UTAR Institutional Repository (Universiti Tunku Abdul Rahman)
Not a member yet
6132 research outputs found
Sort by
Asset and linen monitoring information system framework for UTAR Hospital
Get PDFThe proposed project is developed and designed an Asset and Linen Monitoring Information System Framework by using the Frappe Framework ERPNext for UTAR Hospital, addressing the need for efficient monitoring and tracking asset and linen inventory management. The traditional monitoring method of manually searching and checking assets and linens inventory used in hospital may be time-consuming and prone to human errors. It is leading to unnecessary losses on missing assets or linens and inefficiencies of monitoring and tracking asset and linen. This system offers a user-friendly interface with a dynamic dashboard that streamlines operational hospital and enhances management of asset and linen. It is allowing administrators and staff to monitor assets and linens effectively in real-time data collection. The integrated features of real-time alert and role-based access control fall short in existing system, which are played a crucial role in maintaining operational efficiency and system data security. This project aims to fill this gap by implementing a comprehensive system that tracks and monitors assets and linens as well as sends timely notifications for PPM of asset and updates on newly remarked items. It keeps all the relevant department personnel of UTAR Hospital informed promptly to take action on the notification and make decision. It is implemented the Frappe Framework ERPNext to develop a secure and efficient system supporting those functionalities. Through the proposed project, the management of assets and linens in UTAR Hospital has been enhanced, reducing losses on missing assets or linens and improving operational hospital efficiency. By implementation of the system, it assists the administrator and manager of UTAR Hospital to better monitor and track the assets and linen inventory in real-time
Early childhood interactive education media: Penguinplayground edu
Get PDFThis study presents a comprehensive mobile application multimedia kindergarten learning platform that integrates digital technologies into the educational domain. The platform utilizes the immersive potential of multimedia and mobile applications to create a dynamic and engaging learning environment specifically for kindergarten-aged children. The main motivation behind the platform is the urgent need to bridge the gap between traditional pedagogy and the vast potential offered by digital advancements. Furthermore, this study aims to address the inherent shortcomings of current multimedia kindergarten learning platforms. By integrating interactive multimedia content, the platform facilitates rich educational journeys that inspire curiosity, creativity and personalized learning experiences. The platform works to redefine the way children interact with educational resources, providing accessibility, adaptability, and engagement in a variety of learning environments. Overall, this research has pioneered a transformative approach to early childhood education that promotes a holistic and engaging learning experience. By combining cutting-edge technology with pedagogical expertise, the Multimedia Kindergarten Learning Platform for Online/Offline/Mobile Applications aims to empower parents, educators and early childhood learners, ultimately shaping the future of early childhood education
Parking management with mobile NFC authentication
Get PDFThis project introduces a mobile application aimed at transforming residential parking management through the use of Near Field Communication (NFC) technology for secure and efficient vehicle authentication. Designed to meet the needs of residents, management, and visitors, the application offers a comprehensive solution to common parking challenges such as unauthorized access, poor space utilization, and lack of real-time monitoring. The key innovation of this system is its integration of NFC-based authentication. Vehicles are equipped with NFC tags containing specific data, which management can scan to verify parking authorization. If a tag fails to return valid data, it may indicate a counterfeit, allowing management to take immediate action and prevent unauthorized parking. This improves both security and transparency in the parking process. To support daily use, the application also enables residents to report illegal parking through a built-in reporting module and request temporary spots if their designated space is taken. For visitors, the system streamlines parking access: residents can reserve a spot and generate a unique reference code. Visitors enter this code into the application to receive their assigned parking lot, ensuring that only approved guests gain access. The management team benefits from a real-time overview of parking occupancy, allowing them to monitor space usage and respond promptly to reported issues. A visual floor plan aids visitors in navigating to their assigned spots, further enhancing usability. Unlike existing solutions, which often address only a portion of the parking management process, this application provides a fully integrated system that combines secure vehicle verification, resident engagement, visitor access control, and real-time administrative tools. The objective is to develop a reliable, intuitive mobile application that leverages NFC for secure sticker validation, facilitates efficient space allocation, and improves communication among users. By enhancing both the security and convenience of parking management, this solution aims to reduce unauthorized parking incidents, optimize space usage, and provide a better experience for all stakeholders
Oasis: A computer vision approach to self watering system for green air purifier
Get PDFIn this project, a blockchain-based e-commerce system is proposed to transform the
online marketplace by addressing key challenges found in traditional e-commerce
platforms. This project’s main objective is to eliminate middlemen or intermediaries to
streamline transactions, reduce transaction costs, and enhance transparency and trust
between buyers and sellers. By utilizing blockchain technology, the system provides
decentralized transaction management, user identity verification, and product
authenticity. This will provide a secure and efficient platform for e-commerce activities.
