FYP Titles for Academic Session 2022/2023-Electronic Programme

No FYP Titles Synopsis/Requirement Student Supervisor

1- An improvement of PAPR The transceiver combination technology, of orthogonal frequency division NUR LIYANA
Reduction Using Fireworks multiplexing (OFDM)), provides a viable alternative to enhance the quality of BINTI MAT
Search Optimization Algorithm in service and simultaneously to achieve high spectral efficiency and data rate for FUZI (149787)
OFDM Systems wireless mobile communication systems. However, the high peak-to-average
power ratio (PAPR) is the main concern that should be taken into consideration
in the MIMO-OFDM system. Partial transmit sequences (PTSs) is a promising
scheme and straightforward method, able to achieve an effective PAPR
reduction performance, but it requires an exhaustive search to find the
optimum phase factors, which causes high computational complexity
increased with the number of subblocks. In this paper, a reduced
computational complexity PTS scheme is proposed, based on a novel swarm Dr.
intelligence algorithm, called fireworks algorithm (FWA). Aeizaal Azman
2- PAPR Reduction using Chicken Computational complexity considers as challenging issue in conventional Abdul Wahab
Swarm Optimization (CSO) with Partial Transmit Sequence (PTS) scheme, even though it has good
Partial Transmit Sequences in performance to reduce PAPR. In this research, one of the recently developed
OFDM System bio-inspired algorithm named Chicken Swarm Optimization (CSO) is
implemented with the combination of Partial Transmit Sequences to reduce
PAPR, one of the most featured problems in OFDM system.
3- PAPR Reduction using PTS-PSO In this project, it is proposed that a particle swarm optimization (PSO) based
Technique for OFDM Systems partial transmit sequence (PTS) technique is used so that least Peak-to-
Average Power Ratio(PAPR) is achieved in Orthogonal Frequency Division
Multiplexing Systems (OFDM). Our approach is to apply PSO based PTS on
each antenna of the system helping to find the optimal phase factors, which is
a straight format on how to achieve minimum PAPR in this system.
4- Deep Learning-based Machine The inspection process is essential in an assembly line of the manufacturing GRACE TAN
Vision Inspection Technique for industry as it results in the discovery of many hidden defects in products. YIH CHIN
Defect Detection Manual inspection and conventional machine vision methods are often used (147853)
over the years. However, recent state-of-the-art deep learning methods in the
vision inspection system have been extensively studied. Unlike traditional
machine vision, deep learning technology uses powerful algorithms such as
clustering algorithms, convolutional neural networks, and many more, imitating
human intelligence to identify and distinguish anomalies or abnormalities while
tolerating variations. By combining machine vision inspection systems and
 deep learning, hybrid approaches can offer better performance in terms of
higher accuracy and versatility. This project aims to evaluate the performance
of selected deep learning algorithms in detecting and identifying all possible
cosmetic defects, including scratches, dents, and black marks of an industrial
product. Industrial data will be provided.
Requirement: Python programming
References:
1. J. Yang, S. Li, Z. Wang and G. Yang, "Real-Time Tiny Part Defect Detection
System in Manufacturing Using Deep Learning," in IEEE Access, vol. 7, pp.
89278-89291, 2019, doi: 10.1109/ACCESS.2019.2925561.
2. S. Ren, K. He, R. Girshick, and J. Sun. 2016. Faster R-CNN: Towards Real- Dr.
Time Object Detection with Region Proposal Networks. IEEE transactions on Muhammad
Pattern Analysis Recognition (CVPR). pp. 2818-2826. Firdaus Bin
3. Chuanxia Jian, Jian Gao, Yinhui Ao, Automatic surface defect detection for Akbar Jalaludin
mobile phone screen glass based on machine vision, Applied Soft Computing, Khan
Volume 52, 2017, Pages 348-358, ISSN 1568-4946.
5- Corrosion Under Insulation Corrosion under coatings is described as localized corrosion that forms LEE REN KAI
Assessment using AI-driven because of moisture penetration through the coating materials or due to (146535)
Microwave Technique contaminants' presence within the coating material. The traditional inspection
techniques operating at a low frequency require removing coating material to
enable inspection, due to poor signal penetration. Several high-frequency
inspection techniques such as the microwave technique have shown
successful inspection in detecting the undercoatings corrosion, without
removing the coatings. However, the prediction formulation of undercoating
corrosion severity is still challenging due to the significant presence of outliers
in the inspection data. Thus, an artificial intelligence (AI)-based microwave
technique for undercoatings corrosion inspection is proposed. The proposed
project's working principle is based on recording and analyzing the reflected
wave from the microwave probe, scanned over the insulated sample using a
selected clustering algorithm to assess the undercoating corrosion's severity.
Requirement: In campus FYP, lab measurement, Python/MATLAB
programming References:
1. N. H. M. M. Shrifan, M. F. Akbar, and N. A. M. Isa, ‘Prospect of Using Artificial
Intelligence for Microwave Nondestructive Testing Technique: A Review’, IEEE
Access, vol. 7, pp. 110628–110650, 2019.
 2. N. H. M. M. Shrifan, G. N. Jawad, N. A. M. Isa, and M. F. Akbar, ‘Microwave
Nondestructive Testing for Defect Detection in Composites Based on K-Means
Clustering Algorithm’, IEEE Access, vol. 9, pp. 4820–4828, 2021.
3. N. H. M. M. Shrifan, M. F. Akbar, and N. A. Mat Isa, ‘Maximal overlap discrete
wavelet-packet transform aided microwave nondestructive testing’, NDT E Int.,
vol. 119, p. 102414, Apr. 2021.
