund, for example, to be useful. Markus Pauly teaches and researches here. The professor of Mathematical Statistics and Industrial Applications advises students to specialize already during their studies if they know that they are aiming for a career with AI in agriculture, for example. In case study projects and supervised industrial internships, students can work with companies, gain important experience and expand their network.

The Dortmund scientist is currently involved in a project with AI in agriculture himself. His team is working with the agricultural machinery group Claas, the German Research Center for Artificial Intelligence and the University of Osnabrück to investigate how combine harvesters will drive autonomously across the field in the future or automatically recognize how they can harvest optimally using cameras. Unlike Meyer, Markus Pauly gained experience in agriculture at an early age: his great-grandparents had a farm in the Eifel. "As a child, I already helped with the harvest myself – but I could also support agricultural projects with my statistical and AI know-how without this knowledge," says Pauly.

What technique can farmers use at all?

Agricultural expertise is therefore not a must to work with AI in agriculture. Nevertheless, the large interdisciplinary teams always need someone who is familiar with agricultural topics – and knows which intelligent technology farmers can use at all. So an agricultural farmer has a good chance of getting a job in the field of AI in agriculture. However, he has to familiarize himself intensively: "You have to have a good feeling for relevant data, have basic programming skills, and have the ambition to acquire the methods and pitfalls of machine and statistical learning," says Pauly.

Updated Date: 04 December 2021, 00:01

You need to login to comment.

Please register or login.