Resilient Smart Farming

Digitalization is a major issue in the agricultural sector. The desired positive aspects are an increase in efficiency and effectiveness and also a more resource-friendly production of food. The terms agriculture 4.0 and smart farming are currently used to summarize the technological trends of knowledge management and intelligent, networked systems for agricultural production.

This dissertation is about the infrastructural resilience of data-driven agriculture and requires a basic understanding of both the application scenario of agricultural practice and digital resilience as it is used in this work. Therefore, this chapter first provides a brief introduction to agriculture in Germany by referring to recent statistical investigations. Afterward, this chapter introduces terms, such as smart farming, precision farming, and farm management information systems (FMIS) and characterizes digital technologies in the field of resilient network approaches.

The motivation of this dissertation is to investigate threats to food security posed by the ever-advancing process of digitalization. Therefore, the strategy is to analyze the sector’s current preventive measures, investigate security aspects of modern technology for smart farming, and develop artifacts that help to overcome the limitations of current technologies. This chapter presents the research approach, context, and methods used in this work.

This chapter presents the main findings of the conducted research and its implications for the agricultural ICT. Key findings are separated into the following three areas: (1) Farms’ Digital Dependencies and Vulnerabilities, (2) Towards a Resilient Software Architecture for Farm Management, and (3) LoRaWAN-based IoT Developments Towards Resilient Communications.

This chapter answers the RQs by discussing the findings and draws a bigger picture of how the results could enhance the stability of the farming system by designing crisis-capable software systems. At the end of this chapter, limitations of the conducted research are described as well.

The topic of this dissertation concerns the resilience of modern agriculture, both in the present and in spite of software-based systems. The aim is to improve technical resilience, with the goal of reducing the susceptibility to failure and vulnerability of new software-based systems, and providing greater flexibility for unforeseeable events. It is only through the progressive development of interconnected systems in the field of smart farming that the questions of systemic security arise.

Agriculture is subject to high demands regarding resilience as it is an essential component of the food production chain. In the agricultural sector, there is an increasing usage of digital tools that rely on communication and energy infrastructures. Should disruption occur, such strengthened dependencies on other infrastructures increase the probability of ripple effects. Thus, there is a need to analyze the resilience of the agricultural sector with a specific focus on the effects of digitalization. This study works out resilience capacities of the interconnected technologies used in farm systems based on the experiences and opinions of farmers. Information was gathered through focus group interviews with farmers (N = 52) and a survey with participants from the agricultural sector (N = 118). In particular, the focus is put on the digital tools and other information and communication technologies they use. Based on a definition of resilience capacities, we evaluate resilience regarding energy and communication demands in various types of farm systems. Especially important are the resilience aspects of modern systems’ digital communication as well as the poorly developed and nonresilient network infrastructure in rural areas that contrast with the claim for a resilient agriculture. The result is a low robustness capacity, as our analysis concludes with the risk of food production losses.

Technological progress can disrupt domains and change the way we work and collaborate. This paper presents a qualitative study with 52 German farmers that investigates the impact of the ongoing digitalization process in agriculture and discusses the implications for privacy research. As in other domains, the introduction of digital tools and services leads to the data itself becoming a resource. Sharing this data with products along the supply chain is favored by retailers and consumers, who benefit from traceability through transparency. However, transparency can pose a privacy risk. Having insight into the business data of others along the supply chain provides an advantage in terms of market position. This is particularly true in agriculture, where there is already a significant imbalance of power between actors. A multitude of small and medium-sized farming businesses are opposed by large upstream and downstream players that drive technological innovation. Further weakening the market position of farmers could lead to severe consequences for the entire sector. We found that on the one hand, privacy behaviors are affected by adoption of digitalization, and on the other hand, privacy itself influences adoption of digital tools. Our study sheds light on the emerging challenges for farmers and the role of privacy in the process of digitalization in agriculture.

Farm Management Information Systems (FMIS) are an important core component of modern farming companies as they allow, e.g., to document activities, create fertilization plans, and feed digital equipment with required data. Since the entire agricultural sector is an essential component of food production, high standards of resilience should be established in the involved companies. Accordingly, the used software should also be designed with high standards on reliability and crisis capability. Based on a literature review, we found that software for farmers with certain resilience needs is lacking. Thus, we designed and evaluated a new FMIS concept with the user-centered design method. By conducting focus groups (two rounds, total N = 57) in 2017 and 2019, we raised specific front-end and back-end requirements of farmers. Based on the requirements, we developed our concept for both front- and back-end in terms of a decentralized and offline-working FMIS. Through the evaluation with practitioners (N = 16) of the implemented concept, we derived findings and implications, highlighting the need for privacy, stability, and offline-capability, as well as the UI-requirement to be supportive, e.g., with easy to understand icons and terms.

Reliable IT-based communication in agriculture is becoming increasingly important for regular operations. For example, if a farmer is in the field during a network outage, such as a failure of the mobile network, an alternative communication channel is needed to continue to connect to IT components and required data. With increasing digitalization, Low Power Wide Area Network (LPWAN) technologies are being used more and more frequently, e.g. for sensor networks. The LPWAN technologies offer a high range and can be used autonomously for the most part, but do not allow classic TCP/IP communication. In this work, a popular LPWAN technology, namely LoRaWAN, is experimentally supplemented by AX.25 on OSI layer 2 (Data Link Layer) to allow end devices TCP/IP-based communication over long distances. The evaluation shows that classic low-bandwidth applications are thus functional and can enable reliable, crisis-capable data transmission.

Since communications infrastructure is subject to many impacts, e.g., destructive natural events, it can potentially collapse at any time. Especially in rural areas, the recovery of public network infrastructure can take some time, so a dedicated communication channel would be advantageous. We explore the possibility of transforming commodity LoRaWAN gateways into meshed network nodes for a digital emergency communication channel. In order to obtain the required parameters, we collected farm locations in Germany with OpenStreetMap. Based on the assumptions of LoRa communication range and considering our use case requirements, connecting farm communities seems theoretically feasible in many areas of our data set. To further analyze our idea, we ran simulations of two common DTN routing protocols with different scenarios. A proof-of-concept implementation allows smaller messages to be transmitted using real hardware and demonstrates that a decentralized communications infrastructure based on commodity hardware is possible.