Digital transformation in smallholder agriculture: A systematic review of ai adoption and constraints in the Kenyan potato value chain

Introduction

The Irish potato value chain significantly contributes to global food and nutrition security as the world's fourth food crop [1] . Globally, it produces over 390 million tonnes annually [2] , with hundreds of millions of people deriving their livelihoods from it. In Kenya, the sector faces low farm-level productivity, which has been on a decline that threatens its sustainability [3] . With the global population expected to reach 10 billion by 2050 [4] , and sub-Saharan Africa's population projected to double, food demand will surge. Currently, one in five people is undernourished globally [5] , making enhanced agricultural productivity vital for boosting global food security. This urgent demand requires adopting modern, advanced agricultural technologies like Artificial Intelligence (AI) to ensure the potato sector's productivity and sustainability [5,6] . AI optimizes farming practices such as precision farming, automated machinery, and real-time crop monitoring to improve crop yields, minimize environmental impact, optimize resource use, and reduce labor waste and costs [4,7,8] . AI-powered precision farming has been identified as a key driver for revolutionizing crop yield and climate resilience in East and South Africa [5] . However, high implementation costs, technical expertise gaps, accessibility issues, and ethical data privacy concerns hinder its wide adoption [9][10][11] . Despite these challenges, AI envisions a future of fully autonomous farms and data-driven farming techniques. This study aims to explore the potential of AI in transforming the Kenyan potato sector by synthesizing existing literature applicable to enhancing its productivity and sustainability.

Literature Review 2.1 Current Status of Potato Production in Kenya

The Irish potato (Solanum tuberosum L.) is a critical agricultural commodity in Kenya, ranking as the second most important staple food crop after maize [12,13] . Recognizing its strategic value, the Kenyan government has designated the potato as a priority crop within key policy frameworks, including the Agricultural Sector Transformation and Growth Strategy (ASTGS), Kenya Vision 2030, and the Bottom-Up Economic Transformation Agenda (BETA) [14] . The sector serves as a vital pillar for rural livelihoods and macroeconomic growth. In 2023, national production reached 2.31 million tonnes, generating an annual market value of approximately KSh 66 billion. The sector employs over 3.5 million people across the value chain [12] . Approximately 800,000 farmers cultivate potatoes across an estimated 120,000 to 150,000 hectares [12,13] . The production landscape is heavily dominated by smallholder farmers, who account for roughly 90% of the total output and operate on land parcels smaller than 0.4 hectares [15] . Cultivation is predominantly rain-fed and concentrated in high-altitude regions ranging from 1,200 to 3,000 meters above sea level, with the leading producing zones situated in Nyandarua, Meru, Nakuru, Narok, Bomet, Bungoma, West Pokot, and Elgeyo Marakwet Counties [16] . Potato production has seen growth, potentially due to government implementation of new regulations like the 2019 Crops (Irish Potato) Regulations and promotion of good agricultural practices (GAPs) and use of certified seeds [12] . Potato demand has also increased particularly in urban areas, driven by factors like rapid population growth, urbanization and increased fast-food industry [17] . However, despite the efforts made, a significant challenge of low yields persists [5,14,18] . Production status does not meet the increasing demand [19] . This threatens the livelihood of farmers and their communities, making it difficult to achieve food security and sustainability [5] .

Constraints and Structural Challenges in the Potato Value Chain

The central crisis facing the Kenyan potato sub-sector is the profound disparity between actual and potential farm-level yields. National average yields stagnate between 8 to 10 t/ha, significantly underperforming against a genetic and environmental potential of 25 to 40 t/ha and a global average benchmark of approximately 21 t/ha [1] . This yield depression is driven by a complex matrix of agronomic, institutional, and market failures that includes:

1. Seed System Deficits: Limited access to high-quality, certified seed potatoes remains a foundational bottleneck. Less than 5% of smallholders utilize certified seeds, with the vast majority relying on low-yielding, degenerated, farmsaved seeds or purchases from unregulated local markets, leading to the spread of diseases, low yields, and subsequent low farm incomes [20] . A stark structural deficit exists between the national demand for certified seed (estimated at 30,000 tonnes) and the formal supply capacity, which stands at roughly 7,000 tonnes due to output constraints at research institutions like the Kenya Agricultural and Livestock Research Organization (KALRO).

