DR-NTU (Digital Repository of NTU)
Not a member yet
116018 research outputs found
Sort by
Efficient simulation of electromagnetics and nanophotonic
No full textElectromagnetics can be a complex concept to embark on and understand. Thus, the goal of
this project is to find the efficient ways to simulate electromagnetic wave propagation to help
simplify this complex concept for visualization and understanding. Three types of simulation
methods will be studied and compared: Finite-Difference Time-Domain (FDTD), FiniteElement Method (FEM), and Finite-Volume Time-Domain (FVTD).
For FDTD, custom 2D and 3D FDTD simulations are created based on Maxwell’s equations.
Key components like Yee’s grid and Mur’s absorbing boundary condition were also
implemented. FEM was explored through existing frameworks, though FVTD
implementation remains a work in progress. The report evaluates the strengths and limitations
of each method and discusses potential improvements.Bachelor's degre
Study of metal additive manufacturing defects
No full textX-ray computed tomography (X-CT) has emerged as a vital non-destructive testing method
in demanding sectors like aerospace and biomedical which is used for assessing the interior
integrity of metal additive manufacturing (AM) parts. Although X-CT provides multiple
advantages over conventional techniques, several drawbacks prevent it from being used effectively
for defect diagnosis in AM. Such of these include the intrinsic trade-off between scan throughput
and voxel resolution, the presence of aberrations such as beam hardening, irregularities in defect
quantification, and the obstacles of in-situ monitoring in real time.
In addition to addressing the shortcomings at present in its approach and post-processing
techniques for precise defect identification and prediction, it introduces the importance of X-CT in
AM quality control. It also underlines how techniques like AI-driven analysis, adaptive scanning,
and in-situ monitoring may be used to overcome these drawbacks and enhance X-CT processes for
increased productivity and dependability in the evaluation of metal AM components.
Comprehensive discussions of these concerns and fresh strategies are provided in the literature
review that follows, which also develops the foundation for the methodology intended to address
these constraints by integrating computational post-processing, artificial intelligence, and
experimental data collection.Bachelor's degre
Splat4D: diffusion-enhanced 4D Gaussian splatting for temporally and spatially consistent content creation
No full textGenerating high-quality 4D content from monocular videos—for applications such as digital humans and AR/VR—poses challenges in ensuring temporal and spatial consistency, preserving intricate details, and incorporating user guidance effectively. To overcome these challenges, we introduce Splat4D, a novel framework enabling high-fidelity 4D content generation from a monocular video. Splat4D achieves superior performance while maintaining faithful spatial-temporal coherence, by leveraging multi-view rendering, inconsistency identification, a video diffusion model, and an asymmetric U-Net for refinement. Through extensive evaluations on public benchmarks, Splat4D consistently demonstrates state-of-the-art performance across various metrics, underscoring the efficacy of our approach. Additionally, the versatility of Splat4D is validated in various applications such as text/image conditioned 4D generation, 4D human generation, and text-guided content editing, producing coherent outcomes following user instructions. Project page: https://visual-ai.github.io/splat4dPublished versionThe work is supported by the Hong Kong Research Grants Council - General Research Fund (Grant No.: 17211024)
Study of warm mix technologies for asphalt concrete
No full textIn recent years there has been increased focus on sustainable alternatives in Singapore’s transport
industry. Of particular note is the use of Warm Mix Asphalt (WMA) as a potential replacement for
conventional Hot Mix Asphalt (HMA) due to its lower production temperature whilst also maintaining
similar or better mechanical performance.WMA also has a host of numerous possible properties
depending on additive material added to the mix design which makes it ideal to find one that will
particularly suit Singapore’s environmental conditions.
This study aims to determine the viability of Saraphalt®
as an additive in WMA. Four tests, namely
Marshall Stability and Flow Test, Moisture Susceptibility Test, Indirect Tensile Modulus Test and
Dynamic Creep Test were conducted on W3B mix specimens with Saraphalt®
as an additive, hydrated
lime as filler and at 5% binder content. The results were collated and analysed before being compared
to similar mix designs to determine the extent of improvement in mechanical performance.
