Networking and Transfer Project for Digitalization in Agriculture

Background and Objective

The Federal Government has initiated a large number of research activities, which will make valuable contributions to the development of digitalization in agriculture, to use the opportunities of this sector and to identify possible regulatory requirements.

The aim of this project is to compile and evaluate the huge number of contributions from science and research according to a defined method, to finally network the respective researchers, to give advise to the politicians and to inform the public.

Approach

1. Scientific conception of the procedure and establishment of a steering committee including institutions, organizations and associations
2. Evaluation of technological developments of the research projects of the Federal Ministry of Food and Agriculture (BMEL) and the industry
3. Analysis of the consequences the technology has to use opportunities and minimize risks in the topic area of digitalization in agriculture
4. Suggestions for support initiatives in digitalization and further developments of the political and juridical frameworks
5. Networking and transfer of knowledge to science, society and agricultural practice

Results

The cost-benefit analyses carried out for selected applications of new technologies in weed control, fertiliser management and disease and pest control show that:

1. Spot spraying approaches offer potential savings in foliar herbicide application of the order of 50-70%, in some cases up to 90% of the application rate. However, they have so far only been successful in crops with high weed control costs, such as sugar beet or certain specialty crops. The closer the distance between the nozzles to be shut off, the greater the potential savings.
2. Determining nitrogen requirements using multispectral cameras on drones and satellites is particularly suitable for very heterogeneous sites. On more homogeneous and high-yielding sites and on large farms, tractor-mounted multispectral camera systems often offer a better cost-benefit ratio. Further research is needed to validate the systems in multi-site, multi-year field trials.
3. The approach of predictive models is to estimate the probability of pest occurrence and to omit full treatment measures when the probability of occurrence is low. However, the accuracy of these predictions is currently only in the order of 85%.