6- Material Characterisation and Microwave imaging is a promising technique for dielectric composite TINESH A/L
Composite Testing Using evaluation. Their most significant advantage is their relatively high penetration SELVAM
Microwave Technique depth. Another feature worth noting is that traditional methods could not (148151)
acquire internal content with such a low impact on both the sample and Dr.
surrounding environment, including the test operator, compared to other Muhammad
techniques. This project proposed microwave non-destructive and noninvasive Firdaus Bin
methods for quality evaluation of layered composite materials using an open- Akbar Jalaludin
ended waveguide probe. Pure reflection coefficient parameters only Khan
exceptionally give a clear answer about the location of material cracks.
Therefore, this makes it necessary to analyze these parameters
simultaneously along with several other factors, such as stand-off distance,
probe type, or wave polarization. The purpose of the work was to find the
dependency between the physical state of a layered composite powerplant
pipeline and the S-matrix parameters response (reflection and transmission
parameters).
Requirement: In campus FYP, lab measurement, Python/MATLAB
programming
Reference:
1. Teng,W.S.; Muhammad Firdaus Akbar; Nihad Jawad, G.; Tan, S.Y.; Mohd
Sazali, M.I.S. A Past, Present, and Prospective Review on Microwave
Nondestructive Evaluation of Composite Coatings. Coatings 2021, 11, 913.
2. Ghassan Nihad Jawad, Muhammad Firdaus Akbar, 2021, IFFT-based
Microwave Non-Destructive Testing for Delamination Detection and Thickness
Estimation, IEEE Access, vol. 8, pp 98561 - 98572.
3. Muhammad Firdaus Akbar, Ghassan Nihad Jawad, Laith Danoon Rashid,
Robin Sloan, 2020, Nondestructive Evaluation of Coatings Delamination Using
Microwave Time Domain Reflectometry Technique, IEEE Access, vol. 8, pp
114833 - 114841.
7- 5G Radiation Absorber: Optimal This project involves designing a Hollow Pyramidal Shape 5G Radiation
Dr.
Material Composition and Absorber using rubber and carbon composite material and determining the
Mohd Nazri
Thickness of Hollow Pyramidal optimal configuration of its material composition and thickness using Response
Mahmud
Absorber Surface Methodology(RSM).
8- 5G Radiation Absorber: Robust This project involves designing a Hollow Pyramidal Shape 5G Radiation
Parameter Design of Hollow Absorber using rubber and carbon composite material and determining the
Pyramidal Absorber parameter settings that result in robust performance.
9- 5G Radiation Absorber: This project involves characterising the strength and durability of Hollow
Characterisation of Strength and Pyramidal Shape 5G Radiation Absorber made of rubber and carbon
Durability of Hollow Pyramidal composite material in relation to its material composition and dimensional
Absorber parameters using Design of Experiment(DoE)
10- Solar powered IOT-based Technologies and IoT have the potential to transform agriculture in many HOR XIAN EN
livestock monitoring system aspects. By using smart devices, you can monitor the location, temperature, (145792)
blood pressure and heart rate of animals and wirelessly send the data in real-
time to farmers’ devices. This allows farmers to check in on the health and
location of each individual animal in their herd from anywhere as well as
receive alerts if something falls outside of the normal range. Student is
expected to build an IOT system to monitor livestock that is solar powered.
Experience with Arduino/Raspberry Pi/ESP32 is highly preferred.
11- Digital dashboard: IOT-based Technologies and IoT have the potential to transform agriculture in many WONG YU HAO
livestock management system aspects. By using smart devices, you can monitor the location, temperature, (148514)
Dr.
blood pressure and heart rate of animals and wirelessly send the data in real-
Intan Sorfina
time to farmers’ devices. This allows farmers to check in on the health and
Binti Zainal
location of each individual animal in their herd from anywhere as well as
Abidin
receive alerts if something falls outside of the normal range. Student is
expected to build a digital dashboard to perform livestock herd management
and perform data analytics. Knowledge/expertise in website/webpage
developing skills is highly preferred.
12- Case-study: IOT-based Smart farming using IOT technology is increasingly popular nowadays. For this MUHAMAD
monitoring system using project only soil moisture sensor is used in the monitoring system. Student is ZUHAIRI BIN
Lorawan for community farming expected to build an IOT system with Lorawan technology to connect multiple BASARI
soil moisture sensors throughout community farming. Student is expected to (148309)
have transport to travel in Seberang Perai to deploy technology in real-life case
study.
13- Optimization of microstrip Swarm intelligence algorithms that can be used to optimize antenna LIM CHEAH
antenna array using swarm parameters efficiently by imitating the collective behaviour of various swarms PING (144016)
intelligence algorithm or groups in nature. Student is expected to simulate and optimize a microstrip
antenna array at mmWave frequency band (24 GHz) using swarm intelligence
algorithm. Knowledge on Matlab, CST or ADS software needed. Also
knowledge on fundamentals of RF Engineering required. Must take the
Antenna and Propagation course in 4th year Semester 1.

17- Finger vein presentation attack Recent studies have shown that finger vein biometrics is vulnerable to TEH VEN TING
detection using deep learning presentation attack, i.e. printed versions of authorized individual's finger veins (147289)
and Taguchi method could be used to gain access to facilities or services. In this project the student
needs to develop a compact deep learning-based algorithm that can detect the
presentation attack of finger vein. To design a compact deep learning algorithm
the student needs to utilize EffNet block and Taguchi Method.
