Digitization in agriculture
The digitization in agriculture builds on already existing digital production methods which are now recognized and accepted. They already show a certain complexity, especially in information technology, and lead to better operating results. These are Precision Farming , Precision Livestock Farming and Smart Farming . These processes are to be seen as important components of digital agriculture. However, digital agriculture is significantly expanding these existing systems with new and more comprehensive components.
Differentiation of Precision, Smart and Digital Farming
Precision farming
Since the beginning of the 1990s, the term precision farming has stood for the use of the so-called partial field technique using, for example, mapped variable dosing and precise application technology. In addition, automatic steering systems and section controls are used. This also includes complex machine functions in harvesting machines that automatically adapt to changing operating conditions, as well as the important topic of optimizing complex removal and supply logistics. Precision Livestock Farming means the use of modern sensor-actuator combinations from the exact allocation of high-performance feed components to automatic milking and health monitoring.
Smart farming
The term Smart Farming came up in the 2000s with the sensor-based real-time systems for fertilizer and crop protection application . With this method, for example, the biomass distribution of a crop is recorded and, depending on the sensor value, a quantity of fertilizer is applied in real time. For this purpose, there is no longer any need to carry out laborious and costly soil sampling in the partial field technique. The farmer can define the quantity level and the distribution by simply calibrating the sensors. Such a system thus represents a combination of automation of the process and decision support. In animal husbandry , the term smart farming with sensor-actuator combinations is used from data acquisition to decision support to automated execution and the combination of executive (e.g. Milking robot ) and evaluative functions (e.g. heat detection).
Digital farming
In digital Farming , or also called Farming 4.0, a technology is meant to complement the existing procedures for another four main components:
a) The Internet of Things (IoT) or machine-to-machine communication (M2M)
c) Big data analysis and artificial intelligence (AI)
d) Robotics with mobile and stationary units.
Development approaches for products that can be used in practice are already recognizable in all areas. However, it is expected that in the near future these will grow significantly in maturity and importance.
Environment and society
The processes in nature such as the weather as well as other biotic (living) and abiotic (inanimate) factors of agricultural production can be better observed through digitalization. This is possible through the use of sensors, remote sensing data and digital simulations. The farmer can react earlier and better. As with the weather forecast, only extensive and long-term data collection of many relevant factors at the location helps to enable big data analyzes, which in turn can be used as the basis for better location-specific decisions.
Digitization can also help to make natural relationships easier to understand and describe. This enables potential for optimization for various production goals such as the coupling of environmental protection and productivity. New cultivation systems without the use of pesticides are also conceivable, since overall biological relationships can be better described, possibly predictable and thus used more successfully for various production goals. More intelligent agricultural machinery and process technologies are being developed, which, in addition to productivity and animal performance, can also take into account the needs of the ecology of cultivated landscapes as well as of farm animals and humans.
Components of digital farming
Agriculture 4.0 is the digitization of agricultural production processes in crop production and livestock husbandry. This includes various components that are explained below.
Internet of Things (IoT)
The "Internet of Things" ( Internet of Things or IoT) is a collective term for an information technology infrastructure. It enables physical and virtual objects to be electronically networked and automatically communicated.
Special, advanced technology is required so that agricultural machines can also be controlled accordingly. This must have access to certain standardized vocabularies or so-called ontologies. This enables programs to “look up” the type and meaning of the data in real time and thus learn their content. For example, instead of general text sequences such as “Sieglinde” for a potato variety or “Fertilization with farm manure”, there are references to terms available online, which are technically represented by their standardized URIs ( Uniform Resource Identifier ). Only through these semantic techniques with their uniform vocabularies can the greatest possible flexibility and future security in data storage be achieved.
An established communication standard between machines today is the so-called ISOBUS according to ISO 11783. This enables machines (tractor, device and office IT) to communicate across manufacturers. This standard is not only important for the control of agricultural machinery (Figure 4). Its importance in the context of digitization and an IoT is enormous. However, the communication properties of the ISOBUS are limited and are therefore not suitable for all digital requirements.
