AI in agriculture: What role can the EU play?

by Nathalie Sauze-Vandevyver, Director for Quality, Research & Innovation, Outreach at the European Commission's Directorate-General for Agriculture and Rural Development 

Artificial Intelligence (AI) in agriculture? Really? If you walk in the countryside, between the fields with cows or wheat, one could wonder what on earth one has to do with the other.

Indeed, most of the times not too much or nothing at all, but things are changing. We have the satellites looking over our fields, 3 times per day the milk robots are analysing every little parameter of what the cow is producing and we have thousands of sensors in, on or next to the combines. Data is becoming abundant and when you have data you best do something with it.

Both the private and public sector have realised this, digitisation of agriculture is hot and everybody is talking about it. Platforms are being created, people are working on standardisation and harmonisation of data flows and new applications pop-up every day. Following this proliferation and integration of agricultural data, one of the main obstacles for the development of AI in agriculture has been eliminated. The use of AI in agriculture could therefore be the logical next step.

In other words, Agriculture 4.0, where every cow and plant gets its “personal” treatment, is coming closer. We already have a number of applications based on AI in agriculture, but there is much more to expect. Knowing this, we could question ourselves, what is the role of the public sector in this story? If it is unavoidable do we need to invest in it? If so, what could be the role of the EU in this story?

First of all, it is obvious that when a certain area of innovation carries the potential of serving the common good (e.g. increase of environmental performance), we have an interest in investing in it. AI and digitisation in general could help us to optimise processes and reduce the environmental impact of farming without jeopardising production, producing “more with less”. There is a clear role for the public sector to steer the development of digital applications in this direction. With Horizon 2020 R&I projects and rural development funds we could therefore invest in applications which have a direct positive impact on environment; Together with the MS we could invest in applications which would improve the functioning and performance of government (in agriculture also very much focussed on the environment); etc.

Also economically it could make sense. A second reason why the EU, as supra-national player, is in a unique position, is that it has the potential to create cross-border cooperation. Because of cultural, economic and institutional boundaries at national level, Europe has a more fragmented market than for example the US. In the digital sector, where economies of scale are quite important, the companies have a smaller basis to start with. This puts the EU’s companies at a disadvantage compared to their competitors. With initiatives, such as “Building a European Data Economy”[1], we can overcome these barriers and create a level playing field. Also for certain types of infrastructure, such cross-border cooperation is essential. For example, the developing of Copernicus, would not make sense for each MS separately.

Thirdly, because of the same economies of scale, the digital sector is often confronted with the principle of “the winner takes it all”. This means that especially in strategic sectors such as agriculture, investments in an open environment for the digitisation of the EU’s economy are necessary. Competition, with EU - and non-EU companies, as opposed to a monopoly, is necessary to drive innovation and serve the interest of all. If a farmer cannot switch from one service provider to another, there is no incentive for the provider to provide the best quality service. Such dominant positions can negatively influence the speed of innovation and in the worst cases lead to abuse of power. That is why, especially in the agri-food sector, we need to democratise the digital environment, pushing for standardisation, open platforms and ecosystem building.

In short, the development of AI in agriculture is going to take place. What we should do, is steer it in the right direction in a responsible way that serves the general interest well.