Original article (in Croatian) was published on 27/11/2023; Author: Melita Vrsaljko
Technological progress is helping make agriculture more resilient to climate change. However, these advancements also present significant ethical questions.
The growing global demand for food, coupled with the need to reduce greenhouse gas emissions from food production, presents considerable challenges. Furthermore, adapting to the increasingly severe weather conditions brought on by climate change adds to the complexity facing the agri-food industry.
Droughts, heatwaves, floods, and storms are increasingly impacting crop yields and the reliability of our food supply. The United Nations Food and Agriculture Organization estimates that up to 40% of global crop production is lost to pests. Additionally, about 14% of produced food is lost in the journey from harvest to sale, and an estimated 17% of total global food production is wasted in households, service sectors, and retail outlets.
Farmers are actively seeking ways to overcome challenges, especially those related to crop production and preservation, by embracing digital technologies, including software powered by artificial intelligence (AI).
Globally, innovative projects are being implemented on farms where robots help determine the precise amount of water or fertilizer needed for plants. Additionally, drones are being used to survey fields and analyze soil conditions. In the United States (US), there’s a significant trend of technological adoption, with over 200 startups developing AI-based software to enhance farming efficiency, according to Forbes.
In Croatia, similar advancements are being explored. At the Faculty of Electrical Engineering and Computing, a notable project called AgroSPARC is underway. This experiment involves growing wheat under varying climate conditions to amass extensive data on its cultivation. This information, which will eventually be accessible to farmers, aims to explore how predictive agriculture can bolster resilience against climate change impacts.
However, when it comes to practice, the use of software based on this technology in Croatia can currently only be found on larger farms.
When robots milk cows
These advancements are being utilized by companies within the Fortenova group. Faktograf reached out to them for insights on how they stay abreast of technological progress. They shared with us that Belje, one of their companies, is currently collaborating with various domestic and international academic institutions on projects that integrate artificial intelligence into animal husbandry.
These projects are at different stages of development. Some are in their initial phases, while others are already demonstrating early results and potential future applications in farming. The primary focus of these projects is on gathering and analyzing data related to farm operations and animals.
At one of Belje’s farms, a member of the Fortenova group, a remarkable technological implementation is evident: six robots are in charge of automatically milking cows.
Fortenova explained to us how this automation has changed traditional practices. “On this farm, cows are milked three times a day, diverging from the usual morning and evening schedule of conventional methods. Considering the numerous advantages of the automatic milking system, the plan is to further improve the system and integrate robotic technology with other systems and sensor technology, all with the aim of maintaining a good health condition and well-being of the cows.”
Agrolaguna, another company under the Fortenova umbrella, is also investing in production efficiency. They have recently completed a significant irrigation project, which is their largest investment in the last decade. This project, located in the Istrian town of Cervar-Porat, spans 106 hectares of olive groves, covering about half of the company’s total olive grove area.
The Fortenova Group says that this is how conditions were created for the stability of olive production in a time of increasingly unstable weather conditions and influences.
“A total of over 330 km of lateral network pipes and 20 km of distribution and main pipelines were installed. The system is controlled by 10 main units that are interconnected for automated operation via radionet signals without a built-in cable network. One of the most technologically advanced solutions for the underground installation of pipelines, drop by drop, was chosen in order to completely avoid pipes on the surface of the plantations and enable the maintenance of a grassy substrate and mechanized harvesting”, says Fortenova.
Furthermore, the project incorporates a fertigation system. This system allows crops to receive fertilizer through the irrigation process, ensuring each plant gets an optimal dose of water-soluble nutrients. According to Fortenova, this approach not only maximizes fertilizer use but also contributes to significant savings. The know-how and equipment for this project were sourced from Netafim, an Israeli company renowned for its irrigation solutions.
The technology is too expensive for small farmers
Such advanced technology is still unattainable for a large number of Croatian farmers.
As Mladen Jakopovic, president of the Croatian Chamber of Agriculture (HPK), told Faktograf, the technology required for the application of AI is expensive, which is why he is hoping that funding for this purpose will be available from the European Union (EU).
“Unfortunately, Croatia has an unfavorable structure of farmers because farmers of the older generation still dominate, but there are more and more young farmers who are ready for changes and the introduction of new technologies. What is crucial is that farmers need to be more educated and follow tenders from EU funds, which in the future will abundantly finance the application of new technologies in production, especially with the aim of preserving the climate and the environment in order to work on reducing the use of pesticides and the like”, he tells us.
Faktograf asked Jakopovic how often farmers in Croatia have the opportunity to learn about new trends and what specifically the Chamber is doing to bring technology knowledge closer to its members.
“HPK, through its Committees, informs farmers about all the news regarding new knowledge and technologies in the EU, and we often organize visits to farms in advanced agricultural countries or organize workshops so that our farmers can learn how new technologies are used today. In addition, two years ago we offered our members free use of the Agrodox application, which was designed by the company of Jan Marinac, the president of the Croatian Association of Young Farmers”, said Jakopovic.
Weak digitization is a problem for Croatian farmers
Matija Zulj, the director of Agrivi, a company known for its farm management platform used globally, shared his insights based on a decade of experience in the sector. He pointed out that small and medium-sized farms, which constitute a significant portion of global agriculture, often lack the knowledge and resources to progress independently.
Since every use of advanced technologies is preceded by basic digitization, we asked him how Croatia stands on that path.
