Incentivization of Large-Scale Crop Pest And Disease Digital Surveillance: Model Formulation and Algorithm Design

Abstract

Crop pests and diseases pose a significant threat to global food security, causing substantial yield losses, particularly in sub-Saharan Africa, where economic damages from pest and disease outbreaks could exceed $1 billion in the next decade. Smallholder farmers, who cultivate most of Kenya's farmland, are especially vulnerable due to limited resources and inadequate pest and disease management support. In Homa Bay County, where agriculture is integral to the local economy, challenges such as insufficient surveillance systems and poor extension services exacerbate food insecurity. Traditional surveillance methods are often ineffective, while mobile-based surveillance offers a promising solution for large-scale monitoring. However, despite the presence of barriers such as high data costs, low digital literacy, and poor connectivity, lack of incentive prevents farmers from overcoming these challenges and adopting the technology. This study addresses these challenges by developing a tailored incentive model, underpinned by an algorithm, to promote the adoption of mobile e-surveillance systems, thereby improving pest and disease management and enhancing food security. The study analysed existing surveillance practices, identified adoption barriers, and developed a participatory based incentivization framework to guide the design of an incentive model. A mixed-methods approach was employed, combining a descriptive survey design with experimental approach to evaluate the adoption and effectiveness of mobile e-surveillance among smallholder farmers. The study is piloted in Homa Bay County, Kenya, and involved 367 farmers and 31 extension officers purposively selected for their relevance to the research objectives. Data was collected through structured questionnaires for farmers and interview guides for extension officers, exploring current surveillance practices, adoption barriers, and perspectives on incentive models. The experimental component designed and prototyped e-surveillance application for mobile phones with various incentive structures to assess their impact on adoption. Instrument reliability and validity were ensured through a pilot study conducted in Kitutu Chache South Sub-County, where Cronbach’s Alpha coefficient was applied. Expert evaluations were incorporated to verify the instrument's accuracy and relevance. Data analysis, performed using statistical software, applied both descriptive and inferential statistics to identify key trends and factors influencing adoption, which informed the development of the incentive model. Findings revealed that farmers in Homa Bay County predominantly relied on traditional and labour-intensive methods for pest and disease scouting. Despite the availability of mobile-based solutions, adoption remains limited due to barriers such as inconsistent internet access, high data costs, low digital literacy and inadequate institutional support. This informed the development of an incentivized e-surveillance framework that anchored the incentive model within a collaborative digital ecosystem supported by strategic partnerships. The incentive model, formulated using game theory, fosters sustained engagement by leveraging strategic interactions between farmers and the e-surveillance platform to achieve nash equilibrium. Farmers earn points for participating in surveillance activities, which can be redeemed for benefits such as input subsidies, technical support, training, and subsidized data bundles. Rewards dynamically adjust based on participation, ensuring long-term engagement. By integrating this scalable and adaptable incentive model into digital surveillance solutions, the study aims to boost adoption and enhance the effectiveness of mobile e-surveillance for pest and disease management, thereby strengthening agricultural resilience and food security.