Assessment of Precision Agriculture Adoption by Smallholder Farmers in maize production in Kenya

Thesis Abstract

The adoption of precision agriculture (PA) among smallholder maize farmers in Kenya is constrained by limited access to knowledge, high initial costs, and perceived risk, which impede sustainable improvements in input-use efficiency and yield. This study aims to examine the determinants, extent, and outcomes of PA adoption among smallholder maize producers, with a view to informing policy and extension strategies. Specific objectives are to (i) quantify the prevalence of PA technologies use (soil sensors, variable-rate applicators, drone-assisted imaging, and decision-support tools) among smallholders in key maize-growing counties; (ii) identify socio-economic, institutional, and farm-level factors influencing adoption decisions; (iii) assess the effects of PA adoption on input use efficiency, maize yield, and production profitability; (iv) evaluate perceived barriers and enabling conditions to sustained PA use; and (v) propose a policy and extension framework to foster scalable PA adoption. A mixed-methods research design is employed. The quantitative component uses a cross-sectional survey of 420 smallholder maize farmers selected through multi-stage sampling across Nyeri, Embu, Machakos, Kisii, and Kakamega counties, complemented by archival data from county agricultural offices. Structured questionnaires capture household demographics, farm characteristics, access to credit and information, and PA adoption status. The qualitative component comprises in-depth interviews with 30 extension officers and 20 PA technology providers to deepen understanding of constraint dynamics. Data collection instruments were pre-tested for reliability (Cronbach’s alpha > 0.70 for multi-item scales) and content validity through expert review. The study applies regression analysis to model determinants of PA adoption (binary logistic regression) and to estimate the impact of PA on yields and input use efficiency (stochastic frontier analysis for technical efficiency). Propensity score matching (PSM) is used to address selection bias in estimating the production and profitability effects of PA adoption. Theoretical grounding integrates the Technology Acceptance Model (TAM) and Diffusion of Innovations (Rogers), supplemented by the Resource-Based View to interpret firm-level capabilities. Qualitative data are analyzed using thematic analysis, with coding aligned to the socio-technical dimensions of PA adoption, and triangulated with quantitative findings. Preliminary expected findings anticipate a PA adoption rate below 25% among surveyed farmers, with determinants including higher education, access to credit, proximity to input supply networks, prior exposure to extension services, farm size, and risk preferences. Adoption is expected to be associated with measurable improvements in input-use efficiency, evidenced by reduced fertilizer dosage per hectare and optimized irrigation scheduling, and with modest but positive yield and gross margin gains, conditional on favorable weather and price environments. Barriers likely include high upfront costs, insufficient technical know-how, lack of reliable electricity or connectivity, and limited after-sales support, while enabling conditions may encompass micro-credit schemes, participatory extension models, and farmer field schools that incorporate PA decision-support tools. The study contributes to knowledge by providing empirically grounded estimates of PA adoption determinants and its farm-level impacts in a sub-Saharan African context, clarifying the role of enabling institutions and information ecosystems, and offering a validated framework for evaluating PA interventions among smallholders. The findings will inform policy-makers, development agencies, and private providers on scalable strategies—such as credit-linked PA package pilots, affordable sensor- and drone-based services, and extension curricula—that can mitigate adoption barriers and enhance maize productivity and resource-use efficiency. The main conclusion is that targeted, economically viable PA interventions supported by robust extension services and accessible credit can significantly improve production efficiency and profitability for smallholder maize farmers in Kenya. Recommendations include (i) scaling demand-driven PA demonstrations and farmer field schools with hands-on exposure to cost-effective PA tools; (ii) developing affordable financing models and subsidy schemes for essential PA inputs; (iii) strengthening extension capacity and local service provision for PA maintenance and technical support; and (iv) integrating PA indicators into county-level agricultural performance metrics to monitor progress and refine implementation strategies.