Journal of Advanced Research Design

Gigabit Passive Optical Networks: Undetected Signal Tapping

Khaled F. Abdallah — 2025-08-07

Recently, the customer high demands for increasing bandwidth and high data rates have necessitated the provision of innovative services such as video cameras and high-speed internet. Although Very High-speed Digital Subscriber Line (VDSL) technology addresses the demand for high bandwidth, it has significant drawbacks in terms of long-distance communications and high attenuation. To overcome the attenuation in cables over long distances and also to increase the bandwidth and data rate, the solution is the use of fibre optic networks. These networks are divided into passive and active networks. Among the different optical network choices, Gigabit Passive Optical Networks (GPONs) have acquired popularity due to their energy-efficiency because they use extremely small power to operate effectively. Furthermore, it consumes 95% less energy than copper networks and it offers a very high traffic speed of up to 100 Gbps for subscribers. The purpose of this study is to evaluate the performance of GPONs under varied transmission powers and frequencies while subjected to undetected signal tapping. The recommended approach includes eavesdropping on upstream and downstream networks without detection. This is accomplished by using an optical coupler with predetermined splitting ratios (10:90, 5:95, 2:98 and 1:99) and insertion losses (0.75 dB, 0.4 dB, 0.3 dB and 0.25 dB). Furthermore, the distance between the coupler and the Optical Line Terminal (OLT) is set to 2-10 km. To measure the quality of the intercepted signals before and after the interception, the Optisystem software is used.

A Comprehensive Review on the adverse Effect of Traffic Noise

Chee Yung Chua — 2025-08-07

Traffic noise is an omnipresent issue in urban environments, significantly impacting public health, property values, and wildlife. This review paper examines the multifaceted adverse effects of traffic noise, focusing on health consequences, economic impacts on property values, and disruptions to animal habitats. Chronic exposure to high levels of traffic noise is linked to a range of health problems, including cardiovascular diseases, sleep disturbances, and impaired cognitive function. Property values in noise-polluted areas often depreciate, reflecting the decreased desirability of such locations. Additionally, traffic noise interferes with animal communication, breeding, and survival, leading to ecological imbalances. This paper also explores the permissible limits for traffic noise as established by various national and international regulatory bodies, with a particular focus on Malaysia. Existing noise regulations and policies worldwide are reviewed to provide a comparative analysis. Through this comprehensive review, the paper aims to underscore the need for stringent noise control measures and highlight best practices for mitigating the adverse effects of traffic noise. Recommendations for future research and policy development are discussed to promote healthier, quieter, and more sustainable urban environments.

Development and Analysis on the PLC Based Sorting System for Recycle Purpose

Vasanthan Maruthapillai — 2025-08-07

In the world with frequent of trash-production and wastes that need to be disposed of, it usually takes a long time for these waste products to be recycled naturally. Some of the waste can affect the environment if it is not properly recycled, for example, plastic, metal, and glass. The purpose of this project is to develop an PLC based system in sorting different type of items to be recycled, such as glass, plastic, and metal can. This project involved programming of programmable logic controllers for a sorting process for objects to be recycled. The sensors were used to detect the material or object’s type, and this data was used as an input for the system to automatically separate them into three different categories based on the material. Pneumatic cylinder system was used as a pushing mechanism during sorting process. The probability of the constructed prototype in detecting type of material was analysed by testing it with seven objects: a plastic bottle, a plastic bag, a plastic case, a glass bottle, an aluminium can, a metal box, and a metal rod. The findings show that transparency, material type, and material strength significantly impact the results. A high probability in material detection primarily because of the hardness of the materials, facilitating easier differentiation. However, the detection of transparent materials posed challenges, and metal materials were consistently detected in nearly all tests. From the analysis, it is concluded that the current prototype system is still unable to provide high percentage of reliability to detect the materials of the selected object. Increasing the ability to detect the object type is necessary to ensure the success of developing this prototype for PLC based sorting system for recycle purpose. It is envisaged that this system can fasten the sorting process in the future. Besides, with automation application in such industry, there will be a reduced requirement for human labour for the repetitive tasks.

