In 2017, about 144 million people collected water from untreated water bodies, such as lakes, streams, and rivers. One of the major causes of death is consuming contaminated or polluted water. Measuring and monitoring water quality are usually done using two methods. The conventional method occurs by taking samples of water and then transferring them to the laboratory. The second method is real-time water quality by integrating the Internet of Things (IoT). This method is preferable as it only requires smart sensors and processors to monitor the water quality. Among the widely used processors are the Arduino and Raspberry Pi. However, these two processors have a limitation, including a limited number of hard-coded input/output pins, unlike the Field Programmable Gate Array (FPGA) processor, which has many input/output pins not hard-coded to allow different interfacing of multiple sensors. Based on the literature, an FPGA platform provides more flexibility and reconfigurability features when compared with the Arduino and Raspberry Pi. This research mainly focuses on designing a reconfigurable multi-core Smart Water Quality System (SWQS) measuring the pH, Total Dissolved Solids (TDS), and turbidity parameters. The hardware design was developed based on the system-on-chip (SoC) design methodology on an FPGA to parallelize the SWQS functionality. A Liquid-Crystal Display (LCD) display has been incorporated into the Raspberry Pi to show real-time data. The Platform Designer on Quartus II has been used to instantiate four cores to integrate all functions into one processor. The Eclipse tool on Quartus II, on the other hand, was used to program the sensors using embedded C language. The proposed design has been implemented on DE10 Nano FPGA-SoC consuming 9% of logic resources and 57% of internal memory. To verify the proposed system functionality, the sensors were tested on different liquids. To test the pH level, the pH sensor was tested on pure water, lemon juice, and milk to show the acidity and alkalinity. The pH sensor showed 7, less nearly 2, and less than 8 for pure water, lemon juice, and milk, respectively. The TDS sensor successfully detected the salt added to the water, and the TDS values increased to approximately 1800 ppm. Finally, the turbidity sensor revealed the dust inserted in the solution. The more dust in the liquid, the more TDS value there was recorded. Additionally, results showed that the processing time of all the sensors using FPGA is approximately 300 ms for ten readings; on the other hand, the processing time of using other processors, such as Arduino, took 2 s for ten readings. This is because FPGA is functioning at 100 MHz, while Arduino’s frequency is not more than 24 MHz. All real-time sensor readings were shown on a Linux Terminal. In conclusion, the proposed FPGA-based system can be utilized as a heterogeneous multi-core system for many applications, including the SWQS.

A company that prioritises quality will have an advantage in the global market because not all companies can achieve and sustain high levels of quality. Wahana Interfood Nusantara Company is Indonesia's most complete cocoa bean processing company under the SCHOKO brand. The company processes raw cocoa beans to produce high-quality chocolate and cocoa products. One of the products that experienced defects at Wahana Interfood Nusantara Company was the white couverture chocolate product. The problem statement of this study was what types of defects contribute to the poor quality of white couverture chocolate and what factors contribute to the defects. The aim and objectives of this study were to identify and analyse the current quality state of the production process of Product White Couverture chocolate at Wahana Interfood Nusantara company, Analyse the data collected during the production process of Product White Couverture chocolate to come up with root causes and identify the optimum solution, and Identify improvements made because of proposed solutions. This study uses the DMAIC method (Define, Measure, Analyze, Improve, and Control). They obtained 5 Critical to Quality chocolate attributes: broken, cracked, peeled-off, soft, and porous. Regarding the p-chart control calculation, the CL, UCL, and LCL values were 0.024, 0.026, and 0.022. It means the data result is in control. The DPMO value was 4788.49, and the Sigma value was 2.496, which means that the white couverture chocolate production process is in Indonesia's average Sigma value industry. It was found that 80% of the most dominant defects were peeled off (38.5%), cracked (28.7%), and broken (25.3%). The identified factors include workers or employees (people), tools (equipment), the environment (environment), and methods (methods). Suggestions for improvements are given to maintain regular machine maintenance, conduct employee training, and carry out SOPs correctly.

