Brandenburg’s agriculture is characterized by highly heterogeneous locations and climatic challenges, making it particularly dependent on innovative approaches in precision farming. However, precise site information and climate-adapted fertilizer requirement models are lacking. This project develops a software package for determining fertilizer requirements by incorporating remote sensing and weather data, as well as soil moisture. It includes new functions for precise humus determination and evaluation. A sensor prototype for texture mapping is being functionally enhanced and developed into a working model. A new workflow for soil sampling avoids media breaks and information loss. A cloud-based IT system environment ensures the interaction of all components with operational systems and a new data processing application. Efficient nutrient management helps Brandenburg increase yields, reduce costs, and minimize nutrient losses.

Autonomous field robots have enormous potential for making crop production more sustainable and mitigating the shortage of skilled workers. Although these systems are now market-ready, significant challenges remain, particularly regarding their integration with machine management, process control, and the interaction between the robot and the implement. The ISObot research project aims to optimize the management of field robots through improved interoperability at various levels and automated process control. The starting point is commercially available robot-implement combinations and a web-based telemetry system. For a specific application in weed control in viticulture, a process control system is being developed that uses sensors and intelligent algorithms to monitor the work process, actively control it, and document the quality of work.

Communication between the robot and the implement for controlling various parameters will take place via ISOBUS. Another application in arable farming will focus on monitoring the weeding process and recording process parameters. In both cases, the systems will be linked to a web-based application to exchange process data and recorded parameters in real time, in accordance with the ISO 11783 standard. The application will be integrated into the machine management system to visualize process-relevant information via user-friendly interfaces and make it available for further agronomic use. Dissemination and transfer activities will be used to facilitate exchange with agricultural practitioners. Furthermore, the integration potential of this approach into industry-wide efforts to improve the interoperability of electronic communication in agricultural technology will be evaluated. Finally, recommendations for further standardization efforts should also be derived.

The aim of this research project is to contribute to increasing nitrogen use efficiency and thus to climate and groundwater protection in arable farming. To this end, new, model-based algorithms will be developed to enable demand-based, site-specific calculation of nitrogen fertilization for winter wheat, taking into account differences in site-specific yield and nitrogen mineralization potential, as well as current weather conditions. Furthermore, these site-specific fertilization algorithms for winter wheat, along with an existing site-specific fertilization algorithm for winter rapeseed, will be implemented in practical IT solutions using various data sources.

We expect this to result in savings of approximately 10–15% in the amount of nitrogen applied, while simultaneously increasing yield and quality with a limited overall nitrogen supply. At the same time, it can be assumed that the improved availability of agronomically validated fertilization algorithms will lead to a wider adoption of sensor-based, site-specific nitrogen fertilization. The integration of novel, commercially available sensor systems and collaboration with service platform providers established in both the data collection and marketing of sensor products and in fertilizer consulting (www.isip.de, geo-konzept) will make a concrete contribution to increasing nitrogen efficiency.

Crop production is facing increasing pressure due to population growth, climate change, and societal demands for more environmentally friendly agriculture. This is particularly true because the sub-goals of “higher yields,” “optimized product quality,” and “adaptation to changing production conditions (climate, societal requirements)” are sometimes in conflict with one another.

Breeding is one method that can be used to optimize production with regard to the aforementioned requirements. This is especially true for adapting to the increasing variability of precipitation associated with climate change (heavy rainfall, longer periods of drought), as well as for breeding for resistance to plant diseases and the potential reduction in pesticide use that this may entail.

Breeding progress is directly related to the amount of initial variability generated through crossbreeding and, consequently, the number of field trials conducted to select breeding lines. Simultaneously, the effort required for assessment and laboratory analysis increases with the number of trials. The visual evaluation of plant stands (assessment) and the collection and analysis of samples provide the crucial basis for subsequent selection. However, assessments are often subjective and prone to human error. Currently, breeding progress is largely limited by the scarcity of human labor in the field and in the laboratory.

The increasing developments and research in sensor-based phenotyping offer a growing number of possibilities for reliably and objectively recording plant or stand characteristics in a non-destructive manner, both qualitatively and quantitatively. Furthermore, previous studies indicate that individual sensors (multispectral) can simultaneously capture several traits.

The use of UAVs (Unmanned Aerial Vehicles) is particularly noteworthy in this context. UAVs can be used to survey crop stands with various sensors, such as multi- and hyperspectral cameras, laser cameras, and thermal cameras, quickly and frequently. However, the subsequent processing of the data (compositing, rectifying, georeferencing, and assigning it to individual plots) currently requires specialized qualifications and is very time-consuming. At present, no practical solution is available on the market that would enable the independent collection and analysis of sensor-based data. This is primarily due to the existing automation gap between data acquisition and processing. The main objective of this project is therefore to close this gap and develop a practical, largely automated, and optimized solution for the application of drone-based assessment in wheat breeding.

This is about a digital toolbox that enables every farmer and consultant to address and analyze relevant questions at their own location and with their own machinery.

For partners in the upstream and downstream sectors, this offers the opportunity to collect more data together with farmers and better address practical issues. At the same time, these surveys are designed to meet all official quality standards so that they can also be used in research and product development. Furthermore, this information will provide a foundation for better supporting and evaluating business decisions in a changing environment.