18- Color classification of solid wood Solid wood panels are the main components of various products in the wooden TAN JU WEI
panels based on deep learning furniture industry.The panels always have to be spliced together to construct a (149058)
larger board. To improve the quality of the product, panels with a similar color
are preferred to be spliced together. In this project, the student needs to
develop a compact deep learning architecture that can perform color
classification of solid wood panels from the captured images by the machine AP. Dr.
vision system. To design a compact deep learning architecture, the student Bakhtiar Affendi
needs to utilize Taguchi Method. (This project is in collaboration with Rosdi
Visionlytics Sdn Bhd)
19- Universal barcode localization Nowadays, barcodes have become an essential part of our lives with the STEPHEN
using YOLOv7 algorithm development of digital technology. They are used in supermarkets, libraries, WONG WEI
storehouses, etc. Barcode can be simply classified into 1D barcode and 2D FENG
barcode according to the dimension of the pattern. For example, the European (145818)
Article Number (EAN) and Quick Response (QR) code are common 1D and
2D barcode respectively. In this project the student needs to develop a
YOLOv7-based algorithm to finely detect and classify multi-class barcode in
complex environment. The developed algorithm needs to be integrated with
Dynamsoft Barcode Reader software for decoding purpose. (This project is in
collaboration with Western Digital)
20- Detection of solid wood defect Surface defect detection is an essential process to control the quality of solid SUI YUN HEN
using an improved YOLOv7 wood panel. The traditional way of identfying the defect relies on the human (146928)
algorithm visual inspection. However, this approach will be affected by some subjective AP. Dr.
factors or misjudgements of inspectors, and accuracy and efficiency are not Bakhtiar Affendi
sufficiently stable. In this project the student needs to develop a YOLOv7- Rosdi
based deep learning algorithm that can detect the surface defect of solid wood
panel. To increase the accuracy and reduce the inference time of deep learning
algorithm, the student needs to modify the existing YOLOv7 algorithm. (This
project is in collaboration with Visionlytics Sdn Bhd)
21- Image classifications for multiple Synopsis: In this research, the goal is to develop a machine vision technique LIM BOON HOU
objects for a humanoid vision system to classify multiple objects in a single image using (145592)
an Intel RealSense Depth camera. The developed system must be able to
classify types of materials, and 3D coordinates of the objects with respect to
the camera in a sufficiently fast manner. (This project is part of USM-WD’s
collaborative research work).
Requirements: Good mathematical background and skills in programming
(preferably C++ and/or Python). Willing to learn new things. The camera will
be provided. AP. Dr. Nur
22- Direction sensing for robotic Synopsis: In this research, the goal is to develop a machine learning technique TAN YEONG Syazreen
vision for robotic vision to classify the moving directions of multiple objects in a single MING (145526) Ahmad
view using a commercially off-the-shelf (COTS) intelligent camera.
Requirements: Good mathematical background and skills in programming
(preferably C++ and/or Python). Willing to learn new things. The camera will
be provided.
24- Modeling of unstable dynamic Synopsis: The first part of this project involves modelling and simulation of an MUHAMAD
systems unstable dynamic system using MATLAB/Simulink. A data-driven approach will HUSAINI
then be used to optimize the parameters of the model. Software involved: ASYRAF BIN
MATLAB and NI LabVIEW. Requirements: Good mathematical background ADNAN
and skills in programming (preferably LabVIEW and MATLAB). Willing to learn (149435)
new things.

25- Development of a Mobile-based [Industry: Emeric Bike Solution] Active mobility is a mode of active AVITRA
Application for Green Active transportation for commuting between places through non-motorized means ANASTASIA
Mobility System but based around human physical activity. One of the best forms of active USUN
 mobility is cycling. There are many benefits of active mobility such as increase ANANTHAN
physical fitness, increase mental health, reduce obesity, reduce fossil fuel (149301)
consumption, reduce greenhouse gas emissions, decrease traffic and urban
congestion, and so on. Hence, providing good system for active mobility will
effectively promotes and increases the usage of this type of transportation. In
this project, the student is required to develop a mobile application (App) for
enhancing the efficiency and effectiveness this green active mobility. The app
system should provide user-friendly mode, secure data, promote awareness,
sales, and services for bicycles like folding bike, mountain bike (MTB), electric- AP. Dr. Haidi
bike (e-bike). This project will collaborate with a bike company, i.e., Emeric Ibrahim
Bike Solution and will test the App with their customers to get the feedbacks
and bugs. [Weekly meeting is expected.]
26- Development of Obstacle [Industry: Sky Realm Sdn. Bhd.] Currently, research related to autonomous NG HARN
Avoidance System for vehicles has been carried out vigorously. To ensure the safety of the vehicle TUNG (147261)
Autonomous Vehicle from Video and its passengers, the system needs to be able to detect any obstacles in
Input front of the vehicle, so that this can prevent collisions from happening. In this
project, the student will be provided with a few video sequences. Student
needs to develop an obstacle detection algorithm based on the input from this
video. The algorithm should be suitable for the implementation in embedded
system, has high accuracy, and also has a fast processing speed. [*Allowance
in the form of honorarium will be given to the student of this title.] [Weekly
meeting is expected.]
27- Development of Edge Detection Image segmentation is usually used for medical images. Generally, image MUHAMAD
Method for Medical Images segmentation can be divided into two types; region-based and edge-based. In NAZRUL
this project, edge-based segmentation method will be developed. The HAMIZAN BIN
developed method should produce good accuracy, and fast. [Weekly meeting MOHD YUSOFF
is expected.] (147333)
28- More than SPICE – Nonlinear This project works on developing a novel circuit simulation algorithm that is far THAM ZHI
capacitor. superior to the commonly used SPICE. This project will investigate nonlinear KEAN (147468)
capacitors and implement their simulation algorithms. Project involves Dr.
programming in Matlab. Good programming skill is recommended. Patrick Goh
29- More than SPICE – Nonlinear This project works on developing a novel circuit simulation algorithm that is far CHIN HONG Kuan Lye
resistor & inductor. superior to the commonly used SPICE. This project will investigate nonlinear KAI (148004)
resistors and inductors and implement their simulation algorithms. Project
involves programming in Matlab. Good programming skill is recommended.
