Spyros Fountas in Spotlight for STELAR: Precision Farming Adoption
In the latest episode of the STELAR Project’s Data Stories 360° podcast, Sandra Kolarić, Dissemination and Communication Manager at Foodscale Hub, hosts Professor Spyros Fountas from the Agricultural University of Athens and the CrackSense Project consortium.
With extensive experience in precision agriculture, our guest explores how technology adoption, farm management, and information systems are improving the management of high-value crops. His work spans multiple EU-funded projects, bringing practical insights from research to the field.
Let’s explore how digital tools, AI, and data-driven approaches are redefining the way we approach smart farming. From hands-on training for farmers to the growing need for high-quality agricultural data and precision farming adoption, Professor Fountas shares valuable reflections on the challenges and opportunities in modern agriculture.
Exploring Our Guest's Motivation and Background
Can you tell us a bit about your background and what first drew you to precision agriculture?
I have been working on precision agriculture for more than 20 years. It all started when I was a master’s student in the UK at Granfield University, where I was pursuing a master’s degree in management information systems. That was the initial phase when only a few farmers in the UK, the first 25 innovative farmers, had formed a kind of club to explore how precision agriculture could be applied.
As part of my master’s thesis, I visited all 25 farmers and their farms across the UK to observe how they were using technology, the problems they encountered with technology, and the challenges related to data adoption.
After that, I moved to Denmark to further my work in precision agriculture under the supervision of Professor Simon Blackmore. Simon Blackmore is considered the godfather of precision agriculture in Europe, alongside a few other key figures like Dick Godwin and John Stafford, who helped establish precision agriculture. I was very fortunate to work with them and apply precision agriculture alongside agricultural robotics.
Later, when I was in Greece, we focused on applying agriculture and digital technology to various cropping systems. Greece has a unique agricultural system with small-scale farms and a population that is not very computer literate. In this context, we explored adoption issues and behavioral changes, aiming to make farmers more willing to engage with precision agriculture.
What are some of the key milestones in your career that have shaped your research direction?
It is always nice to see how precision agriculture and digital technologies are progressing. Initially, when I started working on precision agriculture and precision farming, I followed the terms closely. I began in 1999, when I did my first thesis on precision agriculture, so it has been nearly 20 years now. Back then, we focused solely on precision agriculture, examining the variability within just one single field. All the work revolved around analysing one field at a time.
Over time, with the use of various software to cover more farms, things started to change. We began to look more at the farm as a whole and its different states. In the beginning, we were only using geostatistical analysis to analyse the data and identify trends.
One milestone in this progression occurred in 2016 when deep learning and AI models began to emerge. This allowed us to analyse more data from multiple fields and integrate more data within the same field to identify similarities and correlations, ultimately helping us predict yield and quality more accurately. AI played a significant role in this development, as it allowed us to explore, explain, and predict abiotic and biotic stresses more effectively.
Another important point is that in recent years, we have realised that it is not just about technology anymore. Even with the best technology available, we understand that technology often outpaces the development of agronomic models. The models have not kept up with the rapid advances in technology, so we now recognise how crucial it is to align the agronomic models with technological progress.
The second part of the equation is understanding the factors that hinder precision farming adoption and digital technologies. This is why my group has been exploring adoption trends, as well as conducting behavioural research for digital agriculture.
Precision Farming Adoption: Challenges and Opportunities
While research on precision farming is advancing at an increased pace, the end-user adoption is somewhat delayed and not widespread. What are the most significant challenges you have encountered in managing precision agriculture for high-value crops? On the other hand, based on your experience, what are the key obstacles farmers face when adopting precision agriculture technologies?
The challenges are quite numerous. We believe that in high-value crops, especially irrigated crops, the management zones are generally more stable. For example, we could easily divide the fields into management zones and apply different practices for each zone. However, this is not the case with non-irrigated crops, like rainfed cereals in northern Europe, where irrigation is not used. In such cases, it is very difficult to create management zones because the variability every year is so different.
Initially, we expected management zones to be more stable over the years, but we realised that even with irrigated crops, such as permanent trees or vines in vineyards, the variability persists. This means we cannot establish permanent management zones. The variability is influenced by factors such as climate change, which impacts the field every year.
Therefore, we cannot rely solely on historical data, although it can highlight recurring problems. We now need to find ways to manage fields effectively throughout the growing season. This is a critical issue, and we must develop tools – both technological and agronomical models – that can help us address this challenge.
Another challenge is that, while technology is advancing rapidly, agronomy is still lagging behind. For example, in the field of crop protection and pesticide use, we can easily collect a large amount of data on weather patterns and pest infestations. However, there are few models available for predicting diseases or pests, and many of the existing models are outdated and not suitable for different countries.
