Shortly, every milk can, every camera, every animal, agricultural device and a wide variety of sensors in the fields will probably be integrated into a digital agricultural ecosystem as an independent IoT device. And that is also urgently needed!
Modern technology in agriculture? Not a question, just a fact. Even before the great hype of IoT technologies and long before the automotive industry had the first assistance systems in productive use in the direction of autonomous driving, there were already autonomously driving agricultural machines in agriculture, largely fully automated further processing on the farm and first attempts to improve the well-being of the animal’s record or to predict calving through the movement profile and the rumination behaviour.
Animals are uniquely identified via radio technology and “register” by approaching the automated feeding troughs and receiving their intended food. However, since a considerable part of agricultural data is generated in the fields (increasing online connection of agricultural machinery) and there is a need to process it on-site and forward it if necessary, 5G data communication is the next logical, essential step.
Cloud technologies play a crucial role in all of these scenarios. Only with them is it possible to process all this information, efficiently develop and roll out solutions, monitor them and scale them almost infinitely. Accordingly, cloud-based software offers displace the tried and tested farm programs that farmers have installed on their PCs.
The cloud enables both human-to-machine and machine-to-machine interaction on the farm and any communication to the outside. But what happens when precisely this cloud environment is no longer accessible precisely due to the still considerable gaps in mobile phone coverage? There is no secure connection if you are in a mobile phone fringe zone. In addition, there are power outages, failures of individual sectors of the mobile network and the rapidly increasing number of cyberattacks.
Why Is An Internet Connection So Essential For Farming?
Let’s go into two aspects of plant production: 1. production and supply chain management and 2. environmental and climate protection (Green Deal).
Regional Production And Supply Chain Management
The complexity of plant production is determined by nature – accordingly, the actors, from hobby farmers to large companies, face similar challenges. In addition to market risks, agricultural producers are primarily struggling with the uncertainties of the weather and unpredictable disease and pest infestation. Increasing environmental and climate protection requirements increase management requirements equally for all farmers.
The framework conditions for small and large companies (see table: only about 1/3 are large companies) differ mainly in the machine and IT equipment. Since, in contrast to the large farms, most of the smaller farms in certain regions only have their machinery to a limited extent. They are more or less dependent on third parties for cultivation, harvesting and logistics. This means they have to buy/rent machines and logistics services on time. In regions with rather small-structured agriculture, corresponding structures for the inter-company use of machines have been established with agricultural contractors and machine rings.
The innovations of smart farming as part of the digital transformation in agriculture are primarily focused on large farms with their mechanisation. Accordingly, the cloud platforms of the farm management information systems are usually adapted to one or a consortium of agricultural machinery manufacturers. This is understandable, as the management of machine fleets is often the focus of crop production processes.
Farmers are still reluctant to accept IT platforms with purely cloud-based data storage due to concerns about data sovereignty. In addition, the preference for particular machine fleets on the competing IT platforms does not meet the requirements of smaller farms, which use service providers with a wide variety of machine fleets on their land. Farmers who depend on machinery across farms need a manufacturer-independent digital ecosystem that connects as many players as possible, especially at the regional level.
This trend clarifies that internet dependency in digitization in the agricultural sector significantly increases the vulnerability of primary agricultural production, which is an essential part of the critical infrastructure (KRITIS) “nutrition”. The outsourcing of data to the cloud, the associated dependence of machines and production processes on the Internet, and a failure of IP-based telephony and data communication could significantly impact food security in a longer-term Internet failure.
IoT Sensor Networks For Environmental And Climate Protection
A longer-term failure of the voice and data communication networks is a significant risk, especially for regions with small-scale and predominantly labour-structured agriculture. Since local LoRaWAN networks will increasingly be used for modern IoT sensor networks as part of intelligent farming shortly, it would be a socially appropriate option to use this resilient network for emergency communication in rural areas in the event of a need or need crisis.
By using battery-powered LoRaWAN networks, the farmer can receive all the necessary (sensor) data to continue his business over a certain period, even in such a case. The prerequisite, however, is that such networks should be expanded as comprehensively as possible, which could be accelerated by more intensive use of the IoT sensor data obtained from them.
Since agriculture has to make a significant contribution to climate protection, the comprehensive acquisition of environmentally relevant IoT sensor data (e.g. soil moisture and temperature) could represent a service that, as an open data offer, could not only be used to strengthen climate resilience but even in the short term benefit society as a whole in a wide variety of sectors (e.g. ice warning). The EU, for example, with its Common Agricultural Policy (CAP) programs and the Green Deal, could provide the impetus for the development of such an IoT infrastructure with a view to environmental and climate protection.
Operationalization Of A Resilient Edge Computing (REC) In Agriculture
Politicians in Germany have recognized the problem of Internet dependency in primary agricultural production as part of the KRITIS “Nutrition” and have taken the first steps to reduce vulnerability. Significantly since this must not be increased in the critical infrastructures through the digital transformation, both the Food Security and Prevention Act (ESA) and the IT Security Act (IT-SIG) provide directives for prophylactic measures in the area of KRITIS “nutrition”.
Accordingly, the Federal Ministry of Food and Agriculture (BMEL) is promoting the development of an infrastructure for decentralised data management and regional networking in agriculture for precautionary reasons and with the approval of the Conference of Agriculture Ministers. In addition, the BMEL supports the practical transfer of a resilient GeoBox infrastructure through the Experimental Field Southwest in Rhineland-Palatinate.
