Murang'a University of Technology

The increasing global integration of photovoltaic (PV) systems, vital for sustainable development, presents significant operational challenges in rural distribution networks. These feeders, particularly those in Kenya with high resistance-to-reactance (R/X) ratios, are prone to reverse power flow (RPF) and voltage instability, with generic smart inverter controls often proving ineffective due to insufficient voltage rise. This paper addresses this critical gap by developing and validating an adaptive, coordinated smart inverter control algorithm. Employing both Volt-VAR for reactive power support and Active Power Curtailment (APC), the strategy dynamically aligns PV output with real-time local load demand using quasi-static time-series simulations in OpenDSS for a representative Murang'a 11kV rural feeder in Kenya. This grid compliance was achieved at a quantified cost of approximately 1,500 kWh of curtailed renewable energy per day during the dry season. Critically, a comparative seasonal analysis was performed by repeating the simulations under rainy season (long rains) conditions, characterized by significantly reduced and variable solar irradiance. The rainy season results confirmed the absence of reverse power flow due to the naturally lower PV generation, which remained below local load demand throughout the day. This seasonal comparison validates that RPF is predominantly a dry-season phenomenon in this context and that the proposed adaptive control strategy is both critically necessary during high-irradiance periods and benignly non-intrusive during low-irradiance seasons. The study validates a practical methodology for high-penetration PV integration in challenging rural grids, emphasizing the necessity of feeder-specific, seasonally-informed control strategies to balance grid safety with renewable energy yield.

Neonatal mortality poses a major hurdle in global health efforts. As per the 2022 Kenya Demographic and Health Survey (KDHS), the neonatal mortality rate was 21 deaths per 1000 live births. Although the rates have been declining over the years, the rate remains significantly higher than the target set by Sustainable Development Goal 3.2 of 12 deaths per 1,000 live births by 2030 for preventable deaths endorsed by the United Nations meeting in 2015. To achieve this target, we need a predictive model that could provide actionable insights and timely interventions to predict and avert neonatal deaths. Machine learning has great potential to improve health outcomes for newborns, especially in areas with limited resources. The objective of this study was to develop a predictive model based on ensemble machine learning algorithms to predict neonatal mortality in Kenya using the 2022 KDHS dataset and identify the critical features contributing to the prediction of neonatal mortality. The Synthetic Minority Over-Sampling Technique was used to address the imbalance in the dataset. The Categorical gradient boosting model outperformed others in terms of accuracy (98.00%), precision (97.46%), recall (98.00%), F1 score (97.57%), and the Receiver Operating Characteristic (ROC) area under the curve (92.24%). The major factors identified included the number of children below five in the household, the total births recorded over the past five years, region, religion, sex of the newborn, source of drinking water, occupation, sex of the household head, use of mosquito nets, place of residence, ever-terminated pregnancy and duration of pregnancy. The research findings support the integration of machine learning into healthcare approaches, such Kenya Health Information System, to facilitate the accurate prediction of neonatal deaths and enhance survival for our newborns. The government should also promote family planning in the country to lower the number of children below five years in households and reduce the number of births recorded per woman in the last five years. The model's generalizability might be constrained by the exclusion of crucial clinical variables from missing data despite high performance.

Equitable access to government procurement opportunities (AGPO) is a strategic policy mechanism aimed at promoting inclusive participation and enhancing public service delivery. Despite nearly a decade of implementation, AGPO’s impact remains mixed with persistent challenges, including limited access to procurement information, low contract absorption rates, inadequate technical and financial capacity among beneficiary groups, and weak compliance by procuring entities. This study evaluated the effect of access to government procurement opportunities on the service delivery of water companies in Murang’a County. The investigation was guided by the Agency Theory, which emphasises the importance of effective delegation and accountability between procuring entities and suppliers in achieving optimal service outcomes. The study employed a descriptive survey design, targeting management staff across licensed water companies within Murang’a County. The population of this study consisted of 240 employees working in the procurement department and heads of the user department drawn from the five water companies in Murang’a County. Yamane’s formula was used to arrive at a sample size of 150 respondents. Data was gathered through questionnaires and analysed using regression analysis to determine the statistical relationship between access to procurement opportunities and service delivery outcomes. The regression results revealed that AGPO had a positive unstandardized coefficient of 0.472, with a corresponding p-value of 0.000, which is far below the conventional threshold of 0.05, indicating that the relationship between AGPO and service delivery is both strong and statistically significant. Hence, the null hypothesis was rejected, leading to the conclusion that AGPO implementation significantly improves service delivery in water companies. The study recommends that county governments and oversight bodies intensify AGPO policy implementation by increasing transparency, minimising access barriers, and equipping staff with procurement inclusion skills. As a policy recommendation, the Water Services Regulatory Board (WASREB) and the Public Procurement Regulatory Authority (PPRA) should jointly develop AGPO compliance guidelines tailored to water service providers to promote equity, efficiency, and improved public service outcomes.