We need to develop agronomical models that cater to specific needs, particularly for crop protection and fertilisation, as these areas are still behind in terms of technological advancements.
Another challenge is that, despite the rapid development of technology, it is still not easy to use, particularly for mainstream farmers. We need to find ways to make technology more applicable and user-friendly for the average farmer, avoiding unnecessary complications.
Lastly, we still lack robust economic models that can convincingly demonstrate to farmers – both small and large scale – that specific technologies or operations are beneficial for them. Fortunately, the European Union has recognised this issue, and a number of projects are underway to address it. There is a growing trend in this direction, but it is essential that we focus on local or regional levels, rather than applying broad models at the national level.
Is there a way to overcome the barriers to adoption by improving farmers' education and training in using digital technologies?
Yes, I think most of the projects I have been working on included that aspect. One of my first EU projects was Future Farm, which was under FP7, many years ago – 2008. It was actually the first EU project on digital agriculture in Greece. We did work on that, and training and education was one of the main findings.
The second project I coordinated was Smart-AKIS – smart technologies within the AKIS network. Again, what we figured out, in all countries – not only in southern or eastern Europe – was that farmers were a little bit behind in education and training. We need to have good training. I do not know what “good training” exactly means, but it should be something that makes people aware – not just attending a workshop, but really using the technology.
We need hands-on training with the farmers, and we should go to them – visit their fields or neighbouring ones. And I would say, not at the national level. It has to be at regional or local level. They should come, see, and explore, because if you take a farmer outside of their environment to another region, they will not feel comfortable.
It has to be within their own environment – what they know. It should be among neighbouring farmers. That is the way we want to go in order to accelerate the precision farming adoption and use of technologies.
What are the main advantages of precision farming driven agriculture management compared to the traditional?
Precision agriculture has economic advantages. We may not save inputs per se, but we optimise the use of resources. That means we apply the resources where they are needed, rather than using them uniformly across the field. This leads to better organisation and better optimisation of resources.
Secondly, there are of course environmental benefits. We avoid overspraying and reduce spray drift, which means we do not damage neighbouring fields. We also avoid the overuse of fertilisers or spraying, which prevents leaching into the soil and the contamination of underground water sources.
Then, of course, there are social benefits. Precision agriculture makes farming more modern and appealing. We all know that the younger generation considers the mobile phone as an extension of their hands. I do not think there is any young person without a mobile phone constantly in their hands. They all want to use apps, technologies, and techniques. So, we make agriculture more attractive to the younger generation by integrating technology.
They can even control greenhouses remotely. They can monitor what is happening in their fields using cameras and drones. It actually makes agriculture more appealing. And while we do not do this just to make it attractive, it does carry value. We are able to measure and explain the variability and the differences within the field.
Data Accessibility and AI: What’s Next for Precision Farming?
The agri-food sector is filled with vast amounts of data, but its true value is often underutilised due to difficulties in accessing and using data spread across various formats and locations. Do you think that the agrifood sector would benefit from a platform for intelligent data discovery, AI-ready data, and semantic interoperability?
Data is actually the future, and we need a lot of data to train models. We need a lot of data to explain the causes of problems, stresses, and various issues. We have to train models, and as we all know, it is very hard to collect raw data in the field at high quality. This is really difficult.
It is very easy to say, “I want to run models,” but you need good raw data. We need a lot of raw data, and we need well-annotated data that can be used by machine learning models to explain what is going on. Based on that, we can then move toward using generative AI to create new models and new datasets, and to explain causes that we are not yet aware of.
So, we definitely need such datasets and platforms to be openly available, to help the community run models and algorithms – for the good of everybody, actually.
Looking ahead, what do you see as the next big frontier for precision agriculture? How do you envision AI and machine learning advancing in this field?
I think we should see something like customised, individualised decisions – per field, per farmer. AI would actually help to personalise the decision-making process and the outcomes, tailoring them to each individual. Because everybody has their own beliefs, their own problems and issues.
With the advancement of AI, and also Large Language Models, I believe it will become much easier and more user-friendly. This will help significantly in making precision agriculture more applicable to every individual farmer and agronomist in the future.
Conclusion
In this conversation, Professor Fountas shows us why precision agriculture is more than just a buzzword – how it optimises resources, reduces environmental impact, and makes the sector more appealing to younger generations. Whether you are a researcher, policymaker, or agrifood innovator, this episode offers fresh perspectives on bridging the gap between technological innovation and on-the-ground practice.
Subscribe to our YouTube channel and explore our Blog for more insights on data innovation, AI in action, and the evolving landscape of data sharing!
