ASABE Technical Library - Abstract

Click on “Download PDF” for the PDF version or on the title for the HTML version.
If you are not an ASABE member or if your employer has not arranged for access to the full-text, view access and subscription options

PeroHop4.0 – An AI-Based Multi-Stage Spore Detection and Forecasting Framework for Downy Mildew Management in Hop Production

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org

Citation:  2026 ASABE Annual International Meeting  2600062.(doi:10.13031/aim.202600062)
Authors:   Christina Maria Sebald, Lina Liv Willruth, Johann Portner, Johann Hausladen, Florian Haselbeck, Heinz Bernhardt
Keywords:   hop production, downy mildew, spore traps, artificial intelligence, decision support, integrated pest management

Abstract. With the greatest potential for yield losses in hop cultivation in both worldwide leading hop production regions, Germany and the U.S., downy mildew (pathogen Pseudoperonospora humuli) is one of the most significant plant diseases in this field. In the 1980s, the “Hop Research Center Hüll” successfully reduced the chemical plant protection for downy mildew in hops by introducing the “Peronospora warning service” (Peronospora warning service 1.0) for all hop growers in Bavaria.

Since then, the warning service has been operated by the Bavarian government for 40 years without major revisions, serving as an essential pillar of integrated plant protection in hop cultivation. The method of detecting downy mildew in the existing warning service is very labor-intensive and time-consuming, primarily involving manual work. To address upcoming challenges, such as labor shortages in the detection process and regional inaccuracies in the existing warning service, the method needs to be revised.

Now, through the integration of modern AI-supported technologies, the project “PeroHop4.0” is significantly advancing the existing “Peronospora warning service”. The first step involves AI-based automation of spore analysis for Pseudoperonospora humuli based on digital microscopic images, which ensures reliable, reproducible, and less labor-intensive detection of spore presence/occurrence/levels (Peronospora Warning Service 2.0). In a second step, an automated real-time spore trap will be adapted for the application to achieve a scalable solution with low personnel requirements (Peronospora Warning Service 3.0). Finally, based on extensive historical data, an AI-based, area-specific early warning system is being developed that provides forecasts for individual regions on the infestation situation for the next two to three days (Peronospora Warning Service 4.0).

These innovations lead to a resource-saving plant protection strategy that significantly reduces the use of chemical pesticides, working hours for farmers, and improves the detection process in a more time and labor-efficient way. Practical development, integration into the existing warning service, and free provision ensure a high level of acceptance among hop growers. In the long term, the project will make a decisive contribution to the sustainability and future of hop cultivation, and the methods developed can be transferred to other pathogens and crops.

(Download PDF)    (Export to EndNotes)