Cellular structures are well-established in biological materials and are often mimicked in many kinds of structural designs applicable to engineering. This results from their lightweight designs and good mechanical properties. Cellular designs in nature have extremely complex configurations. As a result, the deformation behaviour models for bioinspired hollow parts based on these geometries, that are presently available in the literature, are limited in their capacity to provide detailed descriptions of the mechanisms resulting in deformation. Extensions to the existing deformation behaviour mechanisms of cellular parts are proposed in this paper. First, a general outlook on cellular designs is given. This is followed by a review of the commonly recognised two-stage stress–strain curve for cellular parts and its comparison with the new curve suggested in this paper, which incorporates suggestions more fully accounting for the deformation mechanisms of these structures. Further, analytical models that are available in the literature, outlining the behaviour of cellular parts, are highlighted, together with new models developed here for predicting failure of lattice structures based on the Tresca and von Mises criterion. Next follows a discussion of proposed strategies that could be adopted in deformation behaviour models for optimising the design of hollow structures to improve their mechanical properties. Finally, the anticipated challenges for and future insights into the incorporation of the cellular behaviour models suggested here, in cutting-edge structural design for additive manufacturing (AM), are highlighted.

Addressing rising food demand amid population growth and environmental challenges requires more youth engagement in agriculture. In many developing countries, this role is still dominated by ageing farmers. This study examined the career intentions of college-educated youth in Kenya using the Theory of Planned Behaviour and Entrepreneurial Cognition frameworks. Through a sequential mixed-methods approach involving 266 final-year university students, the findings revealed that over 80% expressed strong intentions to pursue agriculture-related careers. These intentions were largely influenced by attitudinal, normative, and perceived behavioural control factors from the TPB, alongside entrepreneurial cognition frameworks like opportunity recognition and problem-solving. Key determinants included family background—especially prior exposure to farming—and educational experience. Contrary to the prevailing narrative of youth disinterest, respondents displayed strong aspirations toward agriculture, with gender-based differences evident in perceived constraints such as land access and cultural norms. Despite the several challenges faced by the youth, such as limited access to capital, land constraints, labour intensity, and inadequate family support, the majority express strong intent to pursue agriculture as a career. These insights underscore the importance of designing interventions that are responsive to the distinct needs of youth, particularly along age and gender dimensions. The study recommends a multi-pronged policy approach: establishing a Youth Agrifinance Window, promoting exposure to agrifood innovations, creating Land Leasing Platforms, and integrating Agripreneurship Hubs in universities to support incubation, aimed at translating youth intentions into active agricultural engagement.

The modernization of power distribution grids, driven by the integration of DER, necessitates advanced modelling capabilities. A critical challenge is the semantic gap between the extensive geospatial asset data and the detailed electrical models required for engineering analysis. This data is often stored in static, unstructured KMZ formats that contain inherent topological errors. To address this, this paper presents a novel low-cost, Python framework that fully automates the conversion of this raw GIS data into solvable mathematical models for computation. This process generates executable files for industry-standard, script-based simulators such as MATPOWER and OpenDSS. The framework's core technical contributions include an XPath and Regex-based engine for metadata extraction. A graph-theory pipeline then utilizes a Minimum Spanning Tree (MST) algorithm to algorithmically heal topological disconnections. A Dijkstra-based method is then used for model abstraction. The core methodology was first validated using a bidirectional, reverse-engineering process on a 4-bus test case. This confirmed a lossless round-trip conversion to and from a data-rich KMZ format. Subsequently, it was applied to a complex, real-world 11 kV Kenyan distribution feeder. The generated model converged in a Newton-Raphson power flow. This demonstrated its utility as a powerful diagnostic instrument by enabling detailed feeder voltage profiling and loss analysis. Results were cross-validated against DIgSILENT PowerFactory, a graphic-based simulator, showing excellent numerical agreement. This study validates a scalable framework that transforms static, error-prone GIS data into dynamic, multi-platform diagnostic models. This approach provides a feasible pathway to accelerate grid modernization.

The general objective of this study was to establish the effect of digital transformation on the customer relationship management capability of star-rated hotels in Mombasa County, Kenya. The specific objectives were to examine the effect of digital infrastructure on customer relationship management capability, establish the effect of digital strategy on customer relationship management capability and analyse the effect of data analytics on customer relationship management capability of star-rated hotels in Mombasa County. The study was founded on the Technology-Organisation-Environment Framework, the Resource-Based Theory and the Dynamic Capabilities Theory. The target population was the star-rated hotels operating in Mombasa County, and they were 43 in number according to the Mombasa County Government Trade Department report (2024). A census of the target population was done because it was relatively small. A cross-sectional research design was adopted, and data were collected using a structured questionnaire with marketing managers of the hotels being the respondents. Out of the 43 star-rated hotels, marketing managers from 37 of them responded, and this led to an 86% response rate. Data analysis was done using simple regression analysis, and the regression results indicated that Digital infrastructure, Digital strategy and Digital analytics all had a positive and significant effect on customer relationship capability of the star-rated hotels. Digital strategy had the strongest effect on the hotel’s customer relationship management capability, which suggests that improvements in strategic digital planning and alignment are most strongly associated with enhanced CRM capabilities. Based on the results of the regression analysis, the study concluded that digital transformation had a positive and significant effect on the customer relationship management capability of the star-rated hotels in Mombasa County. The study recommended that management of the hotels should prioritise strategic digital planning, invest in scalable digital infrastructure, enhance the hotels’ data analytics capabilities through staff training programs and encourage cross-functional digital collaboration within the hotels. Future studies could consider incorporating qualitative methods to capture deeper organisational perspectives on digital transformation, including challenges, cultural factors, and leadership dynamics.

