East African Journal of Engineering

Effect of Cost of Construction of Medium-Level Building Projects on Project Productivity in Kirinyaga County, Kenya

2025-01-14T14:46:53+02:00

The construction industry is important in society, the economy and the environment. Despite its importance, it is faced with the challenge of the increasing world's urban population at a rate of 200,000 people per day who require affordable housing, social transportation and utility infrastructure. To counter this challenge, the industry will be under a moral obligation to transform by way of reducing construction costs, improving the use of scarce materials and making more eco-efficient over time. Therefore, this study's objective was to evaluate the effect of the cost of construction of medium-level building projects on project productivity in Kirinyaga County, Kenya. The total costs of construction (cost stack) are split into hard costs, soft costs and unforeseen costs. The methodology adopted for data collection was through the administration of structured questionnaires to select seven categories considered to be the main key stakeholders in the construction of building projects. These comprised landlords/homeowners, architects, structural engineers, quantity surveyors, building contractors, construction project managers, and Ministry of Housing and Infrastructure officials. To analyze the results, descriptive statistics and correlational analysis were used. In the analysis, emphasis was placed on the evaluation of the effect of the cost of construction on project completion time, quality and costs. Results from the study show that key factors affecting project productivity through increased costs are delay in payment to contractors and suppliers, high cost of construction materials and choice of construction materials. Other factors are the level of project complexity, site topography, architectural design errors and changes as well as structural design errors and changes.

Assessment of Energy Performance of Domestic Lights and Refrigerators in Sub-Sahara African Countries: A Case of Tanzania

2025-01-24T18:19:56+02:00

With the extensive potential for an increase in greenhouse gas emissions over the next few decades, Sub-Saharan African (SSA) countries need to focus on the energy performance of their domestic appliances. This study assesses the energy performance of domestic lights and refrigerators used in SSA countries, especially in Tanzania. The two key domestic appliances namely, lights and refrigerators were experimented. Four brands of 3, 5, 7, 9, 12, and 15 W white LED lights were tested at 24 C, while four brands of refrigerators with capacities of 205 L, 234 L, 184 L, and 225 L + 69 L were experimented. In light, the energy performance was assessed in terms of luminous efficacy () and energy efficiency (EEFLIGHTING). The findings revealed that the 3, 5, 7, 9, 12, and 15 W white LEDs could attain an average  of 86.1, 92.7, 102.9, 89.5, 86.2, and 96.9, respectively. Contrary, the EEFLIGHTING was found to be 123, 104, 115.3, 70.8, 72.0, and 86.0%. The analysis showed that there exist large deviations in  and EEFLIGHTING which cause the lights to consume more energy while producing low brightness. The 205 L and 234 L refrigerators consumed approximately 428 and 486 kWh/year; and 290 and 236 kWh/year at 24 and 27 C with opening activities. In the 184 L and 225 L + 69 L refrigerators, the annual energy consumption reached approximately 235 and 264 kWh/year; and 670 and 749 kWh/year at 24 and 27 C with opening activities. The opening activities in refrigerators increase energy use by its factor while the change in location temperature increases the daily and annual energy use (from 24 to 27 C) by approximately 12%. These findings are critical for understanding the energy use of residential appliances in relation to energy consumption per capita in SSA, as well as developing energy efficiency strategies to boost market adoption

Development of Maintenance Management System for Real-Time Monitoring of Power Transformer to Improve Availability Performance: The Case of Tagamenda TANESCO Grid Substation

2025-01-29T14:35:52+02:00

Power transformers are critical assets in TANESCO's grid substations, playing a vital role in ensuring a reliable and uninterrupted power supply. However, the growing challenges of ageing infrastructure, increasing energy demand, and reactive maintenance practices often lead to unplanned outages, higher operational costs, and reduced transformer availability. This dissertation focuses on the Development of a Maintenance Management System for Real-Time Monitoring of Power Transformers at TANESCO grid substations, with the goal of improving availability performance. The study explores the design and implementation of a Real-Time Monitoring Management System (RTMMS), leveraging advanced sensors, data analytics, and predictive maintenance techniques. The RTMMS continuously monitors critical parameters such as temperature, oil levels, dissolved gases, and load conditions, providing real-time insights into transformer health. By integrating predictive analytics, the system identifies potential faults early, enabling timely interventions and reducing downtime. Additionally, it supports data-driven maintenance planning, enhances operational reliability, and extends transformer lifespan. The proposed system addresses challenges such as high initial costs, data management complexity, and resistance to technological change through phased deployment, secure cloud solutions, and comprehensive training programs. The expected benefits include improved transformer availability, cost efficiency, enhanced grid sustainability, and a shift from reactive to proactive maintenance practices. This research underscores the transformative potential of real-time monitoring systems in modernizing maintenance management and aligns TANESCO with global best practices in energy sector innovation and operational excellence.

