Digitally-enabled crop disorder management based on farmer empowerment for improved outcomes

Abstract

Crop disorder incidents such as pests and disease attacks are the major reason of crop losses that require timely actions and can adversely affect agriculture production. In order to address this problem, a model was designed that empowers farmers to identify crop disorder incidents instantly and manage them effectively by providing relevant information in context. In contrast, the existing approaches reported in the literature rely on identifying crop disorders from images that depict the presence of symptoms in the crop. However, due to the inherent characteristics of the images, these approaches are effective only in more controlled environments and provide limited support in crop disorder identification. We have created a crop disorder search space model that is composed of mapping between different crop disorders and symptom(s) that provide unique identification characteristics specific to each crop disorder. We call these unique mappings as disorder identifiers. This model was later converted into a mobile-based artifact, and the information required to perform the search operation on the search space was obtained from farmers through it. The artifact was deployed among a group of farmers to evaluate how well it could aid in identifying crop disorders. It was noted that the developed artifact was able to identify most crop disorders instantly, mitigating the issues associated with crop disorder identification. In the rest of the cases, it gives subject experts the ability to identify crop disorders. The experiments conducted on the effectiveness and usability of the artifact indicate that disorder identifiers providing clear and consistent representations of the presence of crop disorders can be used to identify them rapidly. Further, it has been also demonstrated that farmers are capable of correlating their field observations with a list of crop disorder identifiers provided through the artifact. The correct selection of the disorder identifier will lead farmers to know about the presence of crop disorder in the field and recommend control measures instantly. Moreover, farmers’ perception of various impact indicators showed that, as compared to previous cultivation seasons, yield quality and quantity losses were reduced due to the reduced crop disorder incidents. The application of agrochemicals and associated expenses of farmers were also significantly reduced, thereby increasing their revenues.