The congestion at the lower frequency bands is pushing the demands for the usage and operation in the higher bands. The bands above 10 GHz are now required to satisfy the tremendously increasing needs in satellite communication (SatCom) systems. The biggest problem in deploying frequencies above 10 GHz in tropical-equatorial regions is that such areas will experience acute degradation of signal quality due to heavy rainfall throughout the year. A reliable estimation of rain attenuation is required. Dependable fade margins are required in mitigating the harsh signal losses. It is found that data validation for rain attenuation estimation using the frequency scaling technique is not available in any previous studies conducted in tropical regions. It has been suggested that instead of using the conventional method of predicting rain attenuation using the point rainfall rate information, an applicable non-meteorological technique should also be established. A frequency scaling technique can be the alternatives mean to predict rain attenuation when rainfall data is not available. The objectives of the research entail rain-induced attenuation studies for SatCom in tropical regions. They comprise identifying the best fade margin for rain attenuation at various links, formulating a new frequency scaling model with improved accuracy, and validating the proposed model. The methodologies involved in the study encompass the processing of the beacon signals into first-order statistics of rain attenuation. This, later, leads to the generation of monthly and annual rain attenuation Cumulative Distribution Functions (CDFs). The worst month analysis for rain intensity and rain attenuation was also carried out. The signal loss is expected to be appalling in the worst month because of the high occurrence of rain events. The required fade margin was determined from an exceedance at a specific point from the annual CDF. In brief, the frequency scaling model was derived based on the correlation between the attenuation ratio of a higher and lower frequency against the attenuation at a lower frequency. The newly developed formulation was utilized to generate new CDFs of attenuation at different frequencies. The proposed model offers a lower RMSE value and percentage error of 2.8 and 11.3% respectively. In contrast, the generic method suggested by the ITU-R only managed to perform prediction with RMSE value and percentage errors of 28.3 and 28% accordingly. The new model was validated using data set from alternative years and alternate locations. In conclusion, the results from the research demonstrate a model that can be used in the tropical-equatorial region in a way denoting the achievement of fulfilling the stated objectives. The satellite signal performance can be improved by applying developed mitigation techniques with an economically viable cost where dependable fade margins can be attained. The newly developed frequency scaling technique can offer the right margin to achieve the required quality of service (QoS) for future SatCom in supporting near-future 5G communication. Consistent connectivity for high-speed broadband communication demand in delivering digital and internet applications during heavy precipitation can be attained.

Driving has become essential in transporting people from one place to another. However, prolonged driving could cause muscle fatigue, leading to drowsiness and microsleep. Electromyography (EMG) is an important type of electro-psychological signal that is used to measure electrical activity in muscles. This work classifies and predicts muscle fatigue from trapezius muscle of 10 healthy subjects. The EMG signals and the time when muscle fatigue was experienced by the subjects were recorded. The mean frequency and median frequency of the EMG signals were extracted. For classification of muscle fatigue in non-fatigue and fatigue condition, six machine learning models were used: Logistic Regression, Support Vector Machine, Naïve Bayes, k-nearest Neighbour, Decision Tree and Random Forest. From the value of median frequency and slope coefficient of median frequency, mathematical model was developed with respect to driver’s physical factors. The results show that both the median and mean frequency are lower when fatigue conditions exist. In term of the classification performance, the highest accuracy for classifying muscle fatigue due to prolonged driving was obtained by the Random Forest classifier with 85.00%, using both the median and mean frequency of the EMG signals. This method of using the mean and median frequency will be useful in classifying driver’s non-fatigue and fatigue conditions and predict muscle fatigue during prolonged driving. The significant factor influencing muscle fatigue of the driver was Body Mass Index (BMI). This study successfully developed mathematical model of second order polynomial of muscle fatigue and BMI (p<0.05 and the R2 = 0.85). The model was successfully validated where the residual errors compared between predicted values and actual values were less than 10%.

