*This is the first in a series of blog posts published as part of the author’s ongoing research.

Cevdet Bulut | PhD Candidate | School of Business Management, Royal Holloway, University of London

In a hilly village of Banoštor, only about a half-hour drive away from Novi Sad, I stand on the top of the hill surrounded by a 70-hectare vineyard, overlooking the waters of the Danube River.

“Soon, in about 15 days, the grapevine will bloom. It is a very smart plant. It usually does not come to maturity because, for example, last week we had very cold weather,” says Nemanja Mišić, an agronomist and research assistant at DunavNET. “I work with end users in the field for hardware installation, but it’s my own wish to be there, because it’s easier for me to understand the whole process,” he says, highlighting the importance of being close to the end users.

On my second day of research trip with the company, I am offered to join Nemanja for a field trip to Fruškogorski Vinogradi, a vineyard situated on the slopes of the Fruška Gora Mountain, the only mountain in Vojvodina region. I was eager to observe the interaction with farmers and talk to them, as well as to see an IoT-enabled weather station in action.

The agroNet web interface is showing the stations visited during the field trip.

Fruškogorski Vinogradi is one of vineyards included in the pilot project funded by the Association of Producers of Grapevine and Wines in Fruška Gora. A total of 17+ weather stations have been installed by DunavNET to cover multiple vineyards, so that every member of the association would have access to agroNET for the shared data and insights.

We decide to leave the car and walk uphill on this beautiful April day. After a ten-minute walk, we arrive at one of two weather stations installed to cover the vineyard. This expensive piece of hardware, costing about 1000 euros, is used by the company to capture temperature, humidity, and precipitation in order to calculate the water deficit and air respiration. “This weather station has been here since the beginning of the project, which is now three and a half years, and has worked fine until today,” Nemanja says, reaffirming its stability and resistance to adverse weather conditions. When it comes to energy consumption, it doesn’t seem to be an issue thanks to the use of solar panels, according to him. “They usually say five to ten years, but it depends on the size of the solar panel attached to the weather station. The bigger panels can last longer because they charge better.”

Nemanja inspects the weather station

agroNET detects a malfunction with this station, which sends abnormal values for temperature measurements. After some inspection and discussion over the phone with colleagues at the office, Nemanja comes to the conclusion that the capacitor is the source of the issue. “It is not the first time that we have had this problem, and it is not the first time that we have been able to solve it with a capacitor. It is not a big job, but it does take time; I mean, you need to come here.” As an active observer, I assisted Nemanja in dismantling it and taking it to the office to be replaced and activated.

Soon after, we are joined by Ivan, the vineyard manager. “This is an old wine region. In this village [Banoštor] every house makes wine,” says Ivan, who lives in a nearby village and has been managing this vineyard for 13 years. At its peak during harvesting season, Ivan manages a team of over fifty employees in addition to other seasonal workers, spending most of his time managing and organising the daily operations.

Ivan uses agroNET to monitor and read reports on daily basis using his mobile. The app he finds “Very easy, with no problem,” and he likes to use it mostly to read information about the expected “amount of rain” for his decision making on when to spray. “In agriculture, it’s very important when you spray your pesticides. Because it’s important that rain doesn’t fall after washing the pesticides away from the leaves, and sometimes it’s very important to know how much rain actually has fallen,” Nemanja elaborates. However, spraying must be managed manually. Ivan touches upon the lack of automation for spraying but does not yet see the use of drones for spraying pesticides as a viable option. “Not enough amount of water drone can handle, because one hectare needs 600 to 1000 litres and the drone cannot carry that much water,” Ivan explains.

Interviewing Ivan, the vineyard manager

Ivan nods and expresses his appreciation of the service he receives but points out one issue due to the terrain conditions of the vineyard. “More complicated weather station necessary for these vineyards. Here, we may need more stations to cover all of this. There is a big difference. This vineyard is different from the down vineyard; the microclimate is different,” he says.

We skip the second and go directly to the third station which is installed to monitor insects. This device, consisting of an insect trap and a camera, sends a picture every day. Ivan checks the agroNET app every morning to track the number of insects and decides when to apply pesticides. He says that it is helping, but only for one type of insect. In vineyards, they have two or three different types, depending on the year.

“Agriculture is really complex, and every year brings something new and different. With globalisation, there are new insects, new diseases, and other big problems…”

A total of three stations for weather and insect monitoring have been provided to Ivan to support him in making data-driven decisions. “This is just a tool to help farmers make decisions on time according to the data. This cannot solve all their problems, but it can help them to better manage,” Nemanja explains. Ivan certainly sees value in the product, but he underlines the associated cost as a barrier, saying, “In Serbia, the biggest problem is paying.”

“The financial part is a barrier, definitely,” says Spasenija Gajinov, a product manager at dunavNET. She expects the government to act and provide subsidies to farmers and thinks that would accelerate the adoption of IoT. But it’s not yet in their agenda, Nemanja says, “For tractors and machinery. Last year, I think it was solar panels or something similar. So, it is likely that it will come to the digital technologist.”