Post-harvest physiological deterioration (PPD) is a significant challenge in the storage of cassava roots. The shelf life of cassava roots ranges from 24 to 72 hours, which reduces their economic value. PPD is characterized by blue/black to brown vascular streaking and the release of reactive oxygen species. Breeding efforts aimed at improving PPD resistance are hindered by limited information on the genetic diversity of different shelf-life types and cassava's natural heterozygosity. This study aimed to identify cassava varieties with delayed PPD. Twenty varieties were assessed and evaluated changes in metabolite content and gene expression during storage. Cassava varieties were planted at Kenyatta University's Plant Transformation Laboratory. After maturity, After maturity visual scoring and imaged based scoring to assess PPD at 0, 3, 5, 8, and 11 days post-harvest. Spectrophotometry was utilized to measure the levels of phenolic compounds, flavonoids, carotenoids, and anthocyanins in the stored roots. Additionally, expression of PPD related genes was monitored PAL, APX-2, APX-3, and AP using RT-PCR in both tolerant and susceptible varieties. The study's results demonstrated significant differences in PPD responses among the cultivars. Notably, the varieties NASE18 and KCA0013 exhibited a delayed onset of PPD symptoms, whereas TZ130 and Tajirika were more susceptible. Furthermore, the phenolic and flavonoid content were higher in KCA0013 and NASE18, while the carotenoid levels reached their peak in MM96/7151. Interestingly, the anthocyanin content significantly increased in the susceptible variety TZ130. The PPD-tolerant varieties showed elevated levels of PAL, AP, and APX-3 expression, while the expression of APX-2 was higher in the PPD-susceptible varieties. This study identifies promising cassava genotypes for breeding by highlighting the distinct biochemical and molecular profiles associated with PPD tolerance.

With the fast growth of Internet of Things (IoT), there is an increasing security impact, e.g., IoT environments has become the advanced target of complex cyber-attack like Distributed Denial of Service (DDoS) attack. Most conventional Intrusion Detection Systems (IDSs) fail to scale with the heterogeneity, scale, and dynamism of IoT networks. To overcome these challenges, in this paper, we develop a smart hybrid intrusion detection framework that can effectively utilize the power of Machine Learning (ML) and Neural Network (NN) models, namely: cascades backpropagation neural network (CPBNN) and convolutional neural network (CNN) to improving accuracy and adaptability of detection in IoT environments. The proposed system employs a two-layer detection structure where the CPBNN model is used to detect abnormal patterns in the network packet level and identify the abnormal behaviors of packet sequences, whereas the CNN model performs deep feature extraction and classification to predict the abnormality as well as the nature of the abnormality. We develop this hybrid architecture to work well under IoT level large-scale deployments without incurring software overhead. The system has been evaluated on the KDDTest-21 benchmark dataset with the usual metrics, including accuracy, precision, recall and F1-score. Experiments results prove the CNN’s performance with 90% and CPBNN with 82% which valid the efficiency of the proposed method in detecting IoT related security threats. These results suggest the ability of intelligent hybrid IDS systems not only to defend in a pro-active manner using ML and neural networks in synergy, but also to use the adversary's strength against itself.

Proof of Stake (PoS) models are energy-efficient and require limited computational power. These features are critical in telemedicine environments, where resource-constrained devices must handle sensitive data securely. The growing need for auditable and privacy-preserving data storage in telemedicine underscores the importance of PoS models optimized for lightweight devices while complying with strict regulatory requirements, such as the Health Insurance Portability and Accountability Act (HIPAA).This study was guided by two research questions: (i) Which PoS models are lightweight and suitable for telemedicine? and (ii) What features make lightweight PoS models effective for privacy and efficiency in telemedicine? To address these questions, a systematic literature review (SLR) guided by the PICOC framework was conducted to investigate lightweight PoS models that can enhance privacy in telemedicine systems. Out of 2,394 papers studies screened, 55 were included in the analysis. The findings identified Algorand, Ouroboros Praos, Tendermint, Nxt, and Casper CBC as promising candidates. Key enabling features included lightweight voting mechanisms, such as Byzantine Agreement protocols and Verifiable Random Functions, as well as cryptographic techniques like symmetric encryption and multiparty computation. Performance metrics evaluated included latency, throughput, energy efficiency, and battery consumption, with Grey Relational Analysis ranking Algorand highest due to its low latency, high throughput, and minimal energy consumption.