Phytosanitary Threats:

The pervasive use of uncertified seed stock facilitates the recycling and proliferation of devastating pests and diseases. Bacterial wilt (Ralstonia solanacearum), late blight (Phytophthora infestans), and the emerging threat of potato cyst nematodes (PCN) induce severe, compounding yield losses, made worse by the absence of effective curative chemical controls for bacterial infections.

Climatic and Agronomic Pressures:

Reliance on rain-fed systems exposes production to climate change realities, characterized by unpredictable rainfall patterns, prolonged droughts, and thermal stress. Furthermore, smallholder agronomic practices are hampered by declining soil fertility notably widespread phosphorus deficiencies. Soil degradation is aggravated by continuous monocropping, inadequate crop rotation, and historical reliance on blanket fertilizer recommendations that ignore farm-specific soil chemistry [19] .

Socioeconomic and Market Inefficiencies:

High operational costs for essential inputs limit their optimal application by capital-constrained smallholders [12] . This is compounded by weak research-extension linkages and poor rural infrastructure, which limit access to credit and formal extension training. Post-harvest management is equally weak; the lack of on-farm cold storage infrastructure triggers seasonal supply gluts and drastic price fluctuations. Consequently, marketing channels are dominated by unregulated brokers who exploit smallholders through nonstandardized packaging units and artificially depressed farmgate prices.

Sectoral Opportunities and Strategic Interventions

Despite these systemic vulnerabilities, the Kenyan potato value chain presents significant opportunities for optimization and structural transformation. Current public and private sector interventions are increasingly focused on formalizing seed production networks and scaling up the adoption of GAPs to enhance potato sector productivity and sustainability [12] . A major institutional driver is the government-led National Agricultural Value Chain Development Project (NAVCDP), which seeks to systematically improve smallholder access to high-yielding, disease-resistant seed varieties [14] . Public institutions strongly focus on the development of innovations and promotion of GAPs to enhance their adoption to improve yields, food security, and livelihoods among smallholder farmers [12,19] . Furthermore, the integration of agro-processing represents an emerging frontier for sectoral transformation, shifting smallholder focus from raw commodity marketing to high-value, shelfstable derivatives. Commercializing potato processing mitigates seasonal gluts, extends shelf life, and provides entry points into lucrative urban markets, thereby directly boosting smallholder profit margins [21] . To sustain this transition, institutional frameworks are pivoting toward improving business-to-business linkages among seed producers, smallholders, and industrial processors, alongside the agronomic development of dualpurpose potato varieties [22] . Concurrently, expanding minor irrigation infrastructure has been identified as a critical pathway to decouple potato production from absolute rainfed reliance, thereby mitigating the impacts of seasonal rainfall variability and stabilizing annual outputs [16] . Complementary structural reforms focus on strengthening formal farmer associations and cooperatives to enhance smallholder collective bargaining power, improve symmetric market information dissemination, and finance decentralized modern storage facilities to reduce post-harvest losses.

The Trajectory of AI in Agricultural Transformation

Amid these ongoing structural interventions, modern advanced technologies, specifically Artificial Intelligence (AI), offer a highly transformative paradigm to address the persistent yield gap. AI-driven systems enhance agricultural productivity through data-driven precision farming, automated pest and phytosanitary disease detection, and optimized digital market linkages, while simultaneously promoting environmental sustainability through precise input application [8] . By leveraging real-time data streams, AI empowers smallholders to make highly optimized agronomic decisions regarding crop management, localized weather forecasting, and micro-resource allocation [4,8,23] . In the context of the Kenyan Irish potato sub-sector, AI architectures possess the specific potential to close the historic yield gap by transitioning smallholders from lowinput, rain-fed methods to precision-based commercial production. Consequently, identifying and synthesizing the operational entry points for AI deployment becomes critical to reinforcing current state interventions and overcoming the sector's systemic ecological and structural challenges.