Alongside the operational group, an expert advisory board will be convened to provide advice and support on substantive issues. This board will consist of, among others, advisory institutions with their own experimental facilities (Chamber of Agriculture), private consulting firms (e.g., Hanse Agro, GBB), experimental sites (e.g., university experimental farms), and technology providers (e.g., Haldrup, Wintersteiger).

The goal of the SCHEBA² project is to develop an autonomous platform for monitoring individual plants in arable and vegetable farming. This platform utilizes image processing/AI to recognize and assess the condition of crops (primarily vegetables) with regard to factors such as biomass, diseases, and water requirements.

Based on this analysis, each individual plant within a crop can be selectively supplied with the precise amount of resources it requires.

In addition, the collected data will make it possible to distinguish between plants and weeds in the respective plant rows, so that selective, chemical-free weed control can be carried out in a further step.

Precise positioning of forestry machinery is fundamental for sustainable, soil-conserving, and efficient timber harvesting. However, reliable sub-meter positioning in real time within forest stands is not possible with currently available Global Navigation Satellite Systems (GNSS) receivers. The project aims to develop and field-test technical systems that compensate for the unfavorable conditions prevalent in forests for satellite positioning, such as shadowing, diffraction, and a poor signal-to-noise ratio, through innovative GNSS antenna technology and achieve higher accuracy through supplementary sensor fusion. The technologies to be developed within this project will ensure consistent, real-time positioning accuracy in the sub-meter range.

The aim of this project is to establish a digital monitoring system for the early detection and localization of quarantine diseases in commercial fruit growing. Based on image capture and machine learning image processing algorithms, species-specific disease symptoms will be recorded and mapped in large numbers, at various developmental stages, and with high spatial resolution in apple and pear orchards. The initial focus is on the detection of leaf symptoms. Data on the progression of disease symptoms will be compared with geoinformation from weather services. The long-term goal is to establish a labor-saving and cost-effective high-throughput method as a control system for fruit production and breeding, enabling the continuous spatial recording and documentation of infections caused by quarantine pests.

Electronic drawbar for agricultural machinery with environmental sensors and additional geoinformation

The electronic drawbar operates on a master-slave principle. A manned vehicle (master) sets the course, which an unmanned vehicle (slave) follows autonomously. To increase safety and the level of automation, the system is supplemented with geoinformation and sensors for environmental monitoring. The goal is to better meet the legal safety requirements for unmanned machinery in the field.

We are responsible for path planning, steering algorithms, telemetry data transmission between the tractors, and the integration of additional geoinformation.

Im Fokus steht die Entwicklung eines Verfahrens zur Anbauempfehlung und Entscheidungsunterstützung für eine standortspezifische Pflanzenproduktion. Anforderungen an ein solches Verfahren sind eine hohe räumliche Präzision unter Beachtung der großflächig skalierbaren, überregionalen und dabei kosteneffizienten Einsetzbarkeit.
In der auf einer Internetplattform basierenden Prozesskette sind wir für die Berechnung von Managementzonen-, Aussaat- und Pflanzenschutzkarten sowie die Wettermodellierung zuständig. Weiterhin implementieren wir Module für Ertragsdatenauswertung, Erntelogistik und Prozessdatenverwaltung.

Standortangepasstes N-Management auf Basis teilflächenspezifischer Informationsverarbeitung – System-, Hard- und Softwarelösung

Ziel ist die Detektion des betriebswirtschaftlich optimalen Stickstoffdüngereinsatzes im Winterweizen. Dieser wird von einem künstlichen neuronalen Netz (KNN) anhand eines Bodenprozessmodells ermittelt. Als Ergebnis wird die teilflächenspezifische Stickstoffmenge unter Berücksichtigung der von einem Stickstoff-Sensor während der Überfahrt gemessenen Werte ausgegeben.
Die Kommunikation mit dem Düngerstreuer übernimmt dann ein sogenannter In-Field-Controller. Dessen Entwicklung nach dem Standard ISO 11783 liegt genauso in unserer Hand wie die Mobilfunkverbindung zu einem Datenportal, das als Benutzerschnittstelle dient.

Prediction and Modeling of Hybrid Performance and Yield Gain in Oilseed Rape by Systems Biology

Intention ist die Untersuchung des Phänomens “Heterosis” beim Raps in einem systembiologischen Ansatz. Hierfür werden iterative Prädiktionsmodelle mit experimentell erzeugten Daten kombiniert. Wir übernehmen dabei die sensorgestützte Phänotypisierung von Rapshybriden per Drohnenbefliegung inkl. Entwicklung der Auswertealgorithmen, die Charakterisierung des Bodens sowie die Erhebung von Klimadaten.

Ertragsfaktoren von Raps als Nachwachsender Rohstoff: Phänotypisierung unter dynamischen Praxisbedingungen

Das Ziel dieses Vorhabens ist die Entwicklung und Praxiserprobung neuartiger nicht-invasiver Phänotypisierungsmethoden für die Erfassung von ertragsbildenden Faktoren beim Raps sowie deren Einsatz zur Identifizierung neuer genetischer Variation für die Züchtung. Diese neuartigen nicht-invasiven Phänotypisierungsmethoden beinhalten Sensoren für ertragsrelevante Merkmale, die zum Beispiel an einem UAV montiert werden und beim Überfliegen einer Versuchsparzelle objektive Messdaten liefern.