Application of Machine Learning in Estimating California Bearing Ratio from Soil Index Properties in Kenya

2025-01-29T15:26:30+02:00

The California Bearing Ratio (CBR) is an important civil and transportation engineering test. It is normally carried out to assess soil's bearing capacity and strength for road pavement and foundation construction. The test, however, is both time-consuming and labour-intensive, resulting in significant delays during the construction process, ultimately leading to financial losses due to the high cost typically associated with construction projects. As a potential solution to this issue, an investigation is conducted into the application of artificial intelligence (AI) and machine learning (ML) techniques for accurately forecasting CBR values. Three models were used in the study, namely, the random forest model, linear regression model, and extreme gradient boosting (XGBoost) model. These models were employed to forecast CBR values based on several soil index properties. These properties included particle size distribution (i.e., percentage of soil passing through the sieve of diameter 0.425mm and 0.075mm), liquid limit (LL), plasticity index (PI), maximum dry density (MDD), plastic limit (PL), and optimum moisture content (OMC). A dataset containing these soil properties and corresponding CBR values for soils was obtained from the University of Nairobi civil engineering laboratory. The models were then trained on 80% of the data and tested on 20%. Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Coefficient of determination (R²) were used to evaluate the accuracy of the predictions. The findings showed that XGBoost was the most accurate model with the lowest MAE, MSE, and RMSE, and the highest R², making it the preferred model for predicting CBR

Analysis of the Efficacy of Contract Management Practices in Public Projects in the Construction Industry of Lesotho

2025-02-02T15:50:53+02:00

Contract mismanagement in public construction projects is a common occurrence in the construction industry of Lesotho. It has led to significant project delays and disputes, hindering the country's infrastructure development. This paper is a report of a recent study carried out to assess the efficacy of contract management practices, aiming to identify areas for improvement for the purpose of enhancing project delivery in the industry. A mixed-methods approach, combining a documentary search and a survey - of the professionals actively involved in public construction projects in Lesotho - was employed for the research. It was observed that a formal framework for contract management is in place, but various challenges hinder the effective implementation of the framework. The challenges comprise inadequate managerial capacity of the contract managers, weak enforcement of contractual obligations, and lack of transparency and accountability in the contractual process. The researchers concluded that there was a pressing need to enhance contract management practices in Lesotho's public construction sector. Their recommendations for addressing the identified challenges are capacity-building initiatives for contract managers, strengthening the regulatory frameworks for contract enforcement, and promoting a culture of transparency and accountability within the industry

A Conformal Transformation Technique for Mapping an Open Channel Fluid Flow into a Two-Dimensional Plane

2025-02-10T18:23:59+02:00

Due to the emergence of many applications in areas of open-channel fluid flow, interest in this branch of fluid mechanics has increased considerably. These areas are wide-ranging, including electricity generation using tidal waves, control of floods and improvement of irrigation systems. Finance, thermal imaging, harnessing of melting of glaciers, flow of fluids over ramps and sluice gates, and quite notably, petroleum exploration are other areas where open channel flow mathematics find great relevance. Whether the free surface of the open channel is being predicted using known bottom characteristics (which is called the ‘direct approach’), or the bottom is being inferred from an observable surface (referred to as the ‘inverse approach’), researchers often find the need to make computations easier by transforming complex topologies into simpler, more relatable physical equivalents. In this paper, the channel flow of a gravity-influenced Newtonian fluid is treated as the physical plane. The fluid is flowing in the positive direction. The upper half-plane is conformally equivalent to the interior domain determined by any polygon, and the interior points of the physical plane are transformed to the corresponding points above the real axis of the upper half-plane which is then mapped onto the auxiliary half-plane, (the plane) by being treated as an infinite strip by use of the Schwarz-Christoffel theorem. Assumptions are made that the fluid has dimensional quantities such as uniform spee far upstream, and velocity potential . Far upstream before the arbitrary obstacle is encountered, the fluid has a uniform height, Then and are used to nondimensionalize the variables to enable computation in a completely non-dimensional environment. With the fluid assumed as steady, inviscid, irrotational and incompressible, w representing the physical w-plane, (and t) being points on the auxiliary the - plane, the angle made by a tangent to this plane at designated points, the required mapping is found to be .