Methodology

To systematically evaluate the transformative potential of AI within the Kenyan Irish potato value chain, this study adopted a systematic review methodology based on secondary data synthesis. The investigation was executed through a structured multi-stage protocol involving document retrieval, strict eligibility screening, and qualitative thematic analysis to ensure analytical depth.

Data Sourcing and Search Strategy

A comprehensive, systematic document retrieval process was conducted across prominent electronic databases, including Google Scholar, Scopus, Web of Science, and institutional repositories. The search matrix utilized specific keyword combinations and Boolean operators, including: ("Artificial Intelligence" OR "AI") AND ("precision agriculture" OR "smart farming") AND ("potato production" OR "value chain") AND ("Kenya" OR "East Africa") intersecting with topics of productivity and sustainability.

Inclusion and Exclusion Criteria

To maintain academic validity, strict boundary criteria were established for the literature audit: Inclusion Criteria: (i) Peer-reviewed academic journal articles; (ii) published institutional reports from recognized agricultural bodies (e.g., CIP, KALRO, FAO); (iii) industry publications containing quantitative metrics on potato yields, AI applications, or value chain dynamics; and (iv) literature published within the last decade to capture contemporary technological milestones.

Exclusion Criteria: (i) Non-peer-reviewed blog posts, opinion pieces, or unverified commercial whitepapers; and (ii) generic AI studies without clear scalability or relevance to sub-Saharan smallholder agricultural contexts.

Screening and Final Sample Size

The initial search query yielded a broad baseline of literature. Following duplicate removal, titles and abstracts were independently screened against the eligibility criteria. Fulltext assessments were subsequently performed on the remaining shortlist. This filtering process culminated in a final analytical sample of 28 highly relevant, peer-reviewed articles and institutional records that detailed the intersections of AI engineering and potato value chain optimization.

Data Analysis and Thematic Categorization

The retrieved data underwent thematic analysis to map out the explicit mechanisms through which AI impacts the potato industry. Based on the empirical literature, the findings were synthesized and categorized into five distinct thematic domains:

Results and Discussion

Algorithmic Frameworks for AI in Potato Agronomy

The synthesis of the reviewed literature indicates that the integration of AI into the Kenyan Irish potato value chain offers a data-driven paradigm shift from traditional, lowinput husbandry to precision-driven commercial agronomy [5] . Mechanistically, AI-powered predictive architectures leverage advanced data analytics and machine learning (ML) algorithms to synthesize complex, multi-dimensional datasets. By integrating phenotypic, genomic, edaphic, and localized meteorological data streams, these systems generate real-time, actionable insights that optimize crop management, reduce production risks, and augment climate change adaptation [4,23,24] . Empirical evidence indicates that deploying integrated AI frameworks, combining IoT soil sensors, unmanned aerial vehicles (UAVs/drones), and satellite remote sensing, can catalyze potential yield improvements of up to 30% within smallholder farming systems [5] . This optimization is achieved through maximizing input-use efficiency to lower production costs and mitigating phytosanitary risks to close the gap between current smallholder outputs (8-10 t/ha) and the true potential of 30-40 t/ha.