Hydrolysis has been identified as a rate-limiting stage in anaerobic-digestion. While it’s been widely used in biomethane production, biomethane only accounts for 50-60%. A Therefore, an integrated AD-MEC system was developed to increase the biomethane content using food waste. However, high electrode’s cost in the hybrid system poses an economical challenge to the market. Moreover, the microbial community plays a crucial role in the system, yet, minimal studies address the enhancement of microbial community and diversity. Hence, the characterization of food waste was performed in terms of carbohydrates, lipids, proteins, chemical oxygen demand, moisture content, solids, and volatile solids. The enzymatic hydrolysis of food waste was conducted to obtain the hydrolysate by one-factor-at-a-time (OFAT) through various factors including reaction time, temperature, enzyme loading, substrate concentration, and pH. The results showed that the optimum pH of 7, substrate concentration of 6%TS, the temperature of 50oC, and time of 16h gave the best release of reducing sugars. followed by the statistical optimization using faced centred central composite design (FCCCD) of selected factors, namely enzyme loading and substrate concentration,. The optimum conditions were enzyme loading of 6% (w/v) and a substrate concentration of 10% as the total solids (TS). Another pre-treatment, the acidic-enzymatic treatment using different concentrations of acids were performed. An acid concentration of 0.5% (v/v) showed the best hydrolysis effect achieving a value of 20 g/L reducing sugar,34.2% solids reduction, and 90 g/L soluble chemical oxygen demand (SCOD). However, the biogas production and free amino nitrogen release from acidic-enzymatic treated samples were lesser than only enzymatically treated samples. For MEC system, the effect of electrode modification using multiwall carbon nanotubes (MWCNT) and microbial growth into the electrodes was monitored using scanning electron microscope (SEM) images. The MWCNT growth was in-between the carbon felt fibres and the stainless steel mesh strands. The effectiveness of the electrodes was tested by inserting them into the hybrid system with glucose as the main substrate. Stainless steel mesh- modified cathode showed the highest biogas and methane production with a value of 14.4 ml CH4/g glucose. In addition, carbon-felt modified electrodes showed a maximum substrate degradation value of 93% and a current density of 4.5 mA/m2. The SEM imaging of the microbial growth on the electrodes showed that the microbes followed a different growth behaviour in modified and unmodified electrodes. In addition, MWCNT-modified Stainless steel mesh(SSTM) showed a potential hydrogenotrophic growth selectivity, unlike unmodified SSTM, which had a more syntrophic microbial community. Hybrid systems showed a higher hydrolysis efficiency especially modified systems, with a percentage of 39.4% by the 48th hour, followed by unmodified systems. The acidogenesis efficiency results showed that the hybrid systems were dominated by the acetic acid pathway, which is favourable in the hybrid system, unlike the conventional digester, which was dominated by a different pathway. Mixing the original inoculum obtained from a previous AD with cow manure has enhanced and increased the competitiveness of the microbial community. Thus, it was positively reflected on the biomethane production potential and rate, with a value of 38ml/g COD and 1.2 ml/h, respectively. In this study, we successfully enhanced the hydrolysis rate, improved the selectivity of microbes in the system, and introduced a set of commercially available electrodes. Our findings also provided compelling evidence that increasing microbial diversity significantly enhances the overall performance of the system.

In recent years, the advancements in deep learning and high-performance edge-computing systems have increased tremendously and have become the center of attention when it comes to the analyzing of video-based systems on the edge by making use of computer vision techniques. Intelligent Transportation Systems (ITS) is one area where deep learning can be used for several tasks including highway-based vehicle counting systems where by making use of computer vision techniques, an edge computing device and cameras installed in specific locations on the road, we are able to obtain very accurate vehicle counting results and replace the use of traditional and laborious hardware devices with modern low-cost solutions. This thesis proposes and implements a modern, compact and reliable vehicle counting system which is based on the most recent and popular object detection algorithm as of writing this thesis known as the YOLOv5, combined with a state-of-the-art object tracking algorithm known as DeepSORT. The YOLOv5 will be used in the following system for the detection and classification of four different classes of vehicles whereas DeepSORT will be used for the tracking of those vehicles across different frames in the video sequence. Finally, a unique and efficient vehicle counting method will be implemented and used for the counting of tracked vehicles across the highway scenes. A new highway vehicle dataset consisting of four vehicle classes namely: car, motorcycle, bus and truck were collected, cleaned and annotated with a total of 11,982 images which will be published in the following study and used for the training of our robust vehicle detection model. From the results observed over real-world highway surveillance data, the following system was able to obtain an average vehicle detection mAP score of 96.1% and a vehicle counting accuracy of 95.39%, all while being able to be deployed on a compact Nvidia Jetson Nano edge computing device with an average speed of 15 FPS which outperforms other previously proposed tools in terms of both accuracy and speed.