Analysis of the Level of Engagement and Motivation for Early Childhood Sundanese Language Learning with Media Augmented Reality Based Learning

Nichi Hana Karlina — 2025-08-07

Sundanese, as one of the largest regional languages in Indonesia, is facing challenges in maintaining its relevance among the younger generation in the digital era. Research indicates a decline in children's interest in learning Sundanese, which poses a threat to its preservation. This study aims to explore the effectiveness of Augmented Reality (AR)-based learning media in increasing engagement and motivation among early childhood children to learn Sundanese. Using a comparative experimental design with a mixed-method approach, this research involved 30 children aged 4–6 years at KB An Nur Muttaqin. Participants were divided into two groups: one using AR-based learning media and the other using animation videos. Data were collected through pre-tests and post-tests, pictorial Likert scales, structured observations, semi-structured interviews and participatory observation. Results revealed that the AR group showed a significant increase in engagement (mean difference of 3.8) compared to the control group (mean difference of 2.1), with p<0.001. Motivation also improved more significantly in the AR group (mean difference of 0.25) than in the animation group (mean difference of 0.23), with p<0.001. This study demonstrates the potential of AR technology in revitalizing Sundanese language education for young children. By enhancing engagement and motivation, AR provides an innovative approach to addressing the challenges of regional language preservation. This solution not only supports language learning but also ensures that Sundanese culture and heritage are passed on to future generations in the digital age.

Emerging Trends in ChatGPT Research: A Quantitative Literature Review and Clustering Analysis

Yuniar Indrihapsari — 2025-08-07

This study presents a quantitative literature review of ChatGPT research using a clustering approach to analyse knowledge flow through Main Path Analysis (MPA) and Global Main Path (GMP). We constructed a citation network from 1,147 articles published between 2015 and 2023, sourced from the Web of Science. The analysis employed edge-betweenness clustering to identify eight rapidly growing research topics: five clusters focus on ChatGPT in healthcare, two clusters on ChatGPT performance and one on ChatGPT in education. One of the most significant clusters is the impact and ethics of ChatGPT across various sectors, including healthcare, hospitality and cybersecurity, with AGR (0.47) and RGR (0.96) showing very high growth. Meanwhile, other clusters, such as AI chatbots in healthcare and trust and personalization in chatbots, demonstrate more moderate growth rates, with AGR 0.29 and 0.14 and RGR 0.59 and 0.29, respectively. Despite these varying growth rates, key challenges related to ethics, trust and personalization remain critical issues that must be addressed to ensure the successful and effective deployment of ChatGPT across diverse sectors.

Enhancing Flipped Classroom Mastery Learning using ZOOMRBT Apps with Fuzzy Delphi Method Verification

Marina Ibrahim Mukhtar — 2025-08-07

The objective of this research is to investigate expert consensus on designing and developing ZOOMRBT apps for the purpose of enhancing students' mastery learning skills in a flipped classroom environment. This research employed a purposive sampling technique, selecting fourteen experts. The evaluation of flipped classroom deployment and lesson preparation was carried out using the fuzzy Delphi technique and data analysis involved the use of Triangular Fuzzy Numbers and Defuzzification. This research not only seeks to foster independent learners capable of addressing today's challenges but also aims to improve the quality of teaching and learning. In the flipped classroom setting, the ZOOMRBT mobile application is proposed as a model for mastery learning. Proficient instructors participated in this training and the online flipped classroom was carefully planned to integrate pre-class, in-class and post-class activities. The ZOOMRBT app was utilized during these sessions to create a cutting-edge learning environment. The results showed that all the experts unanimously reached a consensus of 100%. The approach of designing and developing ZOOMRBT to enhance mastery learning skills in flipped classroom settings was unanimously accepted by all the experts. The ongoing objective of this study is to assist educators in improving their pedagogical competence and their understanding of digital pedagogy, ultimately empowering students in Malaysian Boarding Schools to become more skilled learners.