30- Artificial intelligence for defects This project will apply artificial intelligence (AI) to identify defects on integrated
detection on integrated circuit circuit chips. Deep neural network techniques will be explored. The project is
chips using deep neural a collaboration with the industry and test data will be provided by Ideal Vision WONG NGEI
networks. Integration Sdn. Bhd. Project involves programming in Matlab and Python. HONG (147312)
Good programming skill is recommended.
31- Deep learning based artificial This project will apply deep learning techniques to model electronic devices LIM JUN LIANG
intelligence modeling of and circuits. Student will learn the basics of AI, methods to train the AI, and (145661)
electronic devices and circuits. apply the AI to analyze high-speed circuit behavior. Project involves
programming in Matlab. Knowledge of microwave circuits is beneficial.
32- Evaluation of fixed charges in SYNOPSIS: High resistivity silicon is increasingly becoming one of the subjects KEE JIUNN
physically-deposited aluminium of interest in optimizing the performance of MIS photodetectors as it offers LIANG (147202)
nitride on high resistivity silicon better bias responses compared to low resistivity silicon. The incorporation of
utilizing metal-insulator- thin AlN as the tunnelling insulator in MIS structure utilizing high resistivity
semiconductor (MIS) structures. silicon has shown promising photocurrent to dark current ratios, suggesting
potential integration of the structure. In this work, MIS structure on high
resistivity silicon with AlN tunnelling insulator will be simulated and empirically
modelled. The presence of positive fixed charges in the physically-deposited
AlN will be incorporated in the simulation model and the corresponding current-
voltage characteristics will be evaluated. REQUIREMENTS, Relevant courses:
EEE133 Electronic Devices and Circuits, EEE241 Analogue Electronics I,
EEE344 VLSI system and EEE348 Introduction to IC design. Software: TCAD
tools
33- Evaluation of metal work SYNOPSIS: High resistivity silicon is increasingly becoming one of the subjects GOH KANG
functions effect on the of interest in optimizing the performance of MIS photodetectors as it offers HOW (148087)
photoelectric properties of metal- better bias responses compared to low resistivity silicon. The incorporation of Dr.
insulator-semiconductor (MIS) thin AlN as the tunnelling insulator in MIS structure utilizing high resistivity Nur Zatil 'Ismah
structures on high resistivity silicon has shown promising photocurrent to dark current ratios, suggesting Binti Hashim
silicon with aluminum nitride potential integration of the structure. In this work, MIS structure on high
tunneling insulator. resistivity silicon with AlN tunnelling insulator will be simulated. The effects of
metal work functions of the metal contacts used in the MIS structures on the
current-voltage characteristics will be evaluated. REQUIREMENTS, Relevant
courses: EEE133 Electronic Devices and Circuits, EEE241 Analogue
Electronics I, EEE344 VLSI system and EEE348 Introduction to IC design
Software: TCAD tools
34- FPGA-based Implementation for SYNOPSIS: Often times, face recognition involves large-scale data processing Law Xin Yi
Face Recognition System. for high recognition accuracy. As such, computation time is one of the
important factors affecting the performance of a face recognition system. Due
(MOHE INTEL ELITE to the sequentially-executed nature of software codes, software-implemented
INTERNSHIP PROGRAMME face recognition algorithms are observed to have low processing speed, low
only) energy efficiency and require longer execution time compared to hardware-
accelerated implementation. In this work, a face recognition system based on
LBPH algorithm will be implemented on FPGA. This work contributes to
speedup improvement of the LBPH face recognition system by implementing
the model on an FPGA with parallel processing function.
35- Vibration-Based Fault Detection Student should possess knowledge in fabricating PCB and familiar with DINAGARAN
Dr.
in Drone using MEMS Sensors arduino device particularly in wireless communication. Activities for this project A/L
Wan Mohd
and Camera are 1) Test vibration signals of the drone using MEMS vibration sensors and TAMILSELVAM
Yusof Rahiman
camera in ground and about to take off mode 2) Analyze the vibration signals (148835)
Wan Abdul Aziz
obtained and determined the performance.
36- Electrical Power Generation The aim of this project is to use the energy wasted by the vehicles while CHAN ZHAO
using Speed breaker crossing a speed breaker & convert it into electricity for powering the electrical MING (148462)
load (Street Light). The mechanism below the speed breaker is connected with
gears & flywheel that rotates when a vehicle crosses. The flywheel is
connected to a generator to power up the loads.
37- Eye Ball Control of Wheel Chair The aim of this project is to control the steering of a wheelchair using your LIEW POH YEN
for Physical Handicapped person eyeball movement. The challenging part of this project is the detection of (146004)
eyeball movement. Image processing by acquiring images through a camera
or an IR sensor placed on custom-built glasses can detect such movement. An
Ir. Dr. Gomesh
Arduino is used for processing the data & controls the DC motors inside the
Nair A/L Shas
chair accordingly. The forward & backward movement with speed can be
controlled with an accelerometer placed in the helmet.