Nuclear fusion is one of the future solutions towards sustainable energy. The concept of the magnetic field is introduced to contain high-temperatureture reaction of the nuclear fusion. However, the presence of the magnetic field affects the fluid flow inside the blanket module, which purpose is to harness the energy, where it reduces the heat transfer efficiency in the channel of a heat exchanger. Thus, to counter this effect, this research examines the flow structure behind multiple bluff bodies arranged in tandem in a channel under the influence of a magnetic field with the aim of increasing the heat transfer efficiency inside the channel. In this study, the effect of gap ratio, G/h = [1-2.4] and Hartmann parameters, H = [0-80,0] are analyzed for the critical Reynolds number, pressure drop and Nusselt number using a Computational Fluid Dynamics open-source software. It is found that the presence of the downstream cylinder with gap ratio, G/h = 1.2, 1.4 and 1.6 improves the flow in terms of critical Reynolds number and Nusselt number while having some pressure drop penalty. The multiple cylinders increased the Nusselt number compared with the flow past a single cylinder. In terms of the Hartman parameter, increasing the value of the parameter increases the critical Reynolds number and decreases the Nusselt number.

One of the objectives and aspirations of scientists and engineers ever since the development of computers has been to interact naturally with machines. Hence features of artificial intelligence (AI) like natural language processing and natural language generation were developed. The field of AI that is thought to be expanding the fastest is interactive conversational systems. Numerous businesses have created various Virtual Personal Assistants (VPAs) using these technologies, including Apple's Siri, Amazon's Alexa, and Google Assistant, among others. While an ongoing effort to increase the friendliness and constancy of informal dialogue systems, most research focuses solely on simulating human-like replies, leaving the features of modeling interlocutors' awareness are unexplored. Meanwhile, cognitive science research reveals that awareness is a crucial indicator of a high-quality informal conversation. To precisely model understanding, Persona Perception (P2) Bot was developed using a transmitter-receiver-based structure. P2 Bot leverages mutual persona perception to improve the quality of customized dialogue generation. Even though many chatbots have been introduced through the years to diagnose or treat psychological disorders, we are yet to have a user-friendly chatbot available. This research aims on improving the quality of conversation generation by implementing the Generative Pre-trained Transformer-2 (GPT-2) model on P2 Bot. GPT-2 is a 1.5B parameter transformer model which produces state-of-the-art accuracy in a zero-shot setting on 7 out of 8 evaluated language modeling datasets. Observations on a large open-source dataset, PERSONA-CHAT, show that the technique is successful, with some improvement above state-of-the-art baselines in both automatic measures and human assessments. The model has achieved 82.2% accuracy on Hits@1(%) in the original data and 68.8% on the revised data. On the human evaluation, the model scored an average of 2.66 meaning the provided responses were coherent and informative. A smart generative cognitive behavioral therapy with spoken dialogue systems support was then developed using the model, which was then implemented using modern technologies in VPAs like voice recognition, Natural Language Understanding (NLU), and text-to-speech. This system is a magnificent device to help with voice-based systems because it can have therapeutic discussions with the users utilizing text and vocal interactive user experience.

Soil erosion is an ecological hazard that, if left unchecked, poses wider threats to the environment. These threats range from inconveniences such as the ruining of landscapes and the reduction of water quality to hazards such as floods and landslides. This thereby necessitates a method to monitor soil erosion, one of which is by monitoring the formation of river sediments. Computer vision techniques have matured in recent years and have been used in many different fields of applications. One form of computer vision technique is called "semantic segmentation," which is a technique that seeks to associate every pixel of an image with its own object class. This presented an opportunity where images of river sedimentation could be analysed and identified accurately to the pixel. In examining further the use of semantic segmentation for river sedimentation purposes, this project looked at three publicly available network architectures: Unet, Linknet, and Feature Pyramid Network (FPN). All these three networks belong to a type of architecture called fully convolutional networks. Three prediction models, one from each architecture, were trained and tested against 100 images of various river sediment formations along the course of the IIUM river. The images are divided into 75 images for training and 25 images for validation. Meanwhile, the model is assessed both quantitatively by Intesection over Union (IoU), and label predictions assessed qualitatively. After training, the sediment IoU scores obtained were as follows: 0.83446103 for Unet, 0.8188789 for Linknet, and 0.20392573 for FPN. The qualitative results outputted however were mixed: the architectures are able to perform somewhat well in identifying sedimentation when the formation of those sediments is uniform, with Unet performing the best, followed by Linknet and then FPN. However, all the architectures struggled in identifying the sediment when non-uniform sedimentation formations are present. One recommendation proposed is to add object classes to reduce intraclass differences and hopefully reduce class confusion by the prediction models. Another recommendation is to develop novel architectures that are able to accommodate intraclass differences while still producing accurate sediment identification.

