Assessing Precision Agriculture Adoption in Kenyan Smallholder Maize Farms

Thesis Abstract

Smallholder maize production in Kenya faces productivity and resource-use challenges amid rising input costs and climate variability, limiting gains from modern agronomic practices. Precision agriculture (PA) offers potential to optimize inputs, enhance yields, and reduce environmental impacts, yet its adoption among Kenyan smallholders remains uneven and poorly understood, creating a gap between technological capability and on-the-ground implementation. This study aims to assess the extent, determinants, impacts, and pathways for scaling PA adoption in Kenyan smallholder maize farms, with a view to informing policy and extension strategies. The specific objectives are (i) to quantify the prevalence and intensity of PA practices among smallholder maize farmers in Embu, Meru, and Kakamega counties; (ii) to identify socio-economic, institutional, and farm-level factors influencing PA adoption using a multinomial logit framework; (iii) to evaluate the relationship between PA adoption and maize productivity, input-use efficiency, and income using a panel data approach and fixed-effects regression; (iv) to examine farmers’ knowledge, perceptions, and constraints related to PA through thematic analysis of in-depth interviews; and (v) to propose a scalable adoption pathway aligned with Kenya’s agricultural development policies and farmer needs. A mixed-methods design integrates quantitative and qualitative data to provide a comprehensive understanding. The population comprises smallholder maize farmers (<5 hectares) across the three counties, with a target sample of 900 farm households for the quantitative survey and 40 in-depth interviews for qualitative insights. Stratified random sampling ensures representation by agro-ecological zones and seed/inputs access levels. Data collection employs structured questionnaires covering demographic attributes, farm characteristics, technology exposure, soil and yield data, input costs, and precise PA practices (e.g., GPS-guided variable-rate fertilization, drone-assisted crop monitoring, sensor-based irrigation scheduling). Secondary data are drawn from county agricultural offices and community extension records. Validity and reliability are ensured through pretesting, Cronbach’s alpha checks for multi-item scales, and triangulation with extension records and field visits. Analytical methods include descriptive statistics to profile PA adoption, a multinomial logit model to identify determinants of adoption intensity (non-adopter, partial adopter, full adopter), and fixed-effects regression to assess impacts on maize yield, input-use efficiency (ratio of output to fertilizer and water inputs), and household income over two cropping seasons. The qualitative component employs thematic analysis of semi-structured interviews to illuminate knowledge, attitudes, perceived benefits and barriers, and institutional factors, guided by the Technology Acceptance Model and the Diffusion of Innovations framework. A convergent mixed-methods approach synthesizes quantitative and qualitative findings to draw integrated conclusions. Expected findings anticipate a spectrum of PA adoption, with full adoption concentrated among relatively larger and better-connected farms and among those with higher education levels, stronger extension contact, greater access to credit, and proximity to input suppliers. PA is expected to be associated with modest yet statistically significant increases in maize yield (5–15%), improved input-use efficiency (reduction in fertilizer use by 10–20% without yield penalties), and higher net household income, though benefits may be uneven due to initial capital costs, reliability of service providers, and infrastructure constraints (electricity, mobile connectivity). Barriers likely to emerge include limited awareness, perceived risk, high upfront costs, data management challenges, and inadequate local service provision. The study contributes to knowledge by empirically validating the socio-economic and institutional determinants of PA adoption in sub-Saharan smallholder contexts, quantifying its agronomic and financial impacts, and providing a policy-oriented adoption framework. It foregrounds the role of extension systems, micro-finance, and adaptive governance in scaling PA, while offering an empirically grounded model for predicting adoption likelihood and expected returns under Kenyan conditions. Policy and practice recommendations include targeted credit and subsidy schemes for PA inputs, capacity-building and peer-learning platforms, public–private partnerships to strengthen PA service ecosystems, development of county-level PA benchmarks, and the integration of PA metrics into agricultural performance monitoring. The main conclusion posits that precision agriculture can enhance productivity and efficiency for Kenyan smallholder maize farms when supported by accessible financing, robust extension, reliable digital infrastructure, and context-specific adaptation of PA technologies; the study recommends staged, farmer-centered rollouts complemented by continuous monitoring and iterative learning to ensure sustainable adoption and equitable benefits.