Cloud systems
With a drastic increase in digitization, there are significantly higher demands on information systems, such as a steadily growing need for computing power and storage capacity. Cloud computing offers a solution for this . Large IT resources that can be made available on demand result in extremely flexible and scalable hardware and software infrastructures. Mobile access to data with different end devices is important today. The basic types of cloud systems are global cloud, regional cloud and private cloud. A private cloud (home server) with internet access from outside can promote failure safety and ensure minimal requirements for data security.
There are already various cloud-based platform concepts such as B. pure data platforms for the cross-manufacturer exchange of machine data. There are also specialized trading platforms such as so-called digital trading markets for buying and selling goods. Management platforms such as Farm Management Information Systems (FMIS) also play a large part. In a way, they are a continuation of the electronic field maps, but on a digital platform level (Figure 5).
Some of today's centralized portals only work excellently within the machine fleet of the respective manufacturer. From the farmers' point of view, however, it is only advantageous if a comprehensive and manufacturer-independent data exchange is made possible, since machine systems from different manufacturers are often used on farms in arable farming.
Big Data and Artificial Intelligence
Today and in the future, data is increasingly being recorded, stored and evaluated by machines, sensors, computers, smartphones and similar technology. This results in very large amounts of data with corresponding data stores that can only be evaluated for meaningful use using so-called big data analysis. The potential use contained in this data is also enormous for agriculture, both for site-adapted arable farming and for improved animal husbandry. If these analyzes are correctly linked and summarized into meaningful evaluations, they support the farmer in his strategic (long-term) and operational (short-term) decisions. In agriculture, a considerable amount of data is already accumulating today, as is the case with modern herd management with automatic milking systems and ISOBUS-controlled field machines in arable farming. So far, however, they have not been able to be used because there is too little networking of the machines (IoT) and few storage options (cloud).
Artificial intelligence (AI) or machine learning is also part of the terminology of digital agriculture. An AI unit learns from recorded or selected training data by looking for patterns and recurring structures from which regularities can be derived. With the networking and storage of a company's data over several years, machine learning algorithms can be trained in such a way that operational processes become more transparent in order to generate success factors that lead to better decisions. Furthermore, the algorithms can also be used to e.g. B. to recognize plant diseases, weeds or pests or to predict depending on weather, location and stand factors.
Automation and robotics
The possibilities of robotics represent a new level of mechanization as well as automation. Autonomous machines are scalable in size and are therefore also an issue for small and medium-sized farms. However, like many new applications in digital agriculture, robotics also needs a stable digital infrastructure that must ensure secure communication between the machines and the integration of communication into the entire operational IT system.
It is already becoming apparent that autonomous robots will mostly be small in size and electrically powered. This leads to considerable reductions in investment costs and vehicle weights. The lower the acquisition and investment costs, the lower the area coverage can be. This effect helps with the acceptance of autonomous agricultural robots, because many tasks that a robot has to perform are much more precise at low travel speeds, but above all can be carried out with less energy. Such devices are light and therefore gentle on the floor. The scaling to larger areas is not achieved by larger and faster machines, but by a swarm of similar and small robots that cooperate with one another.
With the move away from large working widths with heavy machines, the fields no longer have to be as large and cleared as possible. Even traditional landscape elements (e.g. hedges, ponds) can be introduced, as they do not have a negative impact on productivity and area performance for small autonomous machines. Thus, a considerable increase in the biodiversity of our agricultural landscapes can be achieved.
We are still at the beginning of the application of robotics, although there are already the first applications, such as autonomous robots with row hoes in field vegetable cultivation.
Public geospatial data
Public spatial data can be viewed as basic data and, if provided, serve as a valuable source of information. These are mostly field outlines, soil information, erosion cadastre, etc. a. be. In this way, publicly available geodata can be used as the basis for digital location-based services.
Some federal states, such as Rhineland-Palatinate and Baden-Württemberg, have implemented concepts for the provision of public spatial data as an example: With the GeoPortal "MapRLP", the farmers have access to relevant basic geographic data freely and in open formats, with local caching and transfer to mobile devices are.