Early Phytosanitary Disease Detection and Management

Bacterial wilt, late blight, and potato cyst nematodes represent catastrophic threats to Kenyan potato fields, often compounded by delayed manual diagnosis. The literature demonstrates that AI-driven computer vision tools, particularly Deep Learning (DL) convolutional neural networks (CNNs), serve as highly effective early warning systems by detecting morphological abnormalities in foliage long before symptoms become visible to the human eye [25] . By processing high-resolution images captured by mobile devices or drone-mounted optical sensors, these models rapidly classify and diagnose phytosanitary stressors [26] . A prominent operational deployment highlighted in the literature is the Plant Village Nuru mobile application. This platform utilizes on-device machine learning models to analyze leaf imagery, empowering smallholders with instantaneous, offline diagnostic capabilities for late blight and pest infestations [8,23] . Once a pathogen is identified, AI minimizes environmental and financial costs through Variable-Rate Technology (VRT) and automated UAV micro-spraying systems. Rather than resorting to conventional, uniform blanket pesticide applications across an entire field, AI-guided drones precisely target individual infected plants or micro-zones with calibrated chemical dosages. This targeted intervention minimizes crop losses, mitigates localized chemical resistance, and reduces overall environmental pesticide loading, creating a more sustainable production ecosystem [25] .

Precision Farming and Smart Resource Management

Traditional smallholder potato production in Kenya suffers from a lack of site-specific management, leading to structural resource misallocation. AI-driven precision farming resolves this by leveraging predictive analytics to establish sitespecific input schedules based on historical and real-time environmental variables [4,7] .

Edaphic Health and Dynamic Fertilization: Soil degradation and phosphorus deficiencies are major drivers of yield depression in Kenyan potato-growing counties [19] . Machine learning algorithms resolve this by analyzing complex edaphic properties from localized soil sensors to map nutrient variations across a plot. Instead of relying on rigid national blanket fertilizer recommendations, AI systems generate site-specific, adaptive fertilizer rates that align precisely with local soil fertility profiles and historical crop performance [4,27] .

Automated Irrigation and Moisture Optimization:

To decouple potato production from its absolute reliance on erratic, rain-fed systems, AI architectures optimize minor irrigation networks. Automated smart irrigation systems continuously monitor real-time soil moisture levels, evapotranspiration rates, and localized meteorological forecasts. By processing these datasets, predictive models dynamically schedule water application, delivering the exact volume required at critical crop growth stages [4] . This precise resource control prevents water stress, conserves limited rural water reserves, and enhances the climate resilience of the production cycle.

Yield Forecasting and Harvesting Optimization

Accurate crop yield forecasting is a fundamental pillar of modern agricultural planning, permitting smallholders and processors to optimize supply chain logistics and mitigate market volatility. The reviewed literature demonstrates that AI models can successfully predict Irish potato yields by deploying multi-layer data integration frameworks. These architectures analyze highly complex, non-linear relationships between variables, including edaphic parameters, environmental indices, and micro-climate data streams [4] . Beyond macroscopic yield forecasting, machine learning models continuously monitor crop phenological growth stages to mathematically predict the optimal physiological harvest window. Pinpointing this timeframe ensures maximum tuber quality and dry matter content while mitigating the risks of field-level degradation [4] . This highprecision forecasting capacity allows smallholders to shift from reactive harvesting to contract-aligned production, closing structural information gaps before the crop ever leaves the soil.

Automated Machinery and Labour Productivity

A major driver of high production costs in the Kenyan potato sub-sector is its reliance on intensive manual labour for repetitive agronomic tasks. Quantitative evidence highlights that AI-powered automation can significantly streamline field-level operations, augmenting labor productivity and lowering overall operational expenditures [28] . Autonomous, AI-driven tractors and harvesters equipped with computer vision systems execute labor-intensive tasks such as precision seed placement, robotic mechanical weeding, and mechanical harvesting with high spatial precision [4,7] . By replacing manual field operations with autonomous systems, farms reduce soil compaction, minimize mechanical tuber damage during harvest, and substantially compress the time required to complete seasonal field tasks. This operational efficiency drastically lowers variable production costs, enhancing net smallholder returns.

