Search

A Manual visual inspection system in the manufacturing company resulted in decreased inspection performance due to the human error factor. Defects that naturally occurring in the mainboard are oxidation and it cannot be repaired. This study aims to improve the quality of inspection in identifying defects, especially the mainboard which is exposed to oxidation in the area of IC (Integrated Circuit) components. The system uses a digital microscope camera to capture the image of the IC component and using the Microsoft Visual Studio application as the interface. Template matching and the NCC (Normalized Cross-Correlation) algorithm as a method for detecting oxidation. NCC is a calculation of the correlation function between two imageries as a comparison value between the template image and the test object image. The experiment was carried out twice on 10 unoxidized PCB (Printed Circuit Board) samples and 10 oxidation-affected PCB samples. The percentage of successful detection of IC components in this study reached 100% with the accuracy or thoroughness value of 90%. The light intensity greatly affects the percentage outcome due to the inconstant light intensity resulting in a percentage value change from testing.

The quality of production in a company is a major guarantee. It needed a human resources, assistance company to control the product quality. However, human errors causing product failures are still found in many cases. Some disadvantages of using human power for the inspection process are fatigued in the eye when carrying out inspections using manual equipment such as magnifying glasses and microscopes. So in this study a system was designed to detect and identify anomalies on the soldering of ICSOIC on surface mount devices using image processing and neural networks. To separate the soldering colour from the SOIC IC, the HSL method is used while the identification of the type of anomaly in the SOIC IC used a backpropagation neural network. From the results of the system testing carried out, the success rate reached 92.72% from 11 trials.

This paper describes a control system which can provide the assistive torque for the lower limb exoskeleton. To provide the assistive torque for the lower limb exoskeleton need to be known the gait phases in one gait cycle walking as the reference signal. The gait cycle information already got from another work with the same prototype of the exoskeleton. The gait cycle signal has been generated from Inertial Moment Unit (IMU) sensor and built-in encoder on the Maxon motor. The decision tree method has been used as the proposed method in order to provide the suitable assistive torque. The real-time experimental result has been carried out to verify the performance of the proposed method. To examine the proposed method in a real-time application, some walking type has been done. The result shows that the proposed method can provide the assistive torque even in several different walking types.

This paper presents a method which can classify the human walking gait cycle based on the Inertial Measurement Unit (IMU) and hip joint angle of lower limb exoskeleton information happened in one gait cycle. By using the information of IMU dan hip joint angle, each gait cycle on the leg supposed to be classified. In order to get the walking gait cycle prediction, the Neural Network (NN) has been chosen as the control system. The gait cycle prediction from NN, later on, will be classified using a simple comparison programming. This method tested on the prototype of the exoskeleton in real-time application. To verify the proposed method on the real-time application, some experiments have been carried out at different walking speed. The experimental results proved that the proposed method can classify the human walking gait cycle properly.

Designing an effective machine learning model for prediction or classification problem is a tedious endeavor. Significant time and expertise are needed to customize the model for a specific problem. A significant way to reduce the complicated design is by using Automated Machine Learning (AML) that can intelligently optimize the best pipeline suitable for a problem or dataset. This paper demonstrates the utilization of an AML that has been developed with a meta-heuristic algorithm namely Genetic Programming (GP). Empirical experiment has been conducted to test the performances of AML on a real dataset of house prices in the area of Petaling Jaya, Selangor. The results show that the AML with GP able to produce the best pipeline of machine learning with high score of accuracy and minimal error. (Abstract)

Social and Solidarity Economy (SSE) initiative which has been implemented in Malaysia and shows it success when four case study reported SSE communities or organization displayed efforts in helping the public on exiting economic crisis (Tan, 2016). However, the information misgovernance destroys the SSE business value, slow down organization and creates unnecessary risk (Blair, 2011) to the SSE communities when the government and state neglected SSE successful initiatives in combating economic crisis of the public, yet not having them as a solution to the economic downturn. Therefore, the study aims to identify factors of why integration of information governance and SSE develop better economic initiatives. Information Governance for SSE Framework were tested to two focus group using empirical study. The framework was simulated on the focus group to analyze its significant impact. The results of the study important to the application of IG towards SSE since it decreases opportunity of business operation mistakes and dark data.

