Identify priorities to leverage smart digital technologies for sustainable crop production

Drones monitoring fields for weeds and robots tackling and treating crop diseases may sound like science fiction, but in reality they are already happening, at least on some experimental farms. Researchers from the PhenoRob Cluster of Excellence at the University of Bonn are working to advance the smart digitalization of agriculture and have now published a list of research questions to be addressed as a priority in the future. Their article appears in the European Journal of Agronomy.

The fact that the Earth today feeds more than 8 billion people is in no small part due to modern, high-quality agriculture. However, this success comes at a high cost. Current cultivation methods threaten biodiversity, while the production of synthetic fertilizers generates greenhouse gases and agrochemicals pollute water bodies and the environment.

Many of these problems can be solved by using more targeted methods, such as applying herbicides only to those areas in a field where weeds actually become a problem, rather than treating the entire area. Other options are to treat diseased crops individually and only fertilize where really necessary. Yet strategies like these are extremely complex and virtually impossible to manage at scale conventionally.

Utilizing high-tech and AI to become more sustainable and efficient

“One answer could be to use smart digital technologies,” explains Hugo Storm, member of the PhenoRob Cluster of Excellence. The University of Bonn is working together with Forschungszentrum Jülich, the Fraunhofer Institute for Algorithms and Scientific Computing in Sankt Augustin, the Leibniz Center for Agricultural Landscape Research in Müncheberg and the Institute for Sugar Beet Research in Göttingen on a large-scale project aimed at making agriculture more efficient and environmentally friendly with using new technologies and artificial intelligence (AI).

The researchers come from a variety of fields, including ecology, plant sciences, soil sciences, computer science, robotics, geodesy and agricultural economics. In their recently published position paper, they set out the steps that they believe should be taken as a priority in the short term.

“We have identified a few key research questions,” says Storm. One of these involves monitoring agricultural land to detect any nutrient deficiencies, weed growth or pests in real time. Satellite images provide a global overview, while drones or robots enable much more detailed monitoring. The latter can systematically cover an entire field and even record the condition of individual plants.

“One difficulty lies in linking all these pieces of information together,” says Storm’s colleague Sabine Seidel, who coordinated the publication with him: “For example, when is low resolution sufficient? When should things get more detailed? Should drones fly to achieve maximum efficiency in viewing all crops, especially those at risk?”

The data obtained provides an impression of the current situation. However, farmers are mainly interested in weighing different possible strategies and their possible implications: how many weeds can my crop tolerate, and when should I intervene? Where should I fertilize and how much should I put down? What would happen if I used fewer pesticides?

“To answer these kinds of questions, you have to make digital copies of your farmland, so to speak,” Seidel explains. “There are several ways to do this. Something that researchers have yet to discover is how the different approaches can be combined to get more accurate models.” Appropriate methods should also be developed to formulate recommendations for action based on these models. Techniques borrowed from machine learning and AI can play an important role in both areas.

Farmers need to be on board

If crop production is to truly embrace this digital revolution, the people who will actually put it into practice – the farmers – will also have to be convinced of its benefits. “In the future, we will have to focus more on the question of which underlying conditions are necessary to secure this acceptance,” says Professor Heiner Kuhlmann, a geodesist and one of the two speakers of the Cluster of Excellence alongside the head of robotics -group. Cyrill Stachniss.

“For example, you can offer financial incentives or impose legal restrictions on the use of fertilizer.” The effectiveness of these types of tools, individually or in combination, can now also be measured using computer models.

In their paper, PhenoRob researchers also use examples to demonstrate what current technologies are already capable of. For example, a ‘digital twin’ of built-up areas can be created and sensors fed with a steady stream of different types of data, for example to detect root growth or the release of gaseous nitrogen compounds from the soil.

“In the medium term, this will make it possible to adjust the level of nitrogen fertilizer applied in real time to the needs of the crops, depending on how nutrient-rich a particular spot is,” adds Professor Stachniss. In some places, the digital revolution in agriculture is already closer than you might think.