Climate Change Adaptation and Environmental Sustainability

The absolute reliance of Kenyan potato agronomy on predictable rain-fed cycles makes the sub-sector vulnerable to climate variability. AI-powered early warning networks address this vulnerability by processing localized and macroscale meteorological data streams to deliver real-time, hyperlocal insights on shifting rainfall patterns, thermal anomalies, and extreme weather events. This gives farmers the foresight required to execute proactive, climate-smart adjustments [23] . Furthermore, AI models act as decision-support engines that facilitate the implementation of climate-smart agricultural (CSA) frameworks. By calculating dynamic soil-water budgets and mapping local biodiversity indicators, these systems optimize multi-cropping strategies, resource conservation practices, and crop diversification schedules. At the ecosystem level, AI-driven precision input applications significantly reduce the sub-sector's aggregate carbon footprint. Optimizing chemical fertilizer application rates minimizes nitrous oxide emissions, while efficient machinery routing reduces fossil fuel consumption, driving the sector toward a more sustainable, low-emissions future.

Post-Harvest Loss Mitigation and Supply Chain Integration

Post-harvest losses caused by poor rural storage infrastructure and asymmetric market structures severely undermine smallholder profitability in Kenya. AI interventions address these post-harvest bottlenecks through a two-pronged approach: optimizing storage environments and formalizing market access channels.

Smart Storage Optimization: Potato tubers are highly perishable and prone to rapid spoilage from pathogens or improper environmental conditions. AI-enabled IoT monitoring systems deployed in rural storage facilities track internal temperature, relative humidity, and gas concentrations. Predictive maintenance algorithms analyze these telemetry streams to forecast facility malfunctions or identify early signs of localized spoilage, prompting timely interventions before catastrophic crop loss occurs.

Algorithmic Supply Chain Linkages: AI-powered digital platforms optimize logistics by directly connecting decentralized smallholder cooperatives with urban processing hubs and formal markets. These platforms use predictive analytics to forecast upcoming market price fluctuations and consumer demand spikes. By matching realtime supply capacities with commercial demand, AI platforms bypass middlemen, stabilizing market prices and maximizing smallholder margins.

Challenges and Institutional Barriers to Widespread AI Adoption

Despite the clear benefits of digital agriculture, the quantitative synthesis identifies a series of acute systemic, infrastructural, and regulatory challenges that currently inhibit the widespread adoption of AI within the Kenyan potato value chain.

The Rural Digital Divide and Structural Infrastructure Deficits

The operational architecture of AI, particularly cloud-based machine learning models and real-time IoT sensors, depends on stable infrastructure. However, Kenyan potato production is concentrated in remote, high-altitude rural regions characterized by severe deficits in last-mile electricity grid connectivity and reliable mobile internet coverage. Without continuous power and cellular data networks, real-time AI tools cannot function reliably. Furthermore, the substantial initial capital investment required to purchase drones, smart sensors, and automated machinery presents an insurmountable financial barrier for capital-constrained smallholders operating on less than 0.4 hectares. Bridging this digital divide requires deliberate public-private partnership investments to deploy low-cost, off-grid solar energy systems and expand rural broadband networks.

Human Capital Deficits and the Digital Literacy Gap

Deploying advanced agricultural technologies requires a baseline of specialized technical skills to operate equipment, interpret data insights, and maintain hardware. Currently, a pronounced digital literacy gap exists among Kenyan smallholders, many of whom are unfamiliar with the operational principles of machine learning interfaces or algorithmic analytics. To overcome this human capital bottleneck, academic institutions, extension agencies, and technology developers must design user-friendly, culturally accessible AI interfaces. Translating complex analytics into simplified, localized smartphone applications is critical to building user trust and ensuring widespread adoption.