Single-Prime versus Multi-Prime Contracting: Impact on Performance Metrics in Road Infrastructure Projects - A Kenyan Case Study

2025-03-04T13:04:21+02:00

This study explored the difference in performance metrics between single-prime contracting methods and multi-prime contracting methods in terms of Cost and overall performance (project’s financial, timeliness, overall quality, compliance with Safety and utilization of resources) as they are used in road infrastructure in Nairobi City County in Kenya. The study employed primary data collection using a Semi-structured Likert-Scaled questionnaire to collect data from professionals who have been involved in both contracting and methods of road construction in the County over the past 10 years. The study targeted a sample size of 385 out of which 267 respondents (69.4%) participated in the study. Purposive and snowball sampling methods were used to recruit and select the study participants. The Cronbach's alpha for contract cost overruns is 0.933, indicating the excellent reliability of the tool used to collect the data. With a Cronbach’s alpha of 0.615, overall contract performance demonstrated acceptable reliability. Inferential analysis of data collected revealed a statistically significant difference in construction costs between single-prime and multi-prime contracting methods (U = 798.00, z = 0.326, p= 0.017). The mean rank was higher for single-prime (42.71) compared to multi-prime (40.97), suggesting that single-prime projects tend to have higher costs. The analysis of construction costs revealed a statistically significant difference between single-prime (x̄=3.4286, σ=0.53385) and multi-prime (x̄=3.3735, σ=0.53785) contracting methods; t (81) = 0.461, p =0.006. The mean difference of 0.05504 (95% CI: -0.18277 to 0.29285) indicates that single-prime projects had slightly higher construction costs than multi-prime projects. The result suggests that single-prime contracting may actually be associated with marginally higher costs, possibly due to the prime contractor's markup on subcontractor work. This study concludes that multi-primes perform better than single-prime contracting methods in terms of cost and overall performance in large and complex road construction projects in Nairobi City County-Kenya Nairobi City County where a lot more technical expertise may be required.

Oil Well Performance Forecasting Using Decline Curve Analysis: Tharjiath Oil Field, South Sudan

2025-03-24T13:00:27+02:00

This study applied Decline Curve Analysis (DCA) to forecast the performance and reserves of oil wells in the Tharjiath Oil Field, located in South Sudan's Muglad Basin. Historical production data from five wells were analyzed to compare exponential, hyperbolic, and harmonic decline models. The exponential model was found to be the most accurate, as it closely matched observed production trends by assuming a constant percentage decrease, aligning well with the initial production behaviour in the Tharjiath Oil Field. To further validate the findings, Oil Field Manager (OFM) software was used to simulate future production trends based on the exponential model. This robust platform enabled comprehensive analysis and visualization of reservoir performance, confirming the effectiveness of DCA in predicting well performance and reserves and reinforcing the exponential model as the best representation of the reservoir's natural depletion. The findings underscored the necessity of refining forecasts through sensitivity analysis, which involved varying key input parameters such as initial production rates, decline rates, and reservoir pressure. This systematic alteration of variables allowed the study to identify a range of potential outcomes and assess the associated uncertainties in production predictions. Additionally, the study emphasized the benefits of integrating DCA with real-time production monitoring and reservoir simulation models to enhance forecast accuracy. The study also recommended further exploration of Enhanced Oil Recovery (EOR) techniques, including water flooding and gas injection, to optimize recovery and extend the economic life of the field. Some recommendations were made to expand DCA applications to all wells in the field, incorporate real-time data into the forecasting process, and address the economic and policy implications for sustainable reservoir management

Reducing Congestion in The Metropolitan Area: A Case of Greater Kampala Metropolitan Area

