On the basis of environmental and crop growth data and profound expert knowledge, the “Fraunhofer Research Center for Machine Learning” and the cluster of excellence PhenoRob have developed an artificial intelligence that determines environmental conditions in agriculture and their impact on plant growth. With their technology, the scientists took the first place in the Syngenta Crop Challenge 2019.
Better seeds, less fertilizer, high adaptability – intelligent data analysis can support agriculture sustainably and profitably. In search of the best technologies the global agricultural company Syngenta launches the “Crop Challenge in Analytics” once a year. At the final in Austin, Texas, the technology based on “Informed Machine Learning” convinced the jury. First place went to scientists Dr. Bogdan Georgiev, Kostadin Cvejoski, Cesar Ojeda and Jannis Schücker from the Fraunhofer Machine Learning Research Center, led by PhenoRob’s project partner Fraunhofer Institute for Intelligent Analysis and Information Systems IAIS, and agricultural scientist Prof. Anne-Katrin Mahlein, principal investigator in the cluster of excellence PhenoRob. The center is part of the Fraunhofer cluster of excellence Cognitive Internet Technologies (CCIT) headed by the cluster’s principal investigators Prof. Stefan Wrobel and Prof. Christian Bauckhage.
Under the title “Combining expert knowledge and neural networks to model environmental stresses in agriculture”, the winning team investigated the relationship between environmental conditions, such as drought and heat, and plant growth. In addition to environmental and growth data, the Fraunhofer scientists also incorporated Prof. Mahlein’s expert knowledge into the development of the technology.