Authentication has been advanced by implying advancement in security such as cryptography to protect password against data leaking and sniffing; captcha to prevent robot system; Secure Sockets Layer designed to protect users against phishing website and many other solution invented for different types of vulnerabilities. Although there are other methods suggested for preventing shoulder surfing (i.e. exposure of password to nearby observing-person), the techniques still have some drawback. Some of the proposed techniques also required high computational process with high range of possible answer. This paper proposed a shoulder-surfing-proof graphical based authentication with direction scheme that combines technique adapted from Passfaces scheme and a selection of direction. User is required to choose four faces and one direction to serve as the secret authentication. Besides shoulder surfing, the proposed method also caters security in data transmission storage as hashing is applied along the communication. This technique has been proven to demonstrate the robustness, security strength and the functionality advantages of the system. The result from the security testing shows that image based passwords is more secure because the possibility of the image to be guessed is very low.

The image processing is not only used to detect and recognize the object. It also can be used to calculate the amount of a buckle object. This paper presented a method which can calculate the amount of a buckle gas cylinder seal. At first, the camera detected a buckle of the seal in real-time. Then, the HSL color space is used to remove background image so that blob counting can detect the seal easily. The pixel of blob counting obtained will be compared to an original pixel in order to build the relation using the polynomial equation to count the seal. By combining the blob counting result, the buckle of the seal can be counted as seen in the experimental results which are done in the real-time application.

The charge unbalancing decreases the total storage capacity and the entire life cycle of batteries. In order to prevent the charge differences among cells, cell balancing system or charge equalization have been proposed. This paper proposes an active balancing system based on switched-capacitor charge equalization for series-connected batteries. This circuit topology allows two series-connected cells to charge one capacitor and then the charge of capacitor will be transferred to the other cell which is insufficiently charged. In order to achieve better efficiency, cell voltages are monitored to decide which cell is insufficiently charged. Besides, small inductors are inserted into the switching capacitors charging path to reduce the current spikes.

A high efficiency, low profile and high step-up dc-dc converter is proposed for low dc voltage renewable energy system. Generally, the power source such as photovoltaic array and the fuel cell stack have low voltage output because of that a high voltage step-up converter is required to boost the voltage much higher than the voltage level for front-end application. The efficiency and voltage gain of conventional dc-dc converter are limit due to the parasitic components. In order to solve the above mentioned problem, the proposed converter is based on active-clamp technique and the converter is the combination of a forward converter and a flyback converter. Multiple secondary windings can be employed for low profile design and help to reduce high transformer copper loss and high output rectifiers conduction losses. The proposed converter is confirmed with experiments on 250-W prototype and the maximum efficiency of ±94% was measured at full load.

A single ring resonator dual-band bandpass filter using quarter-wavelength resonators suitable for WLAN application is presented in this paper. Both side of ring resonator consist of quarter-wavelength parallel coupled-lines. In further reducing the size of the filter, meander structure is applied to an existing design. Using different number of meander slots, a microstrip dual-band single ring resonator are applied and realized on FR4 substrate to suite filter operating at 2.5 GHz and 5.2 GHz. Simulation is carried out in analyzing the performance of the design using high frequency system simulation software and fabrication is done to validate the topology.

Malaysian Muslims have been facing numerous difficulties in learning Arabic language. This problem has inevitably caused them complications in understanding the basic meanings of Al-Quran words and sentences. Presently, there exists a local technique named Al Furqan, based entirely on surahs from the Al-Quran that teaches Malaysian Muslims in classes to learn Arabic easily without much memorizing and writing, and without any homework. However, the element of interaction and engagement prior and after classes is missing. There is a learning gap which can be filled by mobile learning. Thus, this research work aims to design, develop and test a scaffolded mobile educational game called Easy Arabic for learning Arabic using the Al Furqan technique. The research is based on the developmental approach and consists of four major phases which are theoretical study, prototype design, prototype development and prototype testing. The prototype has been successfully developed and installed on Android platform, and has been tested by a sample of 30 Al Furqan learners. Most of them found this prototype significantly helpful in bridging the learning gap of understanding the basic technique of Al Furqan prior and after classes.