This study aims to design, fabricate, and study the performance of a thermal cell absorber attached to a flat plate absorber collector. Flat Plate Solar Collector (FPSC) is widely used in agricultural products for drying applications. An investigation into the effect of design parameters on FPSC has been carried out. Flat Plate Base-Thermal Cell Absorber (FPBTCA) has been designed and fabricated based on the design parameters experiment results which are; absorber base materials (AL), absorber base thickness (0.5 mm), the air gap distance between glass and absorber base (10 mm), the air gap distance between glass 1 and glass 2 (0.4 mm), double glass, glass thickness (2.0 mm), thermal cell thickness (1.0 mm) and material (SS). The experiment was performed using a solar simulator with solar radiation of 700 W/m2. The solar simulator is used as the artificial sun, which is exposed to solar radiation for 300 seconds and without solar radiation also for 300 seconds. The heat transfer rate of the collector (Q ̇) and efficiency of the collector shows that stainless steel 1.0 mm with aluminum base absorber has a higher value which is 412 kJ, 18.21 kW, and 47.08 %, respectively. The performance of the outlet temperature in the drying chamber of stainless steel with an aluminum absorber has a higher value of energy gain which is 116.08 J at 300 seconds. Evaluation under outdoor conditions revealed that the FPBTCA has a lower temperature discharge rate as compared to the FPSC. FPBTCA also shows the highest heat absorption (Q ̇), which is 96079.37 J on 4 March 2021, then FPSC, which is 49187.07 J. The highest efficiency for FPBTCA at 360 minutes is 30.24 %, and for FPSC is 21.81 %. The efficiency of FPBTCA is consistent while the solar radiation is decreased at 120 minutes and 180 minutes. Mathematical modelling analysis proved that the error for energy balance is below 5%. FPBTCA has been introduced to enhance the thermal performance and efficiency of solar thermal collector systems. It also has higher heat storage capabilities during solar radiation drops when the weather is cloudy.

In the case of amputees, the development of cybernetic hands that closely resemble the functions of real hands is essential for comfort and functionality purposes. Controlled by intrinsic and extrinsic muscles, the human thumb plays a major role in differentiating hand gestures. For those who have lost their intrinsic hand muscles, any information about muscle activities that can be obtained from the extrinsic muscles is essential to control the thumb. Thus, focusing on transradial amputees, this research investigates the relationship between extrinsic muscles to characterise thumb posture. A High-Density surface Electromyogram (HD-sEMG) device and a portable thumb force measurement system were used to collect forearm HD-sEMG signals from a total of 17 subjects. For the flexion motion, the subjects were asked to repetitively place their thumb at rest before exerting 30% of their individual maximum voluntary contraction (MVC) on a load cell by following a designated trajectory presented on a developed graphical user interface (GUI). The measurement system was set to four different postures namely zero degrees, thirty degrees, sixty degrees, and ninety degrees. Feature extraction was then performed by extracting the absolute rectified value (ARV), root mean square (RMS), mean frequency (MNF) and median frequency (MDF) values of the forearm HD-sEMG signals before being classified using four different classifiers namely linear discriminant analysis (LDA), support vector machine (SVM), k-Nearest Neighbour (KNN), and TREE-based classifier. The results revealed that the LDA classified RMS and ARV-RMS features, which were extracted from both posterior and anterior hand sides successfully achieved the highest correctly classified percentage of 99.7%. The findings of the study are significant for the development of a dedicated model-based control framework for prosthesis hand development to be used by transradial amputees in the near future.

Electric vehicles are slowly becoming the first option for consumers. As time passes, the challenges limiting the chances of its existence are being reduced by the efforts of the researchers; they have the idea of having a green world and aim to achieve it by eliminating any source of pollution that found a green replacement. However, how well the performance of electric vehicles is not enough to make an industrial change. While there are challenges yet to be solved additionally, the main problem is the time to fully charge the battery in static and dynamic charging. However, time is rapidly being enhanced for static charging due to the fast charging cable improvements. Still, for dynamic charging, it is all about the efficiency of the connection during the power transmission. In this project, a homogenous coil has been designed, and the main focus is the radius of the coil. The design is tested under different conditions, and the optimal case is when the radius of the transmission coil is bigger than the radius of the receiver coil. The design has recorded an efficiency of up to 95%. Lastly, an enhanced method has been produced to wirelessly charge the electric vehicle with the help of homogeneous pads.