“The state of agricultural digitization in Croatia can be understood by looking at the European Union average, which is between 15 and 20 percent. In Croatia, however, agricultural digitalization is only at about 5 percent. To put this in context, in the USA, the digitization rate in agriculture is around 30 percent,” Zulj told Faktograf.
Matija Zulj emphasized the gap between Croatia’s agricultural potential and its actual food production, noting that the country meets only about 50% of its population’s food needs through domestic production. This shortfall leads to a heavy reliance on food imports. Zulj highlighted the resultant challenges in tracing product origins through all stages of production, processing, and distribution, as well as ensuring food safety. Both factors, he notes, are increasingly important in consumers’ decisions when selecting and purchasing food.
We asked Zulj whether the infrastructural prerequisites for the use of artificial intelligence are met in Croatia since a large part of rural areas is not yet covered by the 5G network.
“5G technology supports the concept of smart agriculture, increases the possibilities of installing additional sensors, even more precise control and warning systems, but also the automation of a large part of the work in agricultural production. Croatia could certainly profit in the event that it promotes the introduction of 5G as quickly as possible”, says Zulj, but adds that “the solution is not only in technology, but also in supporting farmers in adopting new crop management concepts, implementing best agricultural practices focused on sustainability, safety and nutritional value, and timely crop protection to reduce the risk of yield loss and ensure profitability”.
What is AI used for in agriculture?
Artificial intelligence tools in agriculture can be used to monitor crops, and algorithms can also help determine the best time to sow seeds.
A notable example of this application occurred in Andhra Pradesh, India. In 2017, Microsoft launched a pilot project there, where farmers received SMS notifications with guidance on preparing seeds for planting, including the appropriate depth for planting them. Microsoft later reported that farmers participating in the project experienced an increase in crop yields ranging from 10 to 30 percent.
Software and applications can help detect pests and diseases and advise farmers on how to prevent their spread. German entrepreneurs have developed the Plantix application, which works to identify plant pests and diseases, as well as potential nutrient deficiencies in the soil.
Technology can also help monitor animal behavior; a company called CattleEye offered a solution in the form of drones with cameras that detect the atypical behavior of cattle and thus monitor livestock health remotely.
AI algorithms can help predict local weather forecasts and combine measurements from a local weather station with sensors in the field to help producers plan pesticide applications, harvests, etc.
The report of the Committee on the Future of Science and Technology of the European Parliament on this topic states that AI systems, by analyzing photos taken by a drone, can quickly identify which parts of farms need water or fertilizers, which enables the targeted use of resources, i.e. avoids applying water and fertilizers to those areas that don’t need it. AI systems perform such tasks much faster than a human, who would need several days to walk the field and identify areas where crops are not sufficiently hydrated or fertilized. In addition, by analyzing the data, the algorithms can identify situations that indicate a potential water leak from the pipes.
AI-based predictive analytics software can also generate predictions about yields and product quality, which can help farmers project their income as well as make decisions about how much to sell and how much to save for personal consumption. AI systems can also be used to analyze consumption patterns to predict demand for a particular agricultural product.
Artificial intelligence is also capable of harvesting certain plant varieties. For example, few years ago the American company Harvest CROO Robotics presented a robot that picks strawberries.
Although a large part of the innovations that come from technology start-ups are still not in mass use, their use is growing every year.
As the FoodUnfolded portal reports, the use of robots that perform automated actions in the production process is increasing worldwide by almost 20 percent compared to the previous year. This is why it is estimated that the AI technology market in agriculture will grow from the current 1.7 billion dollars to 4.7 billion dollars by 2028 (Markets&Markets).
Ethical dilemmas surrounding the use of AI tools in agriculture
Numerous researchers have raised concerns about the ethical implications of using artificial intelligence in agriculture.
A study published in AI&Society last year highlighted several potential issues. One concern is that robots, sensors, and drones used on farms might emit harmful chemicals and vapors. Additionally, the manufacturing process of these machines itself has an ecological impact.
Another significant worry is the increased vulnerability to cyber threats. Farms that heavily depend on automation and robotics could be more susceptible to hacking and sabotage.
There’s also apprehension regarding the data collected by AI tools. Questions arise about the nature of the data being gathered and its subsequent use. Often, the data collected are automatically sent to the software owner’s headquarters, leaving farmers unaware of the specifics of the information being transmitted from their farms. This lack of transparency in data handling poses a concern for data privacy and control.
The study published in AI&Society also discusses how large agricultural companies are gaining more control over farms through the use of technology. The authors note instances where these companies, through their contracts, restrict farmers from modifying or tampering with the hardware used on their farms. This limitation results in farmers having less control over what happens in their own fields.
Furthermore, there are concerns about the potential for artificial intelligence to create a “digital divide” between different farms and even between countries.
“The use of artificial intelligence only on rich, large, monoculture farms severely limits the adoption of artificial intelligence by smaller farms, poorer farmers and the global south. Another concern is that artificial intelligence will bring economic benefits to agribusiness and technology companies, but not to farmers themselves”, the authors state.
The potential risks associated with the use of artificial intelligence in agriculture have been acknowledged by the UN Food and Agriculture Organization (FAO). Qu Dongyu, the director of FAO, has emphasized that while AI can significantly impact agriculture, it’s crucial to ensure that its implementation does not lead to unintended negative consequences.
In early 2020, FAO, along with major companies like Microsoft and IBM, was among the first to endorse the Rome Call for the Ethics of Artificial Intelligence. This initiative, spearheaded by the Pontifical Academy for Life, aims to foster a sense of responsibility in the application of advanced digital technologies (1, 2).