Advancements and Challenges in Convolutional Neural Networks for Marine Corrosion Detection and Classification

Md Mahadi Hasan Imran — 2025-08-07

Marine corrosion significantly undermines the structural integrity of maritime infrastructure, necessitating the development of sophisticated techniques for its early detection and classification. This paper offers an exhaustive critical review of Convolutional Neural Networks (CNNs) applied in marine corrosion detection and classification, covering research from 2018 to 2023. It compiles insights from various scholarly articles, elucidating the progression of CNN methodologies in tackling the intricate challenges associated with corrosion in marine, offshore and oil & gas sectors. This review meticulously examines the deployment of CNN technologies in evaluating corrosion across a myriad of maritime assets, including ships, marine structures, offshore platforms and oil & gas pipelines, also construction materials. It explores a broad spectrum of methodologies, underscoring the advancements in CNN-based strategies for corrosion monitoring. Importantly, the review pinpoints key obstacles, innovative strategies and forthcoming trends in the field, offering a comprehensive summary of current research on marine corrosion detection and classification through CNNs. The insights gained from this thorough analysis are instrumental in deepening the understanding of technological and methodological progress, serving as a guide for future research endeavours in the crucial field of maritime asset integrity management.

Process Optimization of Salt Recovery from Nypa fruticans Fronds using Response Surface Methodology

Mohd Syafiq Abdullah — 2025-08-07

Salt, primarily sodium chloride (NaCl), has been a pivotal component of the human diet and a universal condiment since ancient times due to its ability to enhance flavor, act as a preservative, and meet physiological needs. Despite the availability of traditional unrefined salts, the production processes of nypa salt, derived from Nypa fruticans, have not been systematically investigated, and there are no standard methods for its production. This study addresses this gap by recovering salt from Nypa fruticans fronds using the calcination (ashing) method and optimizing the process conditions influencing salt recovery using Response Surface Methodology (RSM). The optimal conditions, determined to be 12 hours of drying time, a calcination temperature of 1200°C, a soaking time of 10 minutes, and a soaking temperature of 40°C, resulted in a predicted pH of 10.76 and a yield of 1.441%. These conditions were verified through additional trials, confirming model’s reliability and offering a sustainable method for salt production from Nypa fruticans fronds.

Enhancing Underwater Colour Restoration using a Detectron-Autoencoder Hybrid Model

Mohamed Amgad — 2025-08-07

Image colorization in underwater environment has become a useful process for submarine operations. Colorized images can enhance the detection of specific targets or anomalies underwater, recently, most researchers of oceanology use high quality underwater images to make detailed analysis of submarine resources. This paper presents deep learning method for underwater images colorization. Studies show that training the deep learning models on a large dataset of underwater images and corresponding ground truth colour images improves objects definition and colorization in the image; hence, our model learns to segment objects from the underwater image, increase the intensity of degraded colours, merge the segmented object with the original image and to another image has different objects in the same context. This paper illustrates image processing techniques such as colour correction and restoration, for imaging challenges and extensive experiments that illustrate the high performance of the proposed model compared to existing state-of-arts, recovering true colours and enhancing overall image quality. The proposed model achieved good results in image colorization with accuracy of 92% that indicates model performance in classifying the targeted object.

Development of Building Information Modelling Model for Residential Housing in Enhancing the Maintenance Work

Nur’Ain Idris — 2025-08-07

Building Information Modelling (BIM) is a digital model of a building that can be visualized in 3D physical form and contains various geometric information, functions, features, or parameters generated by various related software. The main goal of this study is to develop a BIM architecture and structure drawing of semi-detached house for maintenance purposes. Clients, developers, and contractors can use BIM as a resource when executing maintenance projects. There is few software can use to develop BIM system. Contractors around the world use a variety of BIM software, including AutoCAD, Navisworks, Data CAD, and others to develop BIM in software. In order to develop BIM system in this study, Revit software has been chosen to create a semi-detached house. Between the conventional method and the BIM method, there is difference. BIM is allowing different design teams to collaborate more easily by using a "live" version of the building model compared to conventional method which is working in storage facilities and taking snapshots. Getting a house plan from a developer is the first step in build BIM. The drawing needs to be converted from a PDF file to AutoCAD format. Next, create architectural and structural drawings using the Revit. The outcome of this study, BIM is developed from architecture and structure drawing using Revit software. Therefore, it may be inferred that using Revit software to create BIM saves more time for maintenance purpose.