On the other hand, the user interface is designed to be intuitive and user-friendly to
facilitate seamless interaction with the smart contract functionalities. This system will
not only redefine the e-commerce experience but also aligns with the sustainable
practices by optimizing logistics and minimizing environmental impacts caused by ecommerce
activities. This project aims to create an innovative e-commerce platform
that is cost-effective, secure and environmentally responsible, which will enhance the
user satisfaction and introducing a new era of digital commerce
To what extent do digital learning platforms impact the learning experience and academic performance of UTAR FICT students
Get PDFThis research examines the extent to which online learning websites impact the learning
process and academic performance of university students based on the Self-Determination
Theory (SDT). As development in online learning continues to grow, understanding
psychological and academic effects of such online websites has become more important.
Three separate SDT constructs—autonomy, competence, and relatedness—were employed,
and learning experience as the mediator variable, with academic performance as the
dependent variable. Quantitative research design was employed, with a standardized
questionnaire completed by 210 undergraduate students at the Faculty of Information and
Communication Technology (FICT), Universiti Tunku Abdul Rahman (UTAR).
Data collection and analysis were conducted using SPSS software, using descriptive and
inferential methods. Descriptive statistics like mean, standard deviation, skewness, and
kurtosis were used to evaluate data distribution and central tendencies. Internal consistency
greater than 0.70 in Cronbach's Alpha confirmed the reliability of measurement constructs.
Pearson correlation analysis indicated significant correlations between variables. Multiple
regression analysis was used to derive the predictive power of autonomy, competence, and
relatedness to academic performance. Additionally, moderated regression analysis was used
to quantify the impact of learning experience on these relationships as well as including
interaction terms for testing its moderation effect.
The study revealed that autonomy, competence, and relatedness are positively associated with
the learning experience of the students that, consequently, affect the academic performance
positively. The study concludes that, when online learning environments are being designed
to serve the psychological needs of the learners, they may improve academic achievement and
engagement in students. These conclusions are valuable for teaching faculty and institutions
to reaffirm student support systems and online learning approaches at the university level
Parallel metaheuristic algorithm for route planning using CUDA
Get PDFThis research will be focusing on developing a Parallel Metaheuristic Algorithm for Route Planning using CUDA to improve the efficiency and performance of route planning. The increasing demand for more efficient route planning approaches, which includes several factors such as cost savings, timely deliveries and reduced carbon emissions, has led to a surge in demand for more advanced route planning algorithms in search of more efficient solutions.
The problem that this research will be tackling is the Travelling Salesman Problem (TSP), which is a specific type of route planning problem where the main objective of it is to find out the optimal set of routes for a given number of vehicles to transport goods to a defined set of destinations. TSPs are known to be NP-hard problems[1] where an increase in the number of vehicles and destinations will significantly increase the computational time required to obtain an optimal solution. Existing works that utilized metaheuristic algorithms have shown their flexibility in solving multiple TSP variants and their capabilities in obtaining near-optimal solutions within a reasonable amount of time. However, due to the limitations of CPUs in terms of parallelization, these algorithms do not perform well as they are highly iterative.