38- Device Load Monitor With The concept system is designed in order to calculate the energy consumption YOUNG SI
Programmable Meter For Energy of a load in units and the cost incurred over based on TNB Rating in ZHENG
Audit accordance to the number of hours that the load is to be used. An energy audit (149069)
is a power consumption survey of the amount of units consumed by a particular
device and the cost incurred over it to determine if the device is economical
use in domestic or industrial sectors. Since electricity cost per unit has
increased a lot these days, energy audits for devices have gained importance.
The project consists of an instant energy auditing system using
 ardino/microcontroller/raspberry pi that is able to calculate the consumed units
and cost incurred by device use over given time. The input data including
energy tariff and the number of hours device is to be used is fed by user into
the system. Based on this data the system instantly displays the cost and units
consumed within no time.
39- Development of real time In this project student need to study first the basic of MIMO algorithm i.e. by
Alamouti MMO Algorithm of implementing the Alamouti algorithm using simulation tool like MATLAB with
various channel matrix sizes various channel matrix sizes. Then, student needs to implement the algorithm
using WARP kit board. using tge WARP board. This board/kit utilizes a high performance FPGA, with
two flexible RF interfaces and other peripherals devices for rapid wireless
system designs. Course Requirement: EEE123 (C++), EEE320
(Microcontroller), EEE332 (Communication), EEE377 (Digital
Communication).
40 - Wireless Sensor Network’s using In this project, student is required to develop a simplified CDMA access MUHAMMAD
a Simplified CDMA Access protocol and design a simple wireless sensor network with a choice of network AZRUL BIN AP. Dr.
Protocol Based Wireless topology. The final implementation should be able to monitor the conditions of OTHMAN Mohd Fadzli
Communication for Grit System the grit system agriculture field. Course Requirements:EEE123 (C++), EEE320 (144391) Mohd Salleh
Agricultural Monitoring. (Microcontroller), EEE332 (Communication), EEE377 (Digital
Communication), EEE449 (Computer Network)
41- Development and In this project, the student is required to design a Quadrature Amplitude
Implementation of various QAM Modulation (QAM) for M=16, 32, 64 etc signals. First, the design should be in
Generated Signals schemes simulation software (MATLAB/MULTISIM). Then, the designed transmitter
using FPGA for Baseband Digital should be implemented in ALTERA FPGA board. Integration with a RF front
Data Transmission end circuit can be included if the time permit. Course Requirements: EEE123
(C++), EEE320 (Microcontroller), EEE332 (Communication), EEE377 (Digital AP. Dr.
Communication) Mohd Fadzli
42- Development of child trap hazard In the last few years there have been incidents where child was kill after being NURUL AQILAH Mohd Salleh
indicator for ignorant car driver. left alone in a car unintensionally for several hours. This project to develop an BINTI ROSLI
indicator to alert the driver based on weight sensor that connect to car alarm (147044)
system and key remote for hazard indication if the driver unintentionally leaves
the car without picking up the child. Course Requirements: EEE123 (C++),
EEE320 (Microcontroller), EEE332 (Communication), EEE377 (Digital
Communication)
43- (Western Digital) This project is part of a collaboration between USM and Western Digital. You CHOY MENG
Augmented/Mixed Reality are required to develop an augmented or mixed reality program that can help HUI (145440)
AP. Dr.
Program Development for engineers/technicians in performing the repairing and maintenance
Nor Muzlifah
Repairing & Maintenance procedures. The tools that you will use are Unity with Vuforia Engine and you
Mahyuddin
Process have the opportunity to learn how to use the Microsoft HoloLens 2. You will be
working in School as well as in USM-WD CiA Lab at Western Digital.
44- Vision-Based Plant Health Synopsys: The identification of plant health status is a significant challenge for MUHAMMAD
Monitoring Using Deep Learning the timely and effective management of plants to ensure food security and AMIRUL
Models sustainable agriculture. A common method for the detection of plant stress is IKHWAN BIN
visual plant monitoring performed by human raters. However, this method is SHAMSURI
time-consuming, labor-intensive, and prone to human errors. Image (145488)
processing is one of the techniques which can help in the assessment of water Dr. Mohd
stress directly. Deep Learning combined with image processing can aid in Shahrimie Mohd
identifying water stress beyond the limitations of traditional image processing. Asaari
Deep learning (DL) techniques have gained momentum recently for image
classification and are being applied widely. In this research, the aim is to
conduct a comparative study to assess several (three to four) well-known DL
classification algorithms for the identification of plant stress in different (two or
three) types of horticulture crops. The state-of-the-art DL models will be
explored and the important hyperparameter tuning will be performed to achieve
superior identification accuracy.
Requirement: The students should have some programming experiences such
as Python and be familiar or willing to explore the knowledge of computer vision
and machine learning (deep learning). The final outcome of this project is
expected to be a software implementation.
45- IoT-Based Plant Health Synopsys: Plant health monitoring is an essential aspect of smart farming that CHEONG
Monitoring Using Yolo Model captures crop information at different growth stages. The health condition of KUANG YEN
plants can be assessed by monitoring any stress development during their (148502)
growing stages. Particularly, the prevention and control of plant stress are the
key steps in producing high yields and quality of agricultural products. Having
the ability to accurately detect stress symptoms at an early stage would give
chances for early intervention for control and change crop management Dr. Mohd
practices before the whole crop is damaged, which eventually minimizes the Shahrimie Mohd
agriculture losses. An emerging and innovative approach for autonomous Asaari
monitoring of plant stress is based on computer vision with deep learning
 techniques. In this research, the aim is to propose an embedded system for
crop health monitoring using the analysis of digital images. The state-of-the-art
of deep learning object detection algorithm; the You Look Only Once (YOLO
latest version) will be explored, analyzed, and implemented on the Nvidia
Jetson Nano developer kit to realize the embedded IoT system for automatic
plant health monitoring.