Automatic License Plate Recognition (ALPR) has become a common study area because of its many practical applications, such as automatic toll collection and traffic law enforcement. However, most existing methods of Malaysian ALPR are not robust enough to be used in everyday situations. The lack of high-quality benchmarked datasets that accurately represent real-world complexities in Malaysian license plates (LP) and the absence of a comprehensive dataset to demonstrate system robustness is a significant limitation. In addition, the reliance on shallow techniques in the studies on Malaysian ALPR causes inefficiency of the systems, particularly in handling complicated scenarios involving different image backgrounds and variations in LP size or shape and also the non-standard LPs. This dissertation presents a robust Malaysian ALPR system based on the single-shot detector You Only Look Once (YOLO) in two stages; license plate detection (LPD) and license plate recognition (LPR). The system is designed by evaluating and optimizing different models with different dataset optimization to achieve the best speed versus accuracy trade-off in ALPR system. The models are trained using a large-scale dataset containing images from several places around Malaysia, with the addition of data augmentation techniques to make them robust under various circumstances (e.g., with variations in lighting, camera position and settings, and license plate types). A dataset augmentation has also been accomplished by systematically generating a large, controlled synthetic dataset. The purpose is to achieve a balanced dataset and ensure the robustness of the dataset in terms of variations that exist in Malaysian license plates in the form of non-standardized license plates and special license plates. Thus, this work introduces a dataset for Malaysian ALPR with more than 176,000 images from real-world scenarios and synthetically produced images covering various aspects. This dataset will be public to the research community. The name of this dataset is Malaysian Number Plate and in short (MYNO). This dataset can be used for further training and evaluation of ALPR models. A separate challenging dataset is created for testing the models. Many experiments are carried out in detail with different models, data size, number of epochs, and real and synthetic datasets. When adding the synthetic dataset, the system performed better with 97.6% mAP compared to 85.5% mAP for the only real-world dataset at the same number of epochs. The proposed system achieved a recognition rate of 98.1% mAP on a real-world dataset collected from different toll plazas around Malaysia containing comprehensive environment distinctions with over 50 thousand labeled images. The system was tested on a challenging test dataset with low visibility and an unconstrained environment, resulting in 95.96% end-to-end accuracy. The results demonstrate the significance of incorporating synthetic datasets into the training process for improved performance in ALPR systems. The inclusion of a synthetic dataset led to a substantial increase in mean average precision (mAP), with a notable improvement of 12.1% when combined with the synthetic dataset. The system showcases its effectiveness in handling diverse environmental conditions by achieving 98.1% mAP on a real-world dataset collected from various toll plazas in Malaysia. In addition, achieving an impressive end-to-end accuracy of 95.96% despite low visibility further validates the system’s performance on challenging dataset.

An efficient solar collector design that effectively absorbs solar energy and converts it into heat is required during intermittent solar radiation. Existing Flat Plate Collector (FPC) and Heat-Pipe Evacuated Tube Collector (HP ETC) are designed for water heating required storage tanks, while an additional heat exchanger is required for air heating application which leads to extra spacing and cost. The installation also needs to be tilted at the correct angle and positioned to south/north facing to optimize the system performance. These could lead to design limitations. Therefore, this study aims to design, develop and investigate the thermal performance of an Evacuated Glass-Thermal Absorber Tube Collector (EGATC) for air heating applications. EGATC was designed from a conventional HP ETC, and the performance was compared through parameter and performance experimental setup. The three days performance experiments showed EGATC performed better with daily outlet temperature increased by 9.0%, 7.2%, and 4.9%, respectively, with an average of 7.0% compared with HP ETC. EGATC also had greater efficiency compared to HP ETC, with the average efficiency for EGATC being 51.3% compared with HP ETC's 41.8%. EGATC's inner absorber was designed to create a double pass flow with the ventilated chamber. The parameter experiment shows the design could increase the outlet temperature by a difference of 6.3% (for stainless-steel inner absorber compared with insulation material inner absorber). Regarding energy storage, the stainless-steel inner absorber also had an advantage compared to the insulation material inner absorber, with a 1.3% difference. On the effect of other parameters such as inner absorber surface area air contact (perforated fin), outer absorber selective coating surface, outer absorber wall thickness, double layer non-vacuum glass tube, single layer transparent outer glass tube, and single-layer thin film inner glass tube was investigated by parameter experimental setup on energy storage. It was proven that the outlet temperature, energy store, and energy buffer could be enhanced with the combination of wind speed 0.9 m/s, zero (0) perforated fin, non-coating outer absorber, and 1 mm outer absorber wall thickness. It was also reported that double-layer vacuum glass tubes promise better thermal performance enhancement compared with double-layer non-vacuum glass tubes, single-layer transparent outer glass tubes, and single-layer thin film inner glass tubes. The mathematical equation of each EGATC component was formulated based on the first law of thermodynamics. The total acceptable error of 5% shows that the model at each node was valid. The performance curves for those 0 fins (equation), 0 fin (experimental), and 3 fins (experimental) were obtained. The results showed that the efficiency (collector + storage) was affected by the number of fins. The efficiency (collector + storage) was 68.7%, 71.2%, and 71.0%, respectively. In conclusion, the application of EGATC in air heating applications proved beneficial to the application of solar drying processes, especially in equatorial climate countries such as Malaysia.

