Politicians, business leaders and academics all say that the agriculture of the future will be digital. But even a thousand sensors, a hundred apps and ten platforms do not in themselves make for “Agriculture 4.0”. Farmers on the ground use whatever they actually need on their farms.
We need a sustainable increase in production, and to achieve this we need innovations, including digital applications and AI. For the DLG, this is the path to the future of arable farming – rather than regulatory measures or bans. However, digitalisation, and AI in particular, have not yet become as widespread in practice as would be desirable, given the need to combine productivity with environmental and climate protection. There is a gap between expectations and necessities on the one hand, and reality on the other, which can lead to disappointment. Here are three observations on this.
1. When it comes to automation, digitalisation often works well.
Field-specific management, on the other hand, still has room for improvement. Automation, precision and organisation are the areas where digitalisation has been most successful. For example, RTK-based steering systems have quickly become the norm. Section control and digital field records are also part of everyday life on many farms. However, when it comes to decision support such as site-specific management, there is still plenty of room for improvement. Until now, the cost-benefit ratio has often been out of balance. The growth in the use of NIRS sensors and, above all, drones shows, however, that there is a lot of momentum in this area.
It is not just about higher yields, but also about limiting the costs imposed by policy and/or high prices. Weed detection using cameras (mounted on machinery or on drones) in conjunction with ‘spot spraying’ follows this logic, for example. However, this is where costs come into play, which for many farmers are the main argument against digital solutions. One thing is quite clear: digitalisation must benefit the farm itself, not some “higher” goals. In the eyes of farmers, these are, at best, a “by-product”.
2. Digitalisation means networking. In practice, this is still too rarely utilised.
What Grandfather used to note down in his little red book is now stored on a computer or, increasingly, in the cloud. The difference is that significantly larger volumes of data are generated, and these can be compared and cross-referenced with reference data collected elsewhere. However, due to a lack of connectivity and incompatible interfaces, the exchange of data still does not run smoothly – at least across different manufacturers.
With a particular focus on crop protection, there could be more regional networking based on pest infestation and weather data from different farms. A small example of this is digital yellow traps. This would not only provide individual farmers with a convenient alert on their smartphone, without having to check for themselves. It would also enable a coordinated regional approach to flower spraying that takes the interests of beekeepers into account.
Artificial intelligence is already behind such apps. This raises the question: is arable farming more successful with digital decision-making models than with one’s own judgement? There is a fundamental difference: self-learning decision-making models follow a digital logic rather than real-world conditions. They prioritise data rather than the problem, which often needs to be solved within a broader context.
Artificial intelligence cannot assess situations that have not been ‘programmed’ in advance. This means that, at least initially, it comes into conflict with the self-image of the typical Western European arable farmer. By contrast, the models could be of enormous help to less experienced farmers or those whose focus lies outside arable farming. They are also likely to be helpful when making decisions under uncertainty – that is, when dealing with complex data and differing objectives or challenges.
3. Large companies in the agricultural machinery and agrochemical sectors, in particular, are pushing digitalisation with remarkable vigour.
Nevertheless, the “triangle” of manufacturer, service provider and farmer could be stronger. Farmers’ desire for independence and their mistrust regarding what happens to their data play a key role here. There is often talk of publicly accessible “data rooms”, but in reality, solutions tailored to individual companies prevail. In this context, companies are both drivers and those being driven. This is because not only the EU but also investment firms are demanding proof of sustainability from larger companies. The agrochemicals sector, in particular, with its outdated business model, does not represent the future. Terms such as “digitalisation” or “Agriculture 4.0” are taking on this role. Agricultural machinery, in turn, needs distinctive offerings that go beyond “steel and iron”.
Digitalisation requires a new way of thinking
Most farmers operate according to the principle: I want to keep improving what I can do, but not do things differently. Everyone is familiar with the factors that influence the business, from the weather to the price of grain, and knows how to deal with them effectively. However, when additional variables come into play – such as the whole range of environmental requirements and targets – it may no longer be enough simply to integrate new “tools” into traditional processes.
“The major challenge of digitalisation is not just to sort through the wealth of data, but to place it within an agricultural context. An IT developer simply cannot be familiar with that. Furthermore, psychology comes into play. Farmers want solutions that help their business move forward, not ones designed primarily to meet political and social requirements,” says Karl-Heinz Krudewig, a farmer and consultant in the Rhineland, Germany. “If I reject the Fertiliser Ordinance, I’m not going to take a positive approach to a digital tool either.” Another reason for their reluctance towards digitalisation could be a lack of trust. Along the lines of: “My farm is my farm; I’m not letting anyone from outside in!” A third reason is a lack of familiarity: digital technology isn’t part of day-to-day business. An SAP accountant who does nothing else all day can quickly work around any weaknesses in the programme. A farmer who is at his wits’ end with the various systems? Probably not.
Digital service providers must therefore succeed in gaining the trust that, for example, business management working groups currently enjoy. There must therefore be real people behind them too.
Conclusion
Farmers will make use of whatever benefits them. In a world of sustainable productivity, where inputs are expensive and in short supply, the benefits of digitalisation will increase. Its business model is based on the reuse of data. The dilemma here is this: anyone who seeks to ‘socialise’ data – for example, in government-run clouds – is cutting off the flow to private-sector developments. Those who privatise it through ‘data theft’ and do not allow farmers to share in the benefits will, at some point, find themselves with none left. Perhaps we need to move away from the notion that digitalisation fundamentally offers added value at no cost. Sensitive data, in any case, belongs only in the hands of trustworthy partners. Digitalisation therefore has not only a technical dimension, but also an economic and a psychological one. When dealing with farmers, providers tend to argue on a technical level. However, all three dimensions determine their success.