2025-04-01T10:44:04+02:00

Traffic congestion poses significant challenges for cities in developing countries, with Kampala, Uganda, being no exception. As the capital and largest city in Uganda, Kampala grapples with a burgeoning population and inadequate transport infrastructure, leading to gridlock, environmental pollution, and economic losses. This paper proposes the introduction of a congestion charge zone as a solution to alleviate traffic congestion in Kampala. Drawing inspiration from successful implementations in cities like London, the proposed system aims to discourage private vehicle use during peak hours by levying charges on entering designated zones within the city. The paper outlines the current state of transport in Kampala, highlighting the dominance of private vehicles and the shortcomings in public transit infrastructure. It explores the concept and mechanics of congestion charge zones, emphasizing the role of technology in enforcement and the potential for cost-effective implementation in Kampala, leveraging existing CCTV and traffic control infrastructure. Furthermore, the paper discusses key considerations for successful implementation, including the need for viable alternative transport options, public awareness campaigns, stakeholder engagement, and robust enforcement strategies. By implementing a congestion charge zone, Kampala stands to mitigate traffic congestion, improve air quality, and enhance overall urban mobility, contributing to sustainable development in the region

Design and Performance Evaluation of a LoRaWAN-Based Communication System for Enhanced Situational Awareness in Armored Vehicles

2025-04-14T15:13:25+02:00

This study presents the design and performance evaluation of a LoRaWAN-based communication system for enhancing situational awareness in armoured vehicles, with a comparative analysis against GSM networks. Performance metrics such as communication range, latency, security, GPS accuracy, data transmission speed, and power consumption were assessed under different environmental conditions, including open fields, urban areas, and forested regions. The results indicate that LoRaWAN offers a reliable alternative to GSM, particularly in environments with limited cellular infrastructure. LoRaWAN demonstrated a communication range of up to 12 km in open fields, moderate security with AES-128 encryption, and superior power efficiency, supporting up to 41.6 hours of continuous operation on a 5000mAh battery. While GSM outperformed LoRaWAN in latency (50–150 ms vs. 150–300 ms) and data transmission speed, LoRaWAN provided better performance in rural areas and secured communication through dynamic key management. These findings highlight LoRaWAN’s potential for military applications where secure, long-range, and energy-efficient communication is required

An Evaluation of Stakeholder Management Performance in County Government-Funded Construction Projects: A Case of Machakos County

2025-04-21T22:16:43+02:00

Stakeholder management is a pivotal aspect of construction projects funded by county governments in Kenya. Effective stakeholder management is essential for aligning project objectives with the diverse needs of all the stakeholders involved. In Kenya, public infrastructure projects often face challenges such as delays and budget overruns. Therefore, robust stakeholder management practices are essential for ensuring successful project implementation. The specific objectives of this study were to determine the current state of stakeholder management performance in County government-funded construction projects, to establish the existing stakeholder management practices in County Government-funded construction projects, and to establish the influence of stakeholder management practices on stakeholder management performance in County Government-funded construction projects. The study adopted a survey research design. The data was collected using questionnaires and measured using a 5-point Likert scale. Simple stratified sampling was used to identify the 254 respondents. The respondents included contractors, project consultants, end users, ward development officers, PMC representatives, ward administrators, and village administrators. The collected data was coded and entered into Statistical Packages for Social Scientists (SPSS) and analysed using descriptive statistics. The multiple regression analysis method was used to determine if a relationship existed between the dependent and independent variables. The overall level of stakeholder management performance had a mean of 2.25. This research established that stakeholder management practices are statistically significant in explaining the stakeholder management performance in construction projects funded by the County in Machakos County. Additionally, this study concluded that the level of stakeholder management performance was low in Machakos County. Finally, the study recommended that counties should prioritise stakeholder management training through well-organised awareness campaigns.

