Artificial intelligence can significantly shorten the plant breeding process while reducing fertiliser use and the human labour needed in agriculture. According to experts at the HUN-REN Centre for Agricultural Research (HUN-REN ATK), it is essential for Hungarian farmers to become more open to AI-based solutions, as these can bring significant time and cost savings even for smaller farms.
Artificial intelligence has by now appeared in almost every scientific field, and agriculture is no exception. Further west, technologies are already being used on huge arable fields where only one out of ten combine harvesters is driven by a person, while the others autonomously follow its movement. Although current regulations in Hungary do not yet allow this, more and more AI-based tools are supporting agricultural work in the country.
The use of drones is also governed by strict rules; nevertheless, they are now used to survey drought damage from the air and help determine the exact nutrient needs of individual plots of land.
“AI provides the greatest help in decision-making,” emphasised Roland Hollós, research fellow at the Agricultural Institute of the HUN-REN Centre for Agricultural Research (HUN-REN ATK MGI). According to him, after analysing processes and data, artificial intelligence can make decisions quickly, even within minutes, that previously took weeks. As a result, identifying plant diseases has also become much faster and more accurate, whereas this work used to be done far more slowly and exclusively by people.
Artificial intelligence could revolutionise plant breeding and Hungarian agriculture
In plant breeding, time is a decisive factor, stressed Nándor Fodor, director of HUN-REN ATK MGI. Currently, it takes about ten years for a new plant variety to reach the market, but with artificial intelligence this time could be reduced to six or seven years. “At first, three or four years may not seem like a big difference, but considering that fewer crossings and less chemical use are needed in the meantime, artificial intelligence can bring significant cost savings in breeding,” the director added.
According to Roland Hollós, researcher at HUN-REN ATK and the institution’s AI ambassador, AI can be applied in all plant breeding processes, as it can analyse complex datasets containing genetic, environmental and phenotypic data. By processing huge amounts of genetic information, machine learning algorithms can recognise the patterns that determine plant traits and predict expected outcomes with high accuracy. As a result, plant varieties better adapted to specific climate and soil conditions can be developed faster and more efficiently.
Although more and more companies offer AI-based services to smaller farms as well, many Hungarian farmers still approach new technologies with caution. “Artificial intelligence has to contend with decades-old habits. Many farmers decide on fertilisation based on experience instead of relying on soil test data and reliable professional recommendations. Yet with the help of AI, the amount of chemicals applied can be reduced, which benefits not only the soil and consumers but also the farmers,” Roland Hollós explained.
Last autumn, the HUN-REN Hungarian Research Network launched the AI 4 Science programme, which aims to provide its researchers with all the AI-based tools, services and advice they need for more efficient scientific work. The initiative has been very successful among HUN-REN ATK researchers too, who have already used AI for prompt writing, analysing scientific papers and various modelling tasks.
“Artificial intelligence is the future of Hungarian agricultural research as well,” Roland Hollós emphasised. He added that although researchers are increasingly open to using AI, work is still needed to bring about a change in mindset. “AI not only saves time, it also helps research results to be applied in practice, thereby creating real social value,” he said.