The agroNET web interface is showing an image of insects captured by the station

DunavNET, the case company, was started in 2005 as a software consultancy by Srdjan Krčo, the founder and CEO of the company. After spending several years working in Ireland, he decided to return to Serbia in 2012 to fully engage with his company and to start an IoT business.

Novi Sad, where the company is based today, is the biggest city in the Vojvodina region – the agricultural heart of Serbia. “This whole area here, if you notice from coming from Belgrade airport, very flat apart from that one mountain [Fruška Gora Mountain] that you kind of cross. The primary domain sector was, and today one top three sectors, is agriculture,“ says Srdjan, the grandchild of farmers who himself spent quite a few years in the field.

DunavNET provides IoT-based solutions for many business domains, including agriculture. Today, they offer a number of IoT solutions for many sub-sectors, including livestock, orchard, vegetables, and arable crops, for various use cases such as weather monitoring, irrigation, and pest detection, using one single platform. “From the very beginning, we tried to build a holistic platform, something that would allow farmers to collect all their data into one place, whether it’s a tractor or weather station, and then build modules on top of that, that would give them recommendations, insights on specific operations, so that they can manage everything in one place. That was the approach,” Srdjan says about the company’s product strategy.

Srdjan Krčo, Founder and CEO

Srdjan believes that the IoT and ML/AI have the capability today to generate at least the same, if not better, insights, wisdom, and knowledge in agriculture that his grandparents had.

“Eventually, that is something that’s already done. So, whether it is just IoT, or machine learning, there is now a blurred line between these technologies, but we have in our system today.. If you can record temperature, humidity, precipitation, when the vegetation started etc. then you can predict diseases. And, with the support of these digital technologies, you can do it in a more efficient automatic manner, supporting you in-house or external agronomists and making their decision taking process more efficient.”

IoT is core to the company’s products but complementary to the services provided by the company. Consultancy makes a major part of their daily business. Domain knowledge is the key to success, according to Srdjan. “The good thing about IoT is that it can be applied or is being applied in different sectors. The difficult part is the same, because if you want to create an IoT solution for agriculture, .. you have to know that domain. Having an IoT platform is not sufficient, you have to provide solutions that the end users can avail of straight out of the box.”

Partners play a crucial role in service delivery. The company maintains a network of partners, ranging from hardware providers, system integrators, and telecom operators to local consultants, who can provide “a connection between us and the farmers,” Srdjan says. “We offer complete turnkey solutions, and we wouldn’t be able to do that without partners. It’s much easier if you work with someone for a longer period of time,” he explains. “We created our ecosystem,” Spasenija adds. “agroNET has been created in collaboration with researchers from universities, agronomists, and other agricultural consultants who have been in the business for twenty years,” she says. “People who are part of agricultural production with their knowledge and expertise. We use and digitise their knowledge on our platform.”

What about the technology? “It’s maturing,” Srdjan says. “For many of the scenarios, the technology as it is today is already there and solves the needs of the farmers (e.g., weather stations, various sensors, etc.) Of course, as the technology evolves, the requirements evolve as well – it’s a kind of push-pull process. Soon, we will see much more ML/AI-based solutions in the field, robots, etc.” Petar Knežević, who works as a software architect at DunavNET, confirms the same, adding, “Connectivity, battery and sensors are self-sufficient. In the last ten years I have witnessed the evolution of these devices and batteries.”

Spasenija is presenting the latest sensor technology

Today, the company exports its products and expertise to a number of countries in Europe, Canada, and the Middle East. “We were always among the first ones to do something in the digital agriculture sector in the region,” Srdjan says proudly. Nevertheless, he says the business of agriculture is “yet to become fully profitable.” “It takes time to earn trust of farmers and for the cost of technology to drop to acceptable levels. Anyhow, since COVID, we are seeing an increased pace of traction, and the engagement with farmers more focused on deployment than explanation of what benefits such solution can bring.”

Agriculture is a tough sector which “needs a lot of time and patience” in comparison to other domains such as smart cities, according to Spasenija. The sales cycle is typically quite long, but they think it will shorten in the near future. The company is already seeing some real interest. “It’s only last year and now this year, and there is that we see an increased number of commercial customers,” Srdjan says.

The team is convinced that the trend is upward, and the future is bright for smart agriculture. “I suppose, from year to year, we will have improved solutions, better devices, and more clients,” Spasenija concludes.

About

This is a series of blogs curated by PhD candidate Cevdet Bulut, who is investigating the adoption of IoT in agriculture. In this series, the author shares a limited version of his field notes and highlights from multiple case studies in a news story format as part of his current research study. The scope of the case studies is to uncover what factors (socio-economic, cultural, technical, etc.) affect the adoption of IoT and to gain experience from the field that can guide the design of viable IoT-based business models for the sector. A new blog will be published on each case study with the participant’s permission.

Other blogs in this series:

• Part II: IoT In Agriculture: Trust Matters