Requirement: The students should have some programming experiences such
as Python and be familiar with or willing to explore the knowledge of computer
vision and machine learning. The outcome of this project is expected to be a
software and hardware implementation. The inference of detection model will
be implemented on the development Nvidia Jetson Nano board connected with
a camera module and communicate to the IoT system.
46- Dynamic Hand Gesture Synopsys: Dynamic hand gesture trajectory is one of the crucial elements for TAN QIAN HUI
Recognition Using Deep many human hand-based computer vision task, such as hand pose estimation (148843)
Learning Technique and gesture recognition for human-computer interaction (HCI) system.
However, reliably detecting and tracking hands to extract the trajectory
information from the video scenes remains to be a challenging task because
of the complex appearance diversities of dexterous human hands (e.g.,
different hand shapes, skin colors, illuminations, orientations, scales, etc.) in
color images. In this research, the aim is to detect, track and extract the hand’s
motion trajectory from a sequence of images with minimal constraints on the
video scene. The shape information from extracted trajectory will be classified
into a meaningful signal as a proxy for hand gesture recognition. The emerging
state-of-the-art deep learning technique will be explored and analyzed to
achieve a fast and robust hand gesture recognition system.
Requirement: The students should have some programming experiences such
as Python and be familiar with or willing to explore the knowledge of computer
vision and machine learning. The final outcome of this project is expected to Dr. Mohd
be a software implementation with the interface connected to the USB camera. Shahrimie Mohd
47- Smart Computer Vision Lighting This project is the industry-based project with Western Digital Malaysia (WD) KHOR HUAI Asaari
Control System and student involve in this project will have a high chance to be their future SHENG
engineer at WD. This project aims to develop a smart lighting control system (146092)
for the manufacturing floor. The system will be based on the Artificial
Intelligence (AI) analytic platform (Nvdia Deep Learning Library and Solution)
that can be expanded to cater to multiple computer vision solutions such as
 social distancing and factory hazardous scenarios on the manufacturing floor.
A smart lighting system will be developed that can provide automatic control of
lighting by adjusting the brightness level based on the presence of the human
in the designated area. AI Computer vision solution will be the core component
in this lighting control system, where the live stream from multiple cameras will
be analyzed to detect and track the presence of humans. Information
processed by the AI Computer vision analytic platform will be translated to
control external circuitries to regulate (ON/OFF) the light in the designated
area. The objective of the project is to develop a computer vision AI analytic
platform for smart lighting control based on human detection and tracking and
also to achieve the best cost efficiency by exploring camera specification and
AI precision to further reduce the number of cameras required and
computational complexity. Requirements: Requirement: The students should
have some programming experiences such as Python and be familiar with or
willing to explore the knowledge of computer vision and machine learning.
48- Developing an auto-size system This project is supported by Western Digital, Penang. In a production system, MUHAMMAD
to hold the different sizes of PCB there are many different sizes of PCB and it requires different hardware RUZAINI BIN
board. platforms to hold a particular PCB board during assembly. However, changing
BADRUL ZAMAN
those different platforms (pallets) is time-consuming and requires manpower
(149463)
to do that. Therefore, in this project, you are required to build the hardware
using different sizes of the linear actuator and require programming to control
the movement in the x,y, and z directions. Besides that, you are required to
build an image processing system for processing to justify the size of the PCB
from the image which is sent the information to the Arduino/FPGA system for Dr. Ahmad Nazri
controlling the movement of the linear actuator. Ali
Requirement: Familiar with good Arduino programming. Good hardware
development skill. Good image processing software development and
hardware/software interfacing.
49- Employing Deep Learning based Finger-based biometrics has been widely applied for the last two decades even LUM KAH
on U-Net Architecture for suffers from several problems. Ease of use, high performance, and less HONG (148022)
Fingerknuckle and Creases expensive are the principles that have been considered when implementing
Segmentation finger-based biometric systems. Due to the higher acceptance of finger-based
biometric systems, it is required to explore the possibilities to deploy the system
with another alternative finger-based system for improving its performance.
One of the textures that can be acquired from hand-based recognition is the
 pattern of knuckles and creases of each finger.
Requirement: Python or Matlab programming and image processing.
50- Target area for spraying paint Labeling car parts in the beginning project. The labeling of these parts will be
using Artificial Neural Network used to train CNN to identify the position for spraying paint purposes
51- Distance identification using There are options to use either image or sensor to capture features to identify TIONG SIEH AP. Dr.
artificial neural network the distance HANG (146001) Syed Sahal
52- Animal repeller using Several sensors need to be identified at the beginning of the project. Monkey MUHAMMAD Nazli Alhady
microprocessor or fpga is the primary animal to be selected as the target to be drive away FAUZAN BIN Syed Hassan
MOHD TAJUL
ARUS (148890)
53- Smart Farming System using This project is about developing a smart farming system using SoC platform. TENG XING
SoC platform Both PS (Arm processor) and PL (FPGA) must be explored in this project. SHENG
Simple sensors such as temperature sensor and humidity sensor should be (148197)
AP Dr.
used in this project. Data should be sent to cloud for further processing.
Zaini bt Ab.
54- Effect of Air Pollution on Covid19 This project will study about effect of air pollution on Covid19. Data will be IZZAH
Halim
analysed to determine the significance of air pollution on Covid19. The result SYAHIDAH
will be used to forecast Covid19. BINTI SALEH
(138584)
55- CMOS Bandgap Voltage 1. Bandgap (BG) voltage reference circuit is most commonly used for all analog LOW CHEE
Reference Circuit (Nexperia) integrated circuits for providing reference voltage. Such circuit provides a WENG (147419)
constant voltage, regardless of process, power supply voltage and temperature
(PVT) variations. These circuit can be designed based on low voltage mode Dr.