Road and path networks, aerial photographs, property maps, etc. are considered official geographic reference data. a. So-called official geospatial data can also be provided, such as erosion cadastre, soil information, protected area boundaries, reference values (N min ) and the like. a.
Blockchain
By block chain , there is potential for automation to be respected agricultural documentation requirements: Thus, for example, a mounted in the container sensor to measure the temperature of food, write the data into the block chain, thus a complete compliance with the cold chain documented. If it were not adhered to, an appropriately set up smart contract could automatically sound the alarm.
Data security and data protection
The meaningful use of data requires extensive data acquisition and storage. This certainly creates risks. Information can be generated from personal and area-related data, such as the quantity and quality of harvested products and, in the long term, the intensity of machine use. That means whoever has location-specific data has a knowledge and competitive advantage. Many are interested in the data: the farmer, the contractor, the agricultural machinery dealer, the machine ring, the agricultural trade, politics, the economy, the authorities and the cloud provider.
The protection of the trade secret - no data use without consent - must be guaranteed for cloud systems with data access from outside, since detailed data on fields, measures, yields, etc. are also a commercially valuable asset. Clear access rights as well as a dedicated data exchange and assignment with service providers must remain under the control of the farmer.
It should always apply to the farmer that he benefits from his data. The farmer is the author of the data. Therefore, decentralized structures with data protection are recommended, as they promote the diversity of providers and reduce the risk of dependencies on a central data partner. In the case of paid offers, you should inquire about the data sovereignty and the server location, as the server location can significantly impair data protection.
Many central data platforms are currently offering their service free of charge if users agree to the transfer of the data. But one thing must be clear to the user: platform users today have no clear legal rights when it comes to non-personal data. For example, he cannot prohibit the platform operator from using the operational data provided for commercial purposes. There is also no possibility of legal action that users have to be economically involved in the utilization of the data. The status of the legal situation is that once recorded and released data are transferred into the possession of the new data holder, i.e. the platform operator.
Many operations managers today do not go to the cloud without any certainty about the secure and protected storage of data. As a result, it is not data protection, but data insecurity that slows down digitization and potential remains untapped.
Furthermore, psychology also plays a role in the storage of production and operating data, as some farmers fear that this will arouse the authorities' desire for information. This has to be rejected for reasons of business and trade secrecy, as the widespread use of digital data is fundamentally hindered considerably. The protection of business and trade secrets must also apply to farmers as to other private companies in the free economy. On the other hand, the farmer must of course comply with the legal requirements for information and documentation.
There is a data protection code of conduct within the EU. However, this is largely non-binding. It refers to the data protection regulations in force in the EU and contains a checklist for farms with important service agreements.
Resilience
In many countries, a country's agriculture is to be understood as a critical infrastructure and of important importance for the state community. In the event of failure or impairment, sustainable supply bottlenecks, significant disruptions to public safety or other dramatic consequences can occur. Storage plays a major role in this.
According to the Federal Office for Information Security (BSI), agricultural production is part of the national critical infrastructure in Germany and must therefore be guaranteed even in exceptional situations. Exceptional situations are natural events, technical or human failure, terrorism, crime and war.
In order to counter the risks of central cloud computing , it makes sense to set up a decentralized implementation of corresponding data infrastructures. The point here is that the IT system basically remains operational with a possibly reduced range of functions even without an external Internet connection. In such systems, even if the network connections fail, the data temporarily stored locally in the components remain available and thus guarantee a certain degree of robustness against interference (Figure 7).
Advantages and disadvantages of digitization in agriculture
The management consultancy PricewaterhouseCoopers sees agriculture as a pioneering role in digitization.
The following benefits are expected:
- Work can be made easier, e.g. B. through reduced documentation and planning effort, task status lists (ToDo) and automation of processes
- Better decisions through greater transparency of the operation should be achieved through monitoring, warnings and recommendations
- By exchanging data and information with third parties, simple order processing, certification and traceability can be made possible
- Overall, it is expected that process improvements can be achieved through continuous monitoring and an increase in knowledge, for example through AI applications.
The disadvantage is that the use of digital options is associated with considerable costs for most farmers, which only pay off when the farm is of a certain size. In addition, the data is not always available where it is needed.