Data Availability, Scarcity, and Quality Integration Challenges

The predictive accuracy of any AI architecture is fundamentally dependent on the availability of large, diverse, and high-quality datasets. In Kenya, there is an acute scarcity of comprehensive, open-access agronomic data specific to Irish potato production. Developing robust AI models requires compiling and integrating diverse data assets, including high-resolution leaf imagery, phenotypic records, genomic profiles, historical edaphic maps, and localized weather data. Currently, these datasets are highly fragmented, outdated, or completely non-existent within smallholder contexts. Accelerating AI research requires substantial institutional investment to establish centralized, standardized agricultural data repositories.

Underfunded Research and Sectoral Imbalances

A review of contemporary AI deployments in Kenya reveals a stark sectoral imbalance: only an estimated 49% of active AI initiatives focus on the broader agricultural sector, with the vast majority directed toward fintech, health, and urban logistics. Within the agricultural allocation, specific research dedicated to Irish potato is critically underfunded compared to dominant export cash crops like coffee or tea. This research deficit limits the development of localized algorithms tailored specifically to the unique biological and ecological profiles of Kenyan potato varieties.

Regulatory and Policy Gaps

Finally, the deployment of AI in Kenyan agriculture operates within a regulatory vacuum. There is a lack of clear legal and policy frameworks addressing data governance, algorithmic accountability, and intellectual property rights within the agricultural value chain. Establishing robust national frameworks is essential to ensure equitable data ownership, protect smallholders from algorithmic bias, and establish clear liabilities for autonomous system failures.

Current Applications and Empirical Field Deployments of AI in Kenya

The practical integration of AI within the Kenyan agricultural sector is no longer purely conceptual. Several localized, empirical field initiatives provide substantial baseline data demonstrating how machine learning interfaces scale smallholder Irish potato production.

• Computer Vision Diagnostics (Plant Village Nuru and Plantix): These mobile platforms deploy on-device convolutional neural networks (CNNs) to evaluate foliage anomalies via standard smartphone cameras.

Active across primary potato-producing zones such as Nyandarua, Nakuru, and Elgeyo Marakwet, these applications provide offline diagnostic readouts for late blight and pest infestations alongside real-time treatment protocols [23,27] . Notably, Plant Village Nuru has been adopted by over 10,000 Kenyan smallholders, driving documented localized yield expansions of up to 40% through rapid, targeted phytosanitary interventions.

• IoT Edaphic Telemetry (Ujuzi Kilimo Soil Pal):

Operating via portable electronic soil sensors, this platform delivers real-time algorithmic diagnostics of macronutrient profiles, moisture retention, and pH levels. Currently serving approximately 30,000 smallholder farmers, this data-driven edaphic stream replaces conventional farming methods with site-specific fertilizer blending, yielding productivity improvements exceeding 40% [23] .

• UAV Remote Sensing (The Third Eye Project):

Deployed in partnership with Future Water, this framework uses low-altitude drone sensors equipped with optical and near-infrared cameras to monitor normalized difference vegetation index (NDVI) mapping and crop water stress indicator variables. Deployed across 2,044 smallholders, this predictive tracking architecture has augmented micro-resource allocation efficiency, resulting in a documented yield expansion of up to 20% [23] . [23] .

Multifaceted Socio-Economic and Environmental Impact Assessment

A synthesis of these empirical deployments indicates that the benefits of AI-powered potato farming span three primary performance dimensions:

Productivity Gains and Economic Livelihoods

By enabling early problem detection and precision crop management, predictive AI models fundamentally shift smallholder farms toward commercial optimization. These applications mitigate crop risks, reduce input losses, and consistently expand potato yields by an average baseline of 30% [5] . This directly translates into augmented household incomes, providing rural producers with enhanced financial security and elevated baseline livelihoods [23] .

Resource Efficiency and Climate-Smart Sustainability

Precision farming techniques, guided by predictive algorithms, optimize the execution of micro-resource application schedules for water, specialized fertilizers, and fungicides. Instead of applying resource-heavy blanket remedies, AI limits input interventions to exactly when and where they are required. This precise micro-allocation drastically lowers overall chemical and water waste, minimizes negative environmental externalities, and limits pesticide runoff, creating an ecologic baseline for climateresilient farming systems.