Autism Spectrum Disorder (ASD) in children can be observed by difficulties with social, executive dysfunction, communication and sensory issues. Although there are various well-established interventions for ASD children, most of them are done manually by human therapists and parents which require attention to details and time-consuming. The objective of this research is to develop Human-Robot Interaction (HRI) platform using partial application of Applied Behavioral Analysis (ABA) techniques for early intervention programmes of ASD children. The ABA techniques were then converted to Finite State Machine (FSM) and were programmed in Python 3 (Linux environment) to enable future researchers for HRI to implement the correct technique in lesser time. This framework is then evaluated on ASDC. Four interaction modules with predictable and consistent teaching structures were developed. For implementing the robot modules in this work, the FSM approach were used during the integration of all three components of interaction modules: (a)learning content (developed with ABA therapist); (b)rigid teaching structures; (c)robot’s response. Interaction modules, proof-of-concept and final framework validation were done by experienced ABA therapist. The modules were converted to FA where it enables future researchers to iterate the correct method more effectively and to lessen the time-consumed in developing HRI for ASDC. Efficiency of the system were tested with 20 ASD children recruited from IDEAS. 95% of the subjects managed to finish the modules. From the 19 subjects, 53% subjects showed excitement over meeting the robot, and the remaining 47% showed no expression. While they showed indifference over meeting the robot, their interest started to peak once reinforcement (video reward - cartoons) were shown on the monitor. Out of the 19 subjects, only 16% showed excitement throughout the session while the others started to get restless by the end of the goodbye module. The last subject (9 years old boy, male) was not able to finish the modules as he showed clear signs of irritability. While the subjects showed varying levels of emotions during the interaction, and some were not able to finish the modules, the preliminary results obtained can contribute to better understand factors that might help determine sub-groups of children with ASD for whom robots could be particularly useful, and the structure they need. In summary, two objectives involving development of HRI framework for ASDC were fully achieved while a mixed results on effectiveness of the system were obtained when evaluating efficiency of the system.

The 5th generation of cellular communication is a highly competitive market that promises some distinct features compared to the legacy LTE (long-term evolution) era. Some of these exclusive features include enhanced mobile broadband (eMBB), massive machine type communications (mMTC), and ultra-reliable low latency communication (URLLC) traffic (1 ms one-way latency, 99.999% reliability). Enabling these cuttingedge technologies requires a very smooth processing of user data and transmitted signals. Channel estimation and acquisition of channel state data is one of the important points in this regard because they can eventually enable signal transmission and subsequent processing. However, most of the estimators in research nowadays suffer from high complexity due to either too many constraints or conditions for unique solutions. This is already a significant problem in the communication industry because of its high dependency on resource allocation and system overhead. This research focuses on the enhancement of the legacy channel estimation processes to fit the 5G cellular standards. The industry standard Least Squares (LS) estimator was used as the basis for the optimized estimation. A dual residual function was enabled instead of a single one to make the estimator adaptive. Results show that making the weight function adaptive reduces the error at the receiver and provides a sharper response curve. A comprehensive study was carried out against the trending compressed sensing (CS) based semi-blind estimators as a second objective. And finally, the optimized algorithm was characterized on MIMO-OFDM systems to show its performance improvements in the cases of large arrays. These objectives were carried out through simulation, and results were constructively discussed based on the earlier points. Results were compared with parameters SNR, SER, PER, and BER. Some potentials regarding the channel estimation in MIMO-OFDM were left as pick-up points for future research interests.