The consequences of a wider use of digital solutions in agriculture are also discussed controversially. The KTBL noted in 2017 that, for example, the loss of autonomy in processes and decisions being overwhelmed by the increasing complexity of decisions and polarization can represent the working principle risks of digitization in agriculture. Organizations such as the Agrarbündnis eV also see in particular the “takeover and commercial use of data, information and experience on climate, genetics, soils, sowing and harvesting times that have been in the hands of farmers and indigenous peoples for thousands of years mostly still are. ”In addition, the question of data sovereignty and sovereignty often remains unsolved: In 2019, Bitkom determined that the sovereignty for data that farmers generate when using their equipment and when managing their farms and identifying the person allow it to lie with the farmers themselves. Bitkom sees this as a “basic requirement for trust in the use of digital applications in agriculture.” DLG specialist committees have a similar view with the position paper Digital Agriculture - Opportunities. Risks. Acceptance. , a specialist article for the Austrian Chamber of Agriculture on this subject describes the problem of the unregulated use of operational, machine and business data in agriculture.
See also
Web links
- Digitization in the agriculture and food industry Federal Office for Agriculture, Bern (CH) (accessed November 30, 2019)
- Digitization in Agriculture Federal Ministry for Sustainability and Tourism, Vienna (AT) 2018 (accessed November 30, 2019)
- Digitization in Agriculture - Seizing Opportunities, Minimizing Risks Federal Ministry of Food and Agriculture, Berlin (DE) 2019 (accessed November 30, 2019)
- Agriculture and digitization (PDF) 27th Hülsenberg Talks 2018, series of the H. Wilhelm Schaumann Foundation for the Promotion of Agricultural Sciences (accessed on November 30, 2019)
Individual evidence
- ^ DLG eV: Digitization in Agriculture. In: DLG-Merkblatt 447th DLG, 2019, accessed on November 26, 2019 .
- ↑ Zimmermann B., Schlepphorst R., Meinardi D., Kraft M .: With sensors against drought stress. In: Top Agrar No. 46 (10), 2019, pp. 68–71.
- ↑ Decipher patterns in nature with BIG DATA. zalf.de, 2017, accessed on February 25, 2020 .
- ^ Leibniz Institute for Agricultural Engineering and Bioeconomy (ATB) in Potsdam-Bornim: high-tech spies in the field. biooekonomie.de, 2016, accessed on February 25, 2020 .
- ^ Thünen Institute: High-tech spies in the field. thuenen.de, 2019, accessed on February 25, 2020 .
- ^ Service centers in the rural area of Rhineland-Palatinate: MAPrlp. 2018, accessed November 26, 2019 .
- ^ DLG eV: Digital Agriculture - Opportunities. Risks. Acceptance. In: Position paper. DLG, 2017, accessed on November 23, 2019 .
- ↑ COPA-COGECA: EU Code of conduct on agricultural data sharing by contractual agreement. 2018, accessed on November 23, 2019 .
- ↑ PricewaterhouseCoopers: Study on Smart Farming: Agriculture is playing a pioneering role in digitization . In: PwC . ( pwc.de [accessed January 15, 2020]).
- ↑ https://www.fnp.de/lokales/hochtaunus/5g-netz-taunus-fluch-oder-segen-landwirtschaft-region-zr-13339159.html
- ↑ Dr. Martin Kunisch, Dr. Stefanie Reith, Dr. Jürgen Frisch, KTBL: Digitization in Agriculture: Opportunities and Risks. KTBL, 2017, accessed November 29, 2019 .
- ^ Stig Tanzmann and Bernd Voß: Digitization of Agriculture. AgrarBündnis eV, 2018, accessed on November 29, 2019 .
- ↑ Data sovereignty and data use in agriculture. Bitkom.org, 2019, accessed November 29, 2019 .
- ^ DLG eV: Digital Agriculture - Opportunities. Risks. Acceptance. In: Position paper. DLG, 2017, accessed on November 23, 2019 .
- ^ Rainer Winter, DLG: "Digital Harvesting" through Big Data. LKO.at, 2018, accessed on November 29, 2019 .