Artificial intelligence algorithms are now used, among other things, to analyse plant images, forecast yields and automatically control machinery. AI now supports agricultural production on many levels – from sowing to harvesting and even sales – and also helps to reduce costs and adapt to climate change. Therefore, although new technologies are currently used primarily by the owners of the largest farms in Poland, the scale of AI use is growing rapidly.
Neural modelling is becoming increasingly widespread in the agricultural industry, but we are not yet fully exploiting the potential of this technology. At the moment, the technology is in its infancy – we have companies that are introducing technology to mobile and touchscreen devices and tractors to support the daily work of farmers. However, we have not yet reached the stage of use that we see on a daily basis, according to Prof. Maciej Zaborowicz, PhD, head of the Department of Biosystems Engineering at the University of Life Sciences in Poznań, in an interview with the Newseria agency.
In Poland, according to the National Centre for Research and Development in its report ‘Agriculture 4.0. Identification of technological trends’, citing a study conducted by the University of Agriculture in Krakow in cooperation with Microsoft (Startup Poland, 2020), Polish farmers are becoming increasingly interested in Agriculture 4.0 technologies. This applies in particular to solutions that improve fertilisation efficiency (72%), modern equipment for mechanical weed control (54%), soil cultivation support systems (48%), automatic machine guidance (54%) and telemetry (48%).
However, there are many more advantages to using AI, and automation on farms brings measurable benefits. Thanks to machine learning algorithms, farmers can make more accurate decisions, optimise costs and care for the natural environment. Artificial intelligence allows for faster reduction in the use of fertilisers, automation of harvesting and detection of plant diseases.
A study published in the journal Computers and Electronics in Agriculture demonstrated the ability of an artificial intelligence system to detect apple tree diseases. Using a neural network trained on a dataset of apple leaf images, the system achieved an impressive 95 per cent accuracy in identifying the presence of disease. Another practical example is the detection of yellow rust in wheat crops. Researchers used machine learning algorithms to analyse images of wheat fields, effectively identifying infected areas with high accuracy. This application of artificial intelligence not only saves time but also minimises losses by enabling early intervention.
From the perspective of agricultural engineering and biosystems engineering, where we are talking about the application of artificial intelligence methods, this certainly includes field mapping, soil fertility testing and determining optimal traffic paths. We read satellite data, look at where the yield is lower, where it could be more evenly distributed, determine the appropriate fertiliser doses, and this is already happening today. When it comes to the application of artificial intelligence methods in food production as such, these are certainly algorithms that support food production processes, says an expert from the University of Life Sciences in Poznań.
Artificial intelligence can significantly help in assessing the quality of vegetables and fruit, both at the cultivation and harvesting stages, as well as during processing. For example, computer vision systems with RGB and near-infrared (NIR) cameras are used to sort apples or tomatoes – AI classifies fruit based on colour, shape and ripeness. Algorithms based on sensors or satellite images help to select the best time for harvesting, while near-infrared spectroscopy and hyperspectral imaging technologies allow the sugar, acid, water and starch content of fruit and vegetables to be assessed.
First, we need to obtain a set of relevant parameters, a set of relevant characteristics of the products we want to evaluate, enter them into such a system and create an appropriate model that will allow us to make an additional objective assessment, untainted by the subjectivity of the person making the assessment, and allow us to evaluate the quality of, for example, a tomato or meat,” explains Prof. Maciej Zaborowicz.
A study by Prescient and Strategic Intelligence shows that the artificial intelligence market in agriculture was worth just over $2 billion in 2024 (compared to $1.6 billion in 2023). By 2030, this figure could reach almost $8 billion. In Poland, according to the NCBiR report, Agriculture 4.0 technologies are most often used by owners of farms larger than 20 ha (38.4% of respondents), 7.6% of respondents on farms between 2 and 20 ha, and only 3.6% on even smaller farms (less than 2 ha).
Young people are much more willing to invest in new solutions and are eager to use them. Sometimes they download artificial intelligence applications for testing to see how they work. As the average age increases, so does resistance to these changes, but we already know this from the stage when we introduced computers to agriculture. Today, no tractor operator can imagine working without a joystick or a handheld pad, but we must slowly enter the stage of using new software, such as artificial neural networks, machine learning and information technology, explains a researcher from the University of Life Sciences in Poznan.
Artificial intelligence can significantly reduce costs in agriculture. Intelligent systems that recognise plants apply plant protection products only to weeds, which reduces the cost of chemicals.
Intelligent irrigation systems can save up to 30-50 per cent of water compared to traditional methods, while AI-driven machines operating on optimal routes reduce fuel consumption. Therefore, according to the expert, their popularity will grow.
New technologies primarily reduce expenses, i.e. they save money, because they are advisory systems. This advice consists in eliminating errors in planned processes, for example, optimising the process of travel, obtaining a given material, and this in itself is already an added value. The cost of acquiring technology has also fallen significantly; it is no longer as expensive as it used to be, so introducing such solutions to the market certainly helps even small farmers without large investments and costs, argues Prof. Maciej Zaborowicz, PhD.