The proposed approach will be utilizing the Compute Unified Device Architecture (CUDA) to enhance the performance and efficiency of metaheuristic algorithms in finding optimal solutions for the TSP by leveraging the parallel processing capabilities of Nvidia Graphics Processing Units (GPUs). This research aims to significantly speed up solution searching for the TSP by using GPUs compared to CPUs. Besides, this research strives to provide a foundation for future research on parallel metaheuristic algorithms, and to further encourage their implementations in real-world instances of route planning.
Area of Study: Massively Parallel Computing, Combinatorial Optimization
Keywords: Parallel Metaheuristic Algorithm, Travelling Salesman Problem, CUDA, GPU, Genetic Algorith
Green synthesis and characterization of copper oxide and zinc oxide nanoparticles derived from aqueous mangosteen (garcinia mangostana) leaf extract for palm oil mill effluent (pome) treatment
Get PDFGreen synthesis of nanomaterials using plant extracts offers a sustainable and eco-friendly alternative to conventional methods. In this study, copper oxide (CuO), zinc oxide (ZnO), and zinc oxide-copper oxide (ZnO-CuO) nanomaterials were successfully synthesized using mangosteen (Garcinia mangostana) leaf extract (MLE). The synthesis process was optimized based on key parameters such as MLE concentration, calcination temperature, and the Zn-to-Cu ratio. The synthesized nanomaterials were characterized using a range of analytical tools to assess their physicochemical properties. The optimized CuO, ZnO and ZnO-CuO nanomaterials were chosen in inhibiting selected bacteria using Broth macrodilution assay and treating palm oil mill effluent (POME) at different conditions under light-emitting diode (LED). In preliminary in vitro antibacterial study, MLE-mediated synthesized CuO nanoparticles (NPs) and ZnO NPs calcinated at 500oC and 0.05 g/mL MLE-mediated synthesized 70ZnO-30CuO nanocomposites (NCs) calcinated at 500oC were chosen due to their small crystalline size. The average particle sizes of the synthesized nanomaterials were 14.31 nm ± 5.368, 29.61 nm ± 12.573 and 23.09 nm ± 10.656 and mostly spherical structure. The ZnO NPs demonstrated the highest antibacterial efficacy with minimum inhibitory concentrations (MIC) of 15.63 μg/mL (Staphylococcus aureus, Bacillus subtilis), 62.50 μg/mL (Escherichia coli), and 125.00 μg/mL (Klebsiella pneumoniae). ZnO-CuO NCs ranked second, with MICs of 62.50 μg/mL, 62.50 μg/mL, 250.00 μg/mL, and 125.00 μg/mL, respectively. CuO NPs were the least effective, showing higher MICs of 250.00 μg/mL, 125.00 μg/mL, 125.00 μg/mL, and 500.00 μg/mL. For POME treatment, ZnO NPs calcinated at 400°C achieved the best performance with COD and AN removal efficiencies at 54.05% ± 0.03 and 38.72% ± 0.00, respectively, using 50 mg of NPs under 120 min of blue LED exposure. In contrast, 150 mg of CuO NPs and ZnO-CuO NCs were required for optimal photodegradation under blue LED for 150 min, achieving COD and AN removal efficiencies of 58.58% ± 0.02 and 38.86% ± 0.01 (CuO NPs) and 59.72% ± 0.02 and 27.80% ± 0.01 (ZnO-CuO NCs), respectively. In conclusion, this study demonstrates the potential of MLE-mediated green synthesis of nanomaterials as a sustainable approach for producing effective photocatalyst and antibacterial agents. ZnO NPs, in particular, exhibit high promise for environmental and biomedical applications. This study introduced a sustainable and environmentally friendly approach to MLE-based nanomaterial synthesis. Additionally, it offered a viable alternative method of bacterial inhibition and a treatment for POME. Moreover, the use of treated POME shown good agricultural development. This study proposed breakthroughs in scientific knowledge of nanomaterial synthesis, antibacterial and photocatalysis mechanisms. This work has a significant contribution to science and society as the results have far-reaching implications for sustainable development, public health and environmental protection – all of which are parallel to the sixth and twelfth sustainable development goals
Valet - Smart parking management system for UTAR Kampar