BG or current mode BG. Since in the recent years the supply voltage is reduced Jagadheswaran
below 1.2V that is less the silicon BG voltage, the current mode BGs are Rajendran
preferred. This project aims to design a CMOS BG voltage reference circuit for
smartphones. 2. Requirement: Fundamental knowledge on analog circuit and
devices. Software tools: Cadence 3. Project sponsored by Nexperia
56- CMOS Constant Current 1. CMOS constant current reference (CCR) circuit is capable of a constant NISHANT A/L
Generator (Nexperia) current to a load despite of supply voltage variation, process variation and wide KUMARAN
range of temperature variations. The CCR circuit is of vital importance as it is (148953)
the essential building blocks of analog systems. Such circuits are used to bias
the Op-amps, oscillators, phase lock loops, data converters etc. This project
aims to design a CMOS CCR for smartphones. 2. Requirement: Fundamental
knowledge on analog circuit and devices. Software tools: Cadence
3. Project sponsored by Nexperia.
57- Wideband CMOS Power 1. Wireless communications have been progressing rapidly in recent years in
Amplifier for 5G mmWave where it is currently paving the way for 5G implementations around the world.
Applications (QED Technology) The 5G implementation is categorized into two frequency range namely FR1 Dr.
(410 MHz – 7.125 GHz) and FR2 (24.25 GHz – 52.6 GHz). FR2 is a part of Ka- Jagadheswaran
Band frequency band. In this project, a wideband CMOS power amplifier which Rajendran
operates from 24 to 30 GHz will be designed and simulated for the 5G
mmWave applications. The PA will be utilizing a 2-stage configuration to
achieve a competitive gain, efficiency and output power level. The PA will be
designed in Cadence Virtuoso software in a 65nm Process Design Kit (PDK).
The schematic and layout of the PA will be developed and simulated for data
analysis.
2. Requirement: Fundamental knowledge on analog circuit and devices.
Software tools: Cadence
3. Project sponsored by QED Technology
58- CMOS Low Noise Chopper 1. In spite of the evolution of scaled-down digital processes, analog circuits will Ameerah Adlina
Amplifier Design always be needed to perform a variety of critical tasks to interface the digital binti Azhar
(INTEL ELITE FYP PROJECT) with the external world. One of these functions is high-precision amplification
with low power consumption. To achieve this, due to the CMOS process
limitation, the circuit suffers in term of flicker noise (1/f). Hence to resolve this,
a chopper amplifier is desirable.
2. Requirement: Fundamental knowledge on analog circuit and devices.
Software tools: Cadence.
3. Project sponsored by Intel under their Intel ellite program.
59- Design of RFID Fabric Tag The project will cover design & simulation of a UHF RFID Tag antenna meant MUHAMMAD
Antenna for fashion industry for tracking & monitoring of retail products ARIF FIKRI BIN
SHAMSUDDIN
(143609)
60- Design of Power Divider The project will cover design & simulation of a microstrip antenna that has the AIMAN
Integrated Antenna capability as power divider application SYAZWAN BIN Prof. Dr.
ABDUL AZIZ Widad Ismail
(142425)
61- Design of Low Noise Amplifier The project will cover design & simulation of a microstrip antenna that is MUHAMMAD
Integrated Antenna Front-End integrated with low noise amplification functionalities FARIS BIN
ZAINUDDIN
(144330)
62- High Power Solar Charge There are many type of solar charge controller available in the market and the MUHAMMAD
Controller price are varies according to power handling of the devices. This research will AFIQ BIN
involve in designing high power and reasonable price for the solar system BURUHANUTH
applications. EEN (149247)
63- 28 GHz Air Substrate Antenna Designing high frequency antenna required high performance substrate and Prof. Ir. Dr.
can be very expensive. In this research, the study on air substrate antenna will Mohd Fadzil Ain
be carried out in the design of 28 GHz antenna.
64- Brushless DC Motor Controller Brushless DC motor can be found in many applications however high power MEHINTHRA
BLDC controller are expensive. This research will involve in the design and A/L VIJAYAN
testing of BLDC motor controller. (148900)
65- SIMULATION OF DIELECTRIC The Organic Thin-Film Transistor (OTFT) technology has advanced
MATERIAL ON THE significantly because of its potential benefits of flexibility, lightweight, low cost,
ELECTRICAL PERFORMANCE and can be constructed to be transparent. However, with the demand of high
OF ORGANIC THIN FILM performing OTFT, the study of gate dielectrics with a higher dielectric constant
Dr. Mohamed
TRANSISTOR FOR LOW than the conventional Silicon dioxide dielectric allows higher gate capacitance
Fauzi Bin
VOLTAGE DEVICE with the same dielectric thickness, hence improving the OTFT electrical
Packeer
APPLICATION performance owing to the dielectric-semiconductor interface. To this day, it is
Mohamed
still very challenging to precisely determine how the aforementioned elements
affect OTFT performance. As a result, a research gap exists for researchers to
further study the impacts of gate dielectric materials on the electrical
performance of OTFTs, which must be filled for them to be used successfully
and reliably in the future.
66- Design considerations of The performance of nanowire as a sensing transducer will be studied using Dr. Siti
nanowire device for sensor Silvaco ATLAS simulation. The structure analysis will be based on the current-
applications voltage characteristics of a field-effect transistor-based device.
