top of page

Smart Arable Farming

Combine and weather station on soya field during harvest.

Typical crops in arable farming, like wheat, corn or soybean significantly contribute to securing food supply for our growing global population. This has been true since a long time and will remain so. But what changes is: arable farming is becoming smart.

But what does that actually mean, agriculture is becoming smart? Well, in a nutshell it means switching from human decisions to sensor driven information networks and the automated execution of operations in the field. This is driven by 1) technological innovation and 2) knowledge accumulation and processing. Both these drivers have been relevant since the very beginning of agriculture in the human history. When it comes to technological innovation this gets easily observeable in the ever-bigger machinery and the development which was made from wooden, neolithic hand hoes to camera-controlled, gps-steered cultivaters with a weight which even a 100 persons together couldn’t pull. The second driver of smart farming which is highligthed here, knowledge accumulation, processing and transfer, apparently isn’t new neither, but inherent to humankind in general. However, what got possible in terms of capacity of and connections between ‚knowledge‘ in the last few years is revolutionary. It offers possibilities which were unthinkable some decades ago. Applications in agriculture are for instance fully automated irrigation systems, determining very precisely the best date and amount of irrigation water, based on very accurate weather forecast, data from soil moisture sensors and digital crop models. But that’s old hat already. Now we have autonomous robots, doing plant protection tailored to the specific demands of single plants. But scaleable and hence executable in fields of hundred of thousands of individual plants.

We have access to high resolution drone and satellite images, processed to e.g. yield prediction models and protocols for variable rate applications of fertilisers and agents . That’s actually whats done in Demonstrator#1: „Organic Soya yield and protein-content prediction“, executed by BioSense Institute and Donau Soja. Therefore we relate environmental (soil, weather) and observed features with final yield and protein outcomes of soybean fields, in order to train prediction models. These models permit to predict the harvest amount and quality way before the harvest. This allows farmers to perform precision operations , like variable rate application of fertilizers and plant protection agents. Further it allows farmers to assess the benefit of such operations, as the model adapts the predicted outcome by considering the effect of the executed operation. Such, farmers get support by estimating the operations benefit in relation to its costs. Maps indicating the productivity of different field zones gives the farmer further insights on the fields he/she is managing and allows a precise assessment on which agricultural techniques (soil cultivation, seed rate, application of agents, harvest) fit best, in order to reach the desired outcome.



bottom of page