Heat Exchanger Network Synthesis Using Node-based Non-Structural Model With Enhanced Dynamics for Stream Matching

2025-04-25T22:19:46+02:00

Node-based nonstructural models (NNMs) for heat exchanger network (HEN) synthesis realize flexibilities for stream matching across the entire ranges of process streams, as the concept of stages is not used. These models can also accelerate the efficiency of optimization algorithms. However, since nodes are considered to be in fixed positions, heat exchangers tend to crowd in the feed regions of the process streams during the later stages of optimization. This crowding hinders the generation of new heat exchangers in those areas, eventually interrupting the randomness of the NNM and impeding structural optimization. This paper proposes a mechanism for adding split groups within existing node groups in process streams to allow for freer generation of new stream matches and reduce exchanger clustering. The random walk algorithm with compulsive evolution (RWCE) is utilized for HEN optimization. The algorithm is particularly suitable because it can evolve only existing heat exchangers while also generating new ones independently. Examples from the literature are solved to illustrate the applicability of the proposed modifications to NNM and the results compare well with solutions reported in the literature

Design and Construction of a Solar Powered Silver - Fish Dryer for Enhancing Food Security and Economic Livelihoods of Fishing Communities in East Africa

2025-05-05T15:47:48+02:00

The aim of this study was to design and construct a solar-powered silver-fish dryer that would help reduce the post-harvest losses experienced by fishing communities in East Africa. The system was developed with the goal of promoting food security, improving preservation methods, and enhancing the quality of dried silver-fish. The construction process involved welding, wiring, and soldering various materials, with the trays made from wood and metallic meshes measuring 37 cm by 33 cm. The fabricated cover was also made of wood and measured 47 cm by 63 cm. Three 100-watt bulbs were utilized as heaters, while a CM 3024Z charge controller was connected to a 20W solar panel to manage energy flow. The system was successfully designed and constructed as planned. The direct current (DC) supplied to the dryer was 12 volts. The temperature sensor operated at 5 volts, the heater at 120 volts, and the fan at 12 volts. The charge controller regulated the solar panel’s energy supply to the battery, and the inverter converted the direct current to alternating current (AC) to power both the fan and the heater, thereby minimizing energy losses. The dryer system ensures continuous drying, regardless of weather conditions, by efficiently harnessing solar energy to power the heating system. This results in improved cleanliness, reliability, and the ability to dry large quantities of silverfish effectively. The solar-powered silver-fish dryer designed and constructed in this study proves to be an effective solution for addressing post-harvest losses in East Africa's fishing communities. The system's integration of solar energy, combined with an efficient drying mechanism, ensures continuous operation regardless of weather conditions, thereby improving reliability and preserving the quality of the dried fish.

Forecasting Maximum Temperature Using Comparable Optimizers in LSTM Deep Learning Model: A Case of Koga Mango Farm, Mkuranga District, Tanzania

2025-05-06T23:13:21+02:00

Mango farming is an important economic activity in Tanzania, contributing to the economy through exports of mango fruits and products and acting as a primary source of income for many farmers. Maximum temperature is one of the critical weather variables affecting the growth of mango, having an impact both on flowering stages and fruits, so failure to correctly forecast extreme maximum temperature and take appropriate measures may pose challenges such as poor quality of mango fruits and hence low income to farmers. Long Short-Term Memory (LSTM) is one of the famous deep learning models used for forecasting time-series variables such as temperature. In the LSTM model, an optimizer is a very important component as it is used to minimize loss during model training. Despite there being a number of optimizers, which can be used in the LSTM model, there is still a research gap, on which one is the best-performing optimizer in forecasting tasks, especially in the context of forecasting maximum temperature in Koga farm, a mango farm located in Mkuranga district, Pwani region, Tanzania which has unique climatic conditions and has a small geographical area. This study aims to fill this gap by comparing the performances of common LSTM optimizers and developing an LSTM model for helping Koga farm officials forecast daily maximum temperature using the best-performing optimizer. The experimental findings reveal that Adam and Adamax are the two best-performing optimizers with both having Root Mean Squared (RMSE) values of 0.089 on the test set (unseen data). The performance of the remaining optimizers on the test set with their RMSE values in brackets are as follows; RMSprop (0.091), Adagrad (0.099), SGD (0.102) and Adadelta (0.107). This study recommends that software developers and researchers use either Adam or Adamax optimizer in LSTM models when forecasting temperature in environments which resemble that of the Koga farm in Tanzania.