Fatimah Abd
Rahman
67-
Covid19 identification based
INTEL ELITE FYP PROJECT AP. Dr. Syed
Ang Zheng Sahal Nazli
implementation on FPGA using
Xiong Alhady Syed
CNN
Hassan
68- Machine Learning Algorithm for INTEL ELITE FYP PROJECT Yeoh Zhien AP. Dr. Zaini
Sales Forecasting
Abdul Halim
69-
Car Plate Recognition System
INTEL ELITE FYP PROJECT Teoh Choon
Hong
70- A comparative study of Crosstalk INTEL ELITE FYP PROJECT
Mitigation Technologies for high-
speed digital design/ A AP. Dr. Nor
comparative study of Passive Liew Choy Tung Muzlifah
Equalizer Design for Single Ended Mahyuddin
High Speed Signal [Co-supervised
by Dr Azniza]
71- VLSI Design for Low Power Home INTEL ELITE FYP PROJECT Tan Yong Liang
Automation System Dr. Nor Azlin
72- Performance Analysis and INTEL ELITE FYP PROJECT Ghazali
Verification of Energy Efficient Lim Ying Tong
Multiplier
73- Human Movement Tracking data INTEL ELITE FYP PROJECT Tan Shu Han Dr. Ahmad Nazri
acquisition on the FPGA platform
Ali
74- FPGA implementation for image INTEL ELITE FYP PROJECT Ng Chee Kheng
segmentation
75- Digital Duty Cycle Correction INTEL ELITE FYP PROJECT
Circuit for Double Data Rate Khor Kai Yang Dr. Mohamad
(DDR) System Adzhar Md
76- High Speed FPGA-to-FPGA INTEL ELITE FYP PROJECT Zawawi
Transfer Controller using Intel Lim Xiao Xuan
Built-in IP
77- Design and simulation bandgap INTEL ELITE FYP PROJECT Cheong Kuang
Dr. Mohamed
voltage refrence circuit in a Fauzi Packeer
Yen
0.18mm CMOS process Mohamed
78- Carry Save Adder (CSaA) using INTEL ELITE FYP PROJECT Teoh Yong
Cadence Keong Dr. Siti Fatimah
79- Ripple Carry Adder using INTEL ELITE FYP PROJECT Ong Shi Mei
Abd Rahman
Cadence
80- 10-Bit Digital-to-Analog INTEL ELITE FYP PROJECT Dr. Mohd Tafir
Converter (DAC) using Partial Tan Yi Ka
Mustaffa
Binary Tree Network (PBTN)
81-
1.2 Volt Power Supply
INTEL ELITE FYP PROJECT Quak Ting
En. Zulfiqar Ali
Sheng
Abdul Aziz
82- 12 bits Hybrid DAC INTEL ELITE FYP PROJECT Ch'ng Ooi Keat
83- Matching Analog Layout OPPSTAR FYP PROJECT
Techniques for Current Mirror Dr. Mohamed
Circuits to mitigate Process Fauzi Packeer
Variation and Layout- Kavilenn A/L Mohamed
Dependent-Effect. Thiruchandra
84- Integrated Circuit Reliability - OPPSTAR FYP PROJECT Dr. Mohd Tafir
Antenna violations handling Mohamad Zaki
Mustaffa
technique Bin Abd Hamid
85- Floorplanning and Layout OPPSTAR FYP PROJECT Dr. Nor Azlin
Considerations for High-Power Navinkumar A/L
Ghazali
Circuits. Sivalingam
86- Layout Techniques for Sensitive OPPSTAR FYP PROJECT Dr. Mohamad
High-Speed Clock and Data Adzhar Md
Signals in SerDes Design Suraz A/L Naidu Zawawi
87- Integrated Circuit ESD Protection OPPSTAR FYP PROJECT Muhammad
Dr. Nur Zatil
- Various Layout Implementation Shafiq Iskandar
Ísmah Hashim
Schemes and Usages Bin Samsudin
88- Integrated Circuit Reliability - OPPSTAR FYP PROJECT
Tackling EM and SH in Early Aiman Danial Assoc. Prof. Dr.
Layout Planning Bin Najit Khan Nor Muzlifah
89- Power Grid Design Do's and OPPSTAR FYP PROJECT Muhammad Mahyuddin
Don'ts for Optimum IR-drop and Irsyad Bin
Performance Improvement. Shukor
90- Electrical characterization of low INFINEON KULIM FYP PROJECT Dr. Mohamed
voltage devices in Power SOI
LEE SAI LINK Fauzi Packeer
Technology [Industry supervisor:
Mohamed
Manoj]
91- Electrical characterization of INFINEON KULIM FYP PROJECT Dr. Mohamad
750V freewheeling diodes AW YONG GI Adzhar Md
[Industry supervisor: Philip] Zawawi
92- Characterization of SOI devices INFINEON KULIM FYP PROJECT AUFA NADIA Dr. Nor Azlin
by IV and CV method [Industry
BINTI SUHANDA binti Ghazali
supervisor: Tan Chan Link]
93- Area optimization of active INFINEON PENANG FYP PROJECT Dr. Mohd Tafir
reference band gap amplifier Teh You Han
Mustaffa
layout design
94- Area optimization of phased lock INFINEON PENANG FYP PROJECT Dr. Nur Zatil
look low-dropout voltage Chai Chern ‘Ismah binti
controlled oscillator layout design Hong Hashim
95- Power and area estimation of link SkyeChip Sdn Bhd
AP Dr Bakhtiar
in Network-on-Chip (NoC) using
Affendi Rosdi
machine learning Cheng Khe Xian