Machine Learning Techniques’ Comparison for Early Detection Model of Transformer Health Index Classification

Md Nasif Rahman — 2025-08-07

Timely detection of anomalies in power transformers is essential to maintain a constant, uninterrupted supply of electricity. Early Detection Model (EDM) is a Machine Learning-based approach to derive the Health Index (HI) of power transformers using routine test features, where the absence of non-routine test features is compensated by use of derived features. Latest research in this area involves the use of only Support Vector Machine (SVM) as the classifier for constructing the sub-models of EDM. This paper explores the use of two other classifiers, Random Forest and Naïve Bayesian to deduce a comparative analysis among all three classifiers. The classifiers were developed in Google Colab platform using already available datasets. The datasets were divided into three parts for training, validation and testing. The results of the simulation showed that Naïve Bayesian achieved the highest average accuracy for all the sub-models, which was 92.45%. On the contrary, Random Forest and SVM appeared to have almost similar level of performance with average accuracies of 89.73% and 89.53% respectively. Naïve Bayesian also seemed to have the lead in terms of other performance metrics as well, making it stand out as the most preferred classifier in the experiment conducted. Further research on this topic may reveal more optimized application of Machine Learning models for early detection of transformer anomalies.

Enhancing Big Data Governance Practices: Addressing Security, Privacy and Ethical Challenges

Yunusa Adamu Bena — 2025-08-07

The exponential growth of data from diverse sources has amplified the challenges of managing security, privacy, and ethical concerns in big data governance. Organizations face significant risks related to data breaches, regulatory compliance, and ethical dilemmas, necessitating robust governance frameworks. This study examines the state of BDG practices in data-driven organizations, focusing specifically on security, privacy, and ethical concerns. Through a systematic literature review of 53 peer-reviewed articles published between 2019 and 2024, the study identifies critical gaps and opportunities in current governance practices. The findings reveal that while BDG practices enhance data security, ensure compliance, and foster ethical accountability, organizations struggle with fragmented frameworks, insufficient scalability, and a lack of unified strategies for addressing interconnected challenges. To bridge these gaps, this study proposes the Security, Privacy, and Ethical Concerns Maturity Assessment Model (SPEC MAM), which evaluates governance maturity across four dimensions: Security, Privacy, Ethical Concerns, and Integrated Concerns. The model categorizes maturity into four progressive stages: Foundational, Policy Development, Integrated Practices, and Proactive, providing organizations with a structured approach to assess and enhance governance maturity. While the model offers a systematic framework for addressing BDG challenges, its real-world application and scalability remain areas for further exploration. This study contributes to academia and practice by advancing BDG research and equipping organizations with actionable strategies to address security, privacy, and ethical concerns in an increasingly complex data landscape.

Enhancing Manufacturing Productivity through Refined Work-Study Techniques: A Critical Review of Allowance Factors in Time Study

Haziq Asyraaf Abu Bakar — 2025-08-07

This systematic review examines the critical role of allowance factors in enhancing work measurement accuracy within mechanical and industrial engineering domains. The study addresses inconsistencies in understanding and implementing allowances across various manufacturing and production environments. Through a comprehensive analysis of extant literature and industry practices, this research categorizes and evaluates the impact of diverse allowance types on productivity and operational efficiency in engineering settings. The investigation encompasses three primary allowance categories: constant (e.g., personal needs, basic fatigue), variable (e.g., abnormal posture, environmental factors), and special (e.g., unavoidable delays, policy-based considerations). These allowances are analyzed for their application and impact in industrial and mechanical engineering contexts. Based on the literature synthesis, a methodological framework for allowance determination and implementation is proposed. This framework comprises systematic observation, empirical data collection through time studies, statistical analysis, and validation processes specific to engineering applications. It aims to enhance the precision of allowance estimation in developing equitable work standards, influencing labor cost calculations, productivity metrics, and operational efficiency in manufacturing systems. This study highlights the importance of accurate allowance integration in work measurement methodologies for mechanical and industrial engineering. It provides a structured approach to allowance application, contributing to the advancement of work measurement techniques in complex manufacturing environments. Findings offer evidence-based recommendations for implementing robust allowance strategies, aimed at optimizing work measurement practices and fostering sustainable operational improvements. This research serves as a comprehensive resource for engineers and researchers seeking to enhance the accuracy and effectiveness of their work measurement methodologies. By synthesizing current knowledge and best practices in allowance factor application, this study bridges the gap between theoretical concepts and practical implementation, ultimately improving productivity and efficiency in manufacturing and production processes.