Influence of Technology on Labour Productivity in the Construction Industry in Nairobi City County, Kenya

2025-05-11T20:32:46+02:00

The construction industry in Nairobi City County, Kenya, significantly contributes to economic growth, infrastructure development, and job creation. However, rapid urbanisation has exposed inefficiencies such as project delays, resource constraints, and low labour productivity. This study aimed to examine the influence of technology adoption, specifically safety technology, Building Information Modelling (BIM), construction automation, and remote monitoring on labour productivity in the construction sector. Guided by the Theory of Constraints, the study sought to identify how technology alleviates productivity limitations in construction processes. The specific objectives were to establish the influence of safety technology and assess the impact of BIM on labour productivity. The research targeted a population of 2,610 construction professionals in Nairobi, including engineers, architects, quantity surveyors, contractors, technicians, and artisans. A sample size of 261 was selected through simple random sampling. Primary data were collected using semi-structured questionnaires and interview guides. The questionnaires featured closed-ended and Likert-scale questions to gather quantitative data, while interviews provided qualitative insights. The study achieved an 88% response rate (230 responses). Data analysis was conducted using SPSS, employing correlation, regression, and ANOVA techniques. The results revealed a significant positive correlation (r = 0.386, p < 0.05) between technology use and labour productivity. Regression findings showed that technology use accounted for 14.9% of the variance in productivity levels. The study concludes that technology adoption enhances efficiency, safety, and project outcomes in Nairobi’s construction industry. It is recommended that construction firms invest in affordable technologies and capacity building. Policymakers should support technological innovation through training and incentives. Further research should explore the impact of artificial intelligence and digital tools in other regions of Kenya and Sub-Saharan Africa.

Optimising Harare’s Water Dissemination Network: Development of an IoT-Driven Leak Monitoring System for Sustainable Urban Water Management

2025-07-18T08:56:50+02:00

Water scarcity persists as a pressing issue in Harare, exacerbated by undetected leaks within the municipal water distribution networks, which significantly contribute to non-revenue water (NRW) losses. A considerable volume of water fails to reach end-users due to these inefficiencies, intensifying supply shortages and elevating operational expenditures for water utility providers. These challenges stem primarily from deteriorating infrastructure, reliance on conventional leak detection methodologies, and inadequate management systems. As noted by Dinar (2024), effective water resource management remains a critical concern, particularly in regions facing water deficits amid escalating demand for potable water. This research project proposes the development of an IoT-enabled real-time water leakage detection system to optimise monitoring capabilities and mitigate NRW losses. The system architecture incorporates a network of sensors, microcontrollers, and IoT communication frameworks to facilitate continuous, real-time data acquisition. Strategically positioned sensors will monitor key hydraulic parameters, including flow rate, pressure, and water levels, with collected data transmitted to a cloud-based analytical platform. A dedicated web interface will be implemented to deliver instantaneous leakage alerts, dynamic graphical representations, and comprehensive diagnostic reports, thereby enabling preemptive maintenance interventions and minimising water loss. The system's efficacy will be empirically validated through prototype testing, ensuring precise leak localisation and rapid response mechanisms. In this context, the proposed method will reduce the non-revenue water in Harare by at least 20% within a year of deployment, as well as reduce the response time to leaks by at least 50%. Ultimately, this initiative aims to advance water conservation efforts, improve operational efficiency, and promote sustainable water management practices that can be deployed to large consumers of fresh water, such as industrial complexes

Assessment of the Causes and Effects of Construction Project Delays in Urban Water Supply Project: A Case Study of Kibremengist and Shakkiso Towns

2025-08-05T19:49:14+02:00

Construction project delays are a widespread challenge, particularly in developing countries like Ethiopia, where they often result in significant time and cost overruns. This study investigates the causes and impacts of construction delays in the Kibremengist-Shakkiso Water Supply Project, a government-funded initiative supported by BADEA. Employing a qualitative, case-based approach, the research draws on project documentation, site observation, and professional insights. The delays were primarily attributed to client-related issues, contractor inefficiencies, consultant shortcomings, and external factors such as power supply and material availability. Key delayed factors included unrealistic time allocation, delayed procurement process, and incomplete design documentation. The study further examines the social and economic consequences of these delays and provides practical recommendations to minimise similar challenges in future water sector projects. The findings provide valuable insights for policymakers, engineers and project managers engaged in infrastructure development in similar countries