Value Chain Resilience and Macro Food Security

On an institutional scale, integrating real-time crop disease forecasting with data-driven market aggregator networks yields a more stable, resilient potato supply chain. Predicting pest outbreaks early preserves domestic yields, while predictive market metrics smooth seasonal gluts and price spikes. By modernizing production methods and insulating supply flows from climatic and infrastructural shocks, AI directly supports national and global food security goals [6] .

Conclusions and Future Research Directions 8.1 Conclusions

The findings of this review demonstrate that Artificial Intelligence possesses a profound, transformative capacity to elevate the productivity, efficiency, and sustainability of the Kenyan Irish potato sector. By transitioning traditional farming into an optimized, data-driven enterprise, AI platforms provide empirical pathways to bridge the historic gap between current smallholder outputs (8-10 t/ha) and the country's actual agricultural potential (25-40 t/ha). From automated computer-vision diagnostics to real-time soil nutrient telemetry and streamlined digital market connections, these technologies reliably expand yields, conserve resources, and boost smallholder income margins. However, the realization of an inclusive digital transformation requires overcoming deep systemic and institutional barriers. High upfront technology costs, severe last-mile electrical and network deficits in rural producing counties, data privacy concerns, and low digital literacy rates among smallholders present critical bottlenecks. To prevent the widening of a rural digital divide, stakeholders must pivot from isolated private field pilots toward cohesive national frameworks that combine targeted technological deployment with aggressive rural capacity-building strategies.

Future Research Directions

To fully capitalize on the benefits of digital agronomy, the following target areas are proposed for future empirical research:

• Localized Pilot Testing and Algorithmic Calibration: Executing rigorously monitored, localized field pilots to calibrate international computer vision models specifically against indigenous Kenyan potato varieties and unique micro-climatic stressors.

Socio-Economic Impact and Equity Assessments:

Conducting longitudinal, empirical socio-economic evaluations to map out how digital tool adoption distributedly impacts gender parity, youth engagement, and wealth allocation among smallholders.

• Low-Data Machine Learning Architectures:

Investigating the design of lightweight, edge-computing AI systems capable of executing predictive diagnostics completely offline in network-deficient rural zones. • Cooperative Data Governance Modeling: Developing inclusive policy frameworks and community-owned data repositories to solve rural data scarcity while protecting individual farmer data ownership rights.

Declarations Funding

The author received no public, commercial, or non-profit grant funding to support the research, analysis, or writing of this manuscript. The Article Processing Charge (APC) will be entirely self-funded by the author.