A ground network of communications satellite system is typically made up of Earth Station(s), Mission Operations Center (MOC), Science Operations Center (SOC), and the supporting infrastructure that connects them all. A ground network grows as more Earth Stations are added, which requires additional considerations to ensure that the MOC can communicate with all the Earth Stations in the network. It also requires continuous upgrade to provide a better reliability for a better performance. The improvements of the reliability of a Geostationary satellite control Earth Station system can be accomplished via redundancies of the subsystems, multiple testing in the planning stage and selection of only the best components for its subsystems. Suitable maintenance activities from time to time also play an important role to prevent the cost blow out and any unwanted failures. Hence, the development of a new reliability model based on identified factors that caused calamity to the system was the main objective of this research. In addition, this research also aims to develop an operational cost model along with the suitable maintenance activities to enhance the robustness of the geostationary satellite control Earth Station system. The models were designed by applying Monte Carlo from MATLAB software. The reliability and cost data that were used for simulations was obtained from MEASAT. Based on the previous studies, configurations with more redundancies in the subsystem can affect the reliability performance, which can decrease the failure rate. At the end of this research, a new reliability model of an Earth Station system which was compared against 2-parallel, 3-parallel, and 4-parallel configurations within the range of affordability (operational cost model) along with the suitable maintenance activities were proposed to enhance the robustness of the geostationary satellite control Earth Station system. The three elements consisting of the reliability model, suitable maintenance activities as well as the operational cost model were integrated together creating a sustainable framework. The obtained results showed that an Earth Station that was configured with the 2-parallel configuration provided the cheapest and optimum reliability system performance even though the 3-parallel and the 4-parallel configurations provided higher reliability. Consequently, the sustainable framework encompassing reliability and cost elements were modelled based on the 2-parallel configuration together with the proposed maintenance activities. Furthermore, root mean square (RMS) values were also calculated for both the reliability and the operational cost models. The results demonstrated that the calculated RMS values for both new reliability and new operational cost models produced the smallest values of 20.84% and 22.82% respectively. Therefore, the calculated RMS values for both reliability and operational cost models showed that the 2-parallel configuration fit to be applied in the Earth Station system design which contributes to the system design with acceptable reliability and most affordable cost.

This research aims to solve some problems related to blockchain peer-to-peer energy trading, which are purchased energy under consumption and transaction authentication and demand response control, which still rely on centralized schemes. The purpose of this thesis is to control and authenticate a fully Decentralized pay-per-use energy trading platform. It presents a blockchain-based peer-to-peer (P2P) energy trading platform where prosumers can trade energy autonomously and without interference from a central authority. Multiple prosumers can collaborate on energy generation to form a single supplier. Customers' electricity consumption is monitored via a smart meter connected to an IoT node connected to a private blockchain network. Smart contracts invoked on the blockchain enable autonomous trading interactions between parties and govern the behaviour of accounts within the Ethereum state. The decentralized P2P trading platform uses autonomous usage-based billing and energy routing monitored by a smart contract. A Deep Learning-based Gated Recurrent Unit (GRU) model predicts future consumption based on past data collected on the blockchain. The predictions are then used to set Time of Use (TOU) ranges using the K-means cluster. The data used to train the GRU model is shared among all parties within the network, making the predictions transparent and verifiable. By implementing K-Mean clustering in a smart contract on the blockchain, the set of TOU is independent and unchallengeable, To ensure the validity of the data uploaded to the blockchain, a consensus algorithm is proposed to detect fraudulent nodes, along with a Proof of Location (POL) to ensure that the data is uploaded by the expected nodes. To address the conflict of interest between prosumers and distribution system operators (DSOs) in decentralized P2P energy trading platforms, where prosumers seek to maximize their profit on the one hand, while DSOs seek optimal power flow (OPF) on the other hand, a novel fully decentralized architecture is proposed for an OPF-based demand response management system that uses smart contracts to force generators into compliance without the need for a central authority or hardware. The study details the proposed platform architecture, operation, and implementation. The results are presented mainly in terms of gas consumption of smart contracts and transaction latency for different loads. The work presented in the thesis is relevant in the sense that we have attempted to address a popular and important contemporary research issue related to blockchain technology. The research work holds great potential for industrial use. In the future era of renewable energy, decentralization of energy trading is a necessity for the future society.