Development of a Maintenance Management System for Enhancing the Maintainability Performance of Masonry Commercial Building Structures: A Case of Bungu Ward - Kibiti District Council

2025-08-06T17:50:00+02:00

The maintainability performance of masonry commercial building structures in developing regions faces significant challenges due to inadequate maintenance management systems. This study developed a comprehensive maintenance management system to enhance the maintainability performance of masonry commercial buildings in Bungu Ward, Kibiti District Council, Tanzania. Using a mixed-methods research design, the study assessed 63 commercial buildings to identify critical factors affecting maintainability performance through Relative Importance Index (RII) analysis. A multiple regression model was developed and validated to predict maintainability performance based on seven key factors: structural movement, salt crystallisation, masonry cleaning viability, render/coating condition, foundation-wall interface, damp penetration, and load-bearing capacity. The regression model demonstrated strong predictive power with R² = 0.770, explaining 77% of the variance in maintainability performance (F = 27.922, p < 0.001). Validation testing across five buildings showed performance scores ranging from 82% (Very Good) to 16% (Very Severe), confirming the model's discriminatory capability. A digital Building Structure Maintenance Management System (BSMMS) was developed using Python, featuring modules for building inspection, team management, performance reporting, and predictive maintenance scheduling. The system successfully standardised maintenance protocols and enabled evidence-based decision-making for resource allocation. The study concludes that systematic, evidence-based maintenance management significantly improves building performance and sustainability. The developed system provides stakeholders with practical tools for proactive maintenance planning, ultimately extending building lifespan and reducing lifecycle costs. These findings contribute valuable insights for building maintenance management in similar developing contexts and establish a framework for sustainable commercial building management.

Wind Speed Prediction Using BiLSTM Deep Learning Model and Comparable Batch Sizes of Training Data: A Case of Singida Wind Farm Site, Tanzania

2025-08-06T17:50:00+02:00

Singida region, located in central Tanzania, has long been identified as a potential location for installing wind farms to generate electric power due to the steady annual wind speed. Apart from the huge potential of contributing to the national grid, wind power also helps to address carbon emissions and environmental problems associated with generating electric power using fossil fuels. Failure to accurately predict wind speed can lead to poor harvest of wind power and low contribution to the national grid, and in the end, affect consumers. Bidirectional Long Short-Term Memory (BiLSTM) is one of the Deep Learning models which can be used to predict time series parameters such as wind speed. In a BiLSTM model, a batch size is an important hyperparameter as it is used to set the number of training data samples to be processed together before the weights of a Deep Learning model are updated. Despite its importance, there is still a research gap on the impact of batch size on the prediction performance of BiLSTM models, especially in the context of predicting wind speed at the Singida Wind Farm Site, located in the Singida region, Tanzania. The goal of the study was to fill this gap by developing a BiLSTM model and comparing the performance of three batch sizes (16, 32 and 64) in predicting wind speed at the Singida Wind Farm Site. The 14-year Singida Wind Farm Site daily wind speed dataset was first pre-processed by scaling (normalizing) it using Standard scaler and then split into training, validation and test sets before used to train and test the developed BiLSTM model which used previous 5 days wind speed values as input to predict the output (next day (6^th day) wind speed). The trained BiLSTM model with the optimal (best performing) batch size was then saved in .h5 format and integrated with a Gradio-based web App to provide a user interface for officials in the Singida region to predict daily wind speed at the Singida Wind Farm Site. The evaluation findings revealed that batch size has an impact on the prediction performance of the developed BiLSTM model, showing that the lower the batch size, the better the prediction performance of the BiLSTM model. The findings also revealed that, 16 is the optimal (best performing) batch size with Mean Absolute Error (MAE) score of 0.58, Root Mean Squared Error (RMSE) score of 0.76 and R² score of 0.79, followed with a batch size of 32 (MAE score of 0.62, RMSE score of 0.79 and R² score of 0.75) and followed by a batch size of 64 (MAE score of 0.66, RMSE score of 0.81 and R² score of 0.72). This study recommends that Artificial Intelligence (AI) software developers and researchers use a batch size of 16 in BiLSTM models when forecasting wind speed at the Singida Wind Farm Site, as well as in environments and climates which resemble that of the Singida Wind Farm Site in Tanzania.