Get PDFThis project aims to create a smart parking system for UTAR Kampar, leveraging AIoT and OCR technologies. The primary focus is on using AIoT cameras to conduct OCR for reading vehicle license plates. The system is designed to facilitate the process of finding parking and enhance the functionality of the accompanying application. The research aims to introduce an advanced parking system powered by OCR technology, which includes data collection, analysis of existing parking setups, and the development of a user-friendly app or dashboard. This system will provide real-time car park information, validate vehicle plate numbers, enforce parking regulations, and enhance campus security. Instead of manually searching for their parked car row by row, users will be able to track their car’s location through the application. Additionally, users can check the availability of parking spaces through the app, making parking more efficient and convenient
Predicting financial distress with time-series and market sentiment Integration for solving real world problem - capital a berhad
Get PDFDuring the COVID-19 pandemic, Capital A Berhad, previously known as AirAsia, encountered substantial financial difficulties. This raised concerns about its financial health and PN17 classification status. Hence, this study examines the company’s financial distress by combining sentiment analysis with financial data using time-series methods. It investigates how market sentiment, drawn from news outlets and customer reviews, influences financial performance and highlights discrepancies between the Altman Z-score and the government’s PN17 classification. Furthermore, the study analyzes the effects of restructuring announcements and market expansion efforts on the company’s financial outcomes. Therefore, three forecasting approaches are compared: time-series analysis of market indices, company-specific financial data, and market sentiment analysis. In this stage, advanced techniques such as LSTM networks for financial data and market indices, alongside BERT model for sentiment analysis, are utilized to construct predictive models. The study follows the CRISP-DM framework, with performance assessed through metrics like mean squared error (MSE) and confusion matrices. This is to evaluate the model's accuracy and robustness. Thus, by outlining the strengths and weaknesses of each approach, this research offers valuable insights to internal auditors and decision-makers at Capital A Berhad, supporting enhanced risk management and financial forecasting practices
Optimize and deploy machine learning algorithms on embedded devices for manufacturing applications
Get PDFThis proposal discusses the techniques of optimizing and deploying machine learning algorithms on embedded devices for manufacturing applications; We investigate problems of printed circuit board (PCB) defects and artificial intelligence in embedded system. PCB defects detection had been an essential problem to solve in manufacturing environments, whether it is quality assurance or the needs for PCB inspection had been ramping since decades ago. Fundamental limitation of human-based judgement of inspection engineers is the primary cause of faulty products including PCB defects exiting the manufacturing environment. On the other hand, artificial intelligence had been ways to enhances embedded system by enabling real-time, accurate detection and management of PCB defects through advanced pattern recognition and automated inspection methods. However, embedded system often been having limited computing power, small memory storage and relies on battery capacity. Not to say the difficulty in deploying either artificial intelligence or deep learning in embedded environments due to significant parameters size and computational complexity. In recent studies, we seen developers and researchers proposing solutions on deep learning algorithms like YOLO, EfficientNet, CNN, MobileNet etc. On the other hand, network compression and acceleration techniques such as pruning and quantization also been the focus of the studies for light-weight algorithms in embedded system. While in our studies, we primarily focusing on the deployment and fine-tuning of deep learning model which is YOLOv5 for PCB defects detection. We aim to levitate the baseline YOLOv5 into a state-of-the-art version that focus on lightweight performance, called the LW-YOLOv5, which can been deploy seamlessly into embedded systems for manufacturing applications. As we conduction evaluation experiment on our model using openly accessible datasets like PKU-Market-PCB and perform comparative studies with the latest proposed solutions