Abstract
The Food and Agriculture Organisation (FAO) trust that farmers should implement sustainable methods of farming to reduce food shortages worldwide (Mateo-Sagasta et al., 2018). Smart agriculture helps to provide sustainable means of producing food using technologies such as cloud computing, big data analytics, and AIoTs (Friha et al., 2021). The use of these technologies facilitates efficiency and improves farm productivity. However, this comes with cyber security challenges, which farmers need to confront.
The Design Science Research methodology was used in this research. The research reviewed the uses and application of Agricultural Internet of Things (AIoT) in crop and livestock farming and investigated the different AIoT architectures utilised in livestock and crop farming. Further to that different threats to AIoT were modelled using STRIDE-LM. Existing literature evinces that the currently existing cyber security frameworks are too broad, and the control measures are not within the scope of farmers. The research proposes a Model for a Safe AIoT Operating Environment (MSAOE) that can help farmers improve the security of Agricultural Internet of Things (AIoT) devices.
The model has three building blocks which are, technical and physical aspects of AIoT security, the role of people in creating a safe AIoT operating environment and the criteria to choose secure AIoT devices. The physical and technical controls play a key role in creating a safe AIoT safe operating environment. The techniques proposed under this building block complement each other to ensure a secure, safe AIoT operating environment. Emphasis by all farmers should be placed on user security awareness, training, and clearly defined responsibilities to help in fostering a culture of security throughout the organisation. AIoT labels with three security levels were proposed, i.e. green color-coded, amber colour coded and red colour-coded labels. The weighted approach was used to determine the security levels of the labels. The labels have a name, background colour and scannable QR code to provide more information about the label. The model was evaluated using the quick and simple evaluation strategy which performed a summative evaluation after the completion of the model development process using expert reviewers, who provided expert opinions regarding the applicability of the model in creating a safe AIoT operating environment.
6
The proposed MSAOE model presents notable contributions to enhancing the security of AIoT particularly in developing countries such as South Africa. The model is comprehensive, clearly defined, and specifically tailored to suit agricultural environments in developing. It provides a robust foundation for protecting AIoT devices against cyber threats, ensuring a more secure digital transformation in the agricultural sector. The model addresses cyber, physical, human, and selection-related security aspects in AIoT devices.