References

1. Thuo CM, Wambugu SM. Strengthening smallholder farmers resiliency for improved sustainable productivity of Irish potatoes in Kenya. World J Adv Res Rev. 2024;21(1):1725-1734.
2. Kilimokwanza. The global Irish potato industry in 2026: Power, plates and the African pivot. Regional Agribusiness Report. 2026:12-18.
3. Atieno EO, Almekinders CJM, Struik PC. Developments and challenges of the Kenyan potato sector: Implications for smallholder farmers. Potato Res. 2026;69(98):1-24.
4. Doris L. AI in agriculture: Precision farming, crop monitoring, and sustainable food systems. ResearchGate Report Series. 2025:45-60.
5. International Potato Center (CIP). AI-powered potato precision farming: Revolutionizing crop yield and climate resilience in Africa. Lima: CIP Technical Broadcast Series; 2024.
6. Kenyaai.ke. Application of AI in agriculture in Kenya. National AI Portal Insights. 2024.
7. Al Kamal. Application of AI in agriculture: Boosting productivity and sustainability. ALK Fertilizers. 2025.
8. Piekutowska M. Using artificial intelligence for sustainable crop production, a comprehensive review with a focus on potato production. J Water Land Dev. 2025;64(I-III):181-193.
9. Manoj Kumar. The future of AI in agriculture: Benefits and challenges. Aeologic Tech Briefings. 2025:8-11.
10. Odume BW. Artificial intelligence in agriculture: Application in developing countries. SSRN Electron J. 2024.
11. Younas M, Akhtar MN, Batool S, Owais M, Sahar SS, Anum W. The integration of artificial intelligence in agriculture: Emerging trends, benefits and challenges. J Asian Dev Stud. 2025;14(1):105-118.
12. Mwaura A, Mbogo M. Unlocking potato value chain for employment opportunities in Kenya. Nairobi: Kenya Institute for Public Policy Research and Analysis (KIPPRA); 2025. Policy Paper No. 42.
13. National Potato Council of Kenya (NPCK). Mechanization in potato production in Kenya. NPCK Technical Bulletins. 2025.
14. Agroberichten Buitenland. Advancing Kenya's potato sector through innovation, improvements, and impact. Netherlands Ministry of Agriculture, Nature and Food Quality; 2025.
15. Soethoudt H, Castelein B. Strategies for potato smallholders in Kenya: A positive business case with reduced greenhouse gas emission. Agronomy. 2021;11(9):1857.
16. National Irrigation Authority. Exploiting irrigation to improve Kenya's potato farming. Ministry of Water, Sanitation and Irrigation Insights. 2022.
17. National Potato Council of Kenya (NPCK). Innovations and solutions in potato production in Kenya. NPCK Industry Reports. 2024.
18. Kihoro D, Gathungu GK, Mwirigi RN, Wairimu VN. Effects of adoption of agricultural technologies on Irish potatoes yield in Ol Kalou Sub County, Kenya: Application of endogenous switching regression model. Turk J Agric Food Sci Technol. 2025;13(1):151-157.
19. Momanyi WC. Assessment of the effects of use of good agricultural practices on Irish potato production and marketing in Kenya [master's thesis]. Nairobi: University of Nairobi; 2021.
20. Kabochi FN, Ochola SO. Irish potato production and consumption amid climate change and variability among small scale farmers in Narok South Sub-County, Kenya. Int J Nat Resour Ecol Manag. 2025;10(3):55-169.
21. KTN News Kenya. Defying the odds in the Irish potato value chain. Nairobi: Standard Media Group Broadcast Archive; 2025.
22. Naziri D, Devaux A, Hareau G, Wauters P. Supply and demand of processed potato products in Kenya, Rwanda, and Uganda: Variety requirements of processing companies and implications for trait prioritization for breeding. Potato Res. 2024;68:1915-1939.
23. Okoth A, Njoki P. Enhancing sweet potato production through artificial intelligence for job creation in Kenya. Nairobi: Kenya Institute for Public Policy Research and Analysis (KIPPRA); 2025. Research Report No. 118.
24. Musau GE. Sustainable agriculture leveraging artificial intelligence systems in Kenya's agri-food supply chain. Agric Sci. 2024;7(2):89-104.
25. Kaur A, Randhawa GS, Abbas F, Ali M, Esau TJ, Farooque AA, et al. Artificial intelligence driven smart farming for accurate detection of potato disease: A systematic review. IEEE Access. 2024;12:193902193922.
26. Sharma SK, Raheman H. Development of an AI-based early disease detection algorithm integrated with realtime variable spray. Cogent Food Agric. 2025;11(1):210-229.
27. Elbehri A, Chestnov R, editors. Digital agriculture in action: Artificial intelligence for agriculture. Rome: Food and Agriculture Organization (FAO) and International Telecommunication Union (ITU); 2021.
28. Judijanto L. Exploring the potential of artificial intelligence and digital technologies in transforming the palm oil industry: A review. J Inf Technol Policy. 2025;3(1):1-13.