As mobile cloud computing applications are becoming widely used to mitigate the limitations of mobile devices and exploit cloud services, testing these applications becomes critical. Testing mobile cloud application is complicated further due to dynamic assign of cloud resource to applications that run on mobile devices, due to the increased multiple factors in test case generations and execution. This paper will critically review the three related areas of mobile, cloud and mobile cloud applications testing, in terms of features and models. Furthermore, it will show the necessity of a mobile cloud computing applications testing model due to its uniqueness from both native mobile application testing models and cloud applications testing models. Based on this review, although testing mobile cloud computing applications is critical, there is currently no model to test mobile cloud computing applications.

The firewall deployments have become a favorite research area within IT world. The design of framework and the algorithm of rule set can affect the overall performance of the filtering system. This review paper discusses some of the unique researches to build an effective and efficient filtering model. The application level and system level performance must work in efficient manner to build up a fast and accurate filtering process. The mechanism or how the packet is processed from being received at the physical layer or the network card until further processed by the application layer support software such as an IDS is highly influential for overall packet filtering infrastructure.

While studies investigating emotional intelligence have been extensively reported in the literature, very few have focused on librarians. Hence, a study was carried out using qualitative approach to identify the dimension of emotional intelligence of Malaysian Public Librarians. Based on the findings of the qualitative study, an instrument measuring librarian emotional intelligence was developed and empirically tested using survey method involving 180 librarians. Accordingly, Librarian Emotional Intelligence Assessment Tool (LEIAT) was developed. LEIAT can be used by Malaysian public libraries to continuously assess the level of emotional intelligence among their librarians. Based on the outcome of the assessment, training needs can be identified. In addition, It can also be used as a tool during entry interview for the post of librarians. LEIAT can be commercialized as a software product. The potential clients of this product would be libraries and corporate resources centers. Keywordsemotional intelligence, librarian, assessment tool, Malaysia

In Higher Education learning environment, managing students learning means managing emotions. This paper presents the efforts in recognizing student's emotion to promote interactive teaching learning atmosphere in a Higher Education environment. As to now not many variety of tools provided to help lecturers to assess their teaching, our research suggests a new system to help lecturers to identify whether their teaching style has positive impact on students by recognizing student's emotion in the class. This system propose the positive class emotion (as result of positive student's emotion) to encourage lecturers to retain their teaching style or vice versa. This software is considered as Emotion Based System to enhance teaching and learning. In this paper we compare some of the existing works and identify how emotions can be modeled in order to enhance the system services and response by suggesting meaningful information to users. We also define and discuss the appropriate capacity of meaningful information to generate analysis and recommendations according to theory of teaching and learning. We do our experiments in INTI International University Malaysia involving approximately 300 students from various faculties.

Online businesses in Malaysia suffer from the lack of non-repudiation in the ordering, delivery and payment process. Most businesses cannot sustain their growth because of this limitation. In this paper we propose a new model for the implementation of online business. The solutions can also be adapted to any country for the most part except for the delivery process that requires the use of a smartcard-based national identity card. The non-repudiation evidences that are generated from the system can be used by both the consumer and the merchant to resolve possible future disputes. The evidence is also admissible in the court of law since they are generated from the Public Key Infrastructure (PKI) that is recognized by the authority as legally binding.

This paper presents modeling concern to construct context data format that is applicable for context-aware computing. The first concern deals with developing abstract model at application entity which is defined independently from presentation entity. Regarding to this matter, the context data format have been constructed as higher and lower abstraction model presented in OWL and XML instance. The second concern deals with mapping lower abstraction model to exchange format. JSON exchange format is proposed in this paper as the alternative to the incumbent XML. The context data format, which is constructed at lower abstraction model, has been transformed to JSON by applying XSLT rule. As a result, JSON exchange format is obviously smaller than XML in term of exchange format size.

Currently, power demand in the world has grown at a rate faster than the infrastructure development. This has caused the power system network to be stressful that could disrupt system stability leading to voltage collapse in the entire system. This incident led to a cascading blackout in the whole system and required a voltage stability analysis to be properly conducted. Line outage in a power system could also lead to the event of voltage collapse which implies the contingency in the system. Line outage contingencies are ranked so that the line which highly affects the system when there is an outage occurs in this line in terms of voltage instability could be identified. The contingency ranking process can be conducted by computing the line stability index of each line for a particular line outage and sort them in descending order. The contingency which is ranked the highest implies that it contributed to system instability. This paper presents a new voltage stability index referred to a line used to rank the line outage contingency. The information from the contingency ranking indicates the severity of the voltage stability condition in a power system due to line outage.