The GeoBox infrastructure is an intelligent data hub in the form of cloud-independent mini-servers that ensure secure data provision from heterogeneous data sources. This includes time-critical agricultural production and management data, basic geodata, and knowledge sources and platforms. For example, the responsible district or the public sector can contribute.
“Through our networking approach, we are developing a multidimensional digital ecosystem with the GeoBox infrastructure. We currently have a network via the GeoBox project, the federal government’s “Digitalization in Agriculture” experimental fields, and the KlimAgrar research support. There is also cross-state cooperation, such as the competence network West, an association of several federal states that jointly promote introducing the Geobox infrastructure in practice. We refer to this concept as resilient smart farming (RSF),” explains Daniel Eberz-Eder from the DLR Rheinhessen-Nahe-Hunsrück, coordinator of the projects in the digitization of agriculture.
The technological basis for decentralised edge computing is the open-source framework “Open Horizon”. IBM developed it and gave it to the Linux Foundation to encourage further development in a more diverse ecosystem. Open Horizon forms the basis for the IBM Edge Application Manager, which manages containerized applications on thousands of edge devices. In addition, the framework creates the prerequisite for the offline-first principle for resilient smart farming (RSF) based on a decentralised edge computing infrastructure via so-called “yard boxes”. All programs running on it can also be used without an Internet connection. Accenture
The Offline-First Function Of The Edge Devices Is Decisive
Depending on the requirements, a Hofbox can be a Raspberry PI IV, Jetson Xavier, Win10 computer or a larger server. All sorts of devices, machines, vehicles and even your small data centre within the farm can be considered an edge device.
As soon as an edge device has an Open Horizon agent installed, Docker containers can be transferred to it, managed and monitored. A cloud connection is never necessary. A reference to the higher-level instance, the IBM Edge Application Server, is also not required for operation. In this way, hundreds of thousands of edge devices and every single milk can be actively managed.
This offline-first function of the edge devices based on Open Horizon is crucial for resilient edge computing or intelligent farming. The vast majority of solutions available on the market require a cloud connection to manage and monitor edge devices. For the farmer, this means that he can set up his edge environment with the help of his farm box. Expanded with the LoRaWAN Gateway, this can connect local sensor networks and process the data locally. The farmer does not have to program the applications himself. The farm box pulls them from a central location. The state and private-sector organisations provide a wide range of data necessary for the daily work of a farmer.
By bringing together different data sources, the farmer has significant added value. Publicly provided geoinformation data combined with regional sensor data provide agriculture with a timely, improved basis for decision-making. Therefore, the goal should be that standardised databases of geodata can be delivered nationwide automatically and always synchronised to edge devices. Here, too, an Edge Application Manager is conceivable, which supplies all connected farms in the region with a comprehensive catalogue of existing applications or even specifies some. Operation on the farm works even after the cloud infrastructure has failed for days. As soon as a connection is established again, necessary updates are exchanged via the cloud,
Sensors and actuators can be found in the lowest communication layer under the individual edge devices, for example, the moisture sensors in the soil that belongs to the irrigation system of the fields or a simple counter at the end of a processing line that records the number of tomato or apple crates. This data also migrates to the Hofbox but remains there. They are only sent to the cloud when required.
Several Advantages For The Farmer
This resilient innovative farming concept, as a kind of backup solution for existing cloud infrastructures, meets all the criteria set in advance: a standardised infrastructure, decentralisation, offline capability, data sovereignty and user-friendliness. Above all, data sovereignty is a crucial point for many farmers. You can always decide when which data is sent where. Nobody can get to the Hofbox itself from the outside. If the information remains on the Hofbox and is not sent to the cloud, it does not have to be encrypted. Each user decides whether the information is sent to the cloud for backup purposes. Thanks to easy handling and local data sovereignty, this “simple” IT environment improves access to smart farming even for smaller companies,
Resilient Edge Computing as a necessary extension of Cloud Computing
The background to regional networking at the district level is provided by the ESA, which states that district administrators – comparable to the pandemic – must coordinate agricultural production and food supply to the population at the regional level in a crisis. Using the example of the Donnersberg pilot region, the model of a dRaaS is to be further developed and tested in a public-private partnership.
For the system to be suitable for crises, as many farms as possible must voluntarily connect to resilient edge computing. For this reason, District Administrator Rainer Guth from the Donnersberg district in the Palatinate attaches particular importance to the direct tangible benefits of this resilient digitization, which is expressed in particular in the fact that farmers, even in regions with small structures, benefit from intelligent farming with a demanding inter-company use of machines and thus ensure operational and regional added value be able.
This makes Resilient Edge Computing a further expansion stage of cloud computing mandatory for KRITIS industries. This decentralisation approach serves society with improved security of supply, climate resilience with comprehensive and fail-safe IoT sensor networks. Finally, due to the data sovereignty inherently implemented in the REC, acceptance among farmers, as Dr Christian Koch from Hofgut Neumühle in Donnersbergkreis confirms: “I see the advantage in the fact that my data is stored locally on the Holbox. Suppose you imagine the Internet failing in Germany. In that case, it is still possible for me to document my areas with this local data set and continue cultivating the areas in the future.”
At the same time, this technological approach can be a blueprint for numerous other areas of application in industrial companies or the utility sector – wherever the cloud already plays an important role and failures can have a critical impact on production or regional supply.