The results showed that specimens with the addition of Saraphalt®
indicated slight improvements in
mechanical performance and more notably conformed to LTA workmanship specifications to a greater
degree compared to previous studies done. However, as the study done might not be comprehensive
enough, further exploration into characteristics of the mix such as rutting resistance may prove
necessary to obtain a more well-rounded evaluation.Bachelor's degre
Dual-polarized directional finding circular array
No full textPolarization plays a crucial role in wireless communication by enhancing channel capacity and reducing multipath interference. This project focuses on designing and analyzing a wideband dual-polarized circular antenna array to improve signal performance in wireless and direction-finding applications.
Operating at 2.4 GHz, the proposed array supports dual linear polarization, enabling better signal separation and reduced interference. Its 360-degree azimuthal coverage makes it ideal for omnidirectional applications, while a gain of over 6 dB ensures strong signal propagation. The antenna is designed and optimized using HFSS on ANSYS Electronics, with key performance metrics such as radiation patterns, impedance matching, and polarization purity evaluated through simulations.
The results demonstrate the feasibility of dual-polarized circular arrays for modern communication systems, offering improved signal integrity, spatial coverage, and spectral efficiency. This study provides valuable insights for developing high-performance antennas in wireless networks, radar, and sensing applications.Bachelor's degre
Anisotropic properties and behaviours of carbon sequestrated 3D printed concrete
No full textThe construction industry is under increasing pressure to reduce its carbon footprint while maintaining structural integrity and performance. This project investigates an innovative approach to 3D printed concrete (3DCP) by integrating CO₂-Steam technology with Polyethylene Glycol (PEG) reinforcement. The primary objectives are to enhance the mechanical performance of 3DCP and to assess its potential for carbon sequestration and decarbonization. The study employs a comprehensive experimental and numerical framework. The concrete mixes are modified by incorporating different concentrations of PEG (0%, 1%, 2%, and 3%), and half of the samples are subjected to CO₂ treatment (CO₂-Steam Integration into the printing Process.) Testing protocols include flexural, compressive, and interlayer bonding assessments at curing durations of 7 and 28 days. These tests are designed to capture the inherent anisotropic behaviour of 3DCP, particularly the directional dependencies associated with the layered manufacturing process. Notably, while PEG significantly enhances in-layer properties through improved internal curing and moisture retention, its ability to reinforce the weak interlayer interfaces is limited due to fibre confinement within individual layers. Numerical simulations are conducted using ABAQUS, where the Concrete Damaged Plasticity (CDP) model is augmented with cohesive zone modelling (CZM) to accurately replicate the nonlinear behaviour of the composite material. This integrated approach provides a robust framework for evaluating the performance of 3DCP under varied loading conditions and offers insights into the interplay between material composition, curing conditions, and structural anisotropy. Overall, this research advances our understanding of sustainable 3D printed concrete and contributes valuable insights into the integration of carbon sequestration technologies within the construction industry, paving the way for more environmentally friendly and structurally efficient building practices.Master's degre
Controllables: a universal framework for control and modeling of dynamic systems
No full textWe propose Controllables, a unified reactive framework for modeling and controlling
dynamic systems. The architecture is centered on two primitives —
Variable and Callback — which together support composable state modeling
and reactive behavior. The framework emphasizes separation of concerns
between sensors, control logic, and agents, and enables introspectable, real-time
interaction with control workflows. It is designed to bridge simulation-based
development with live system operation, while remaining lightweight and extensible.Master's degre
Investigation of optimal gear ratio of radial coaxial magnetic gear for low-speed direct-drive applications
No full textMagnetic Gears are a technological leap in the mechanical and electrical domains in transmission of power and torque in today’s world. For the past two centuries, mechanical gears bore the responsibility of high and low torque transmission in the industrial revolution and the automotive industry. However, due to limitations posed by mechanical gears such as wearing of teeth due to prolonged persistent friction and vibration, constant maintenance and replacement has become a financial toll on manufacturing companies together with increased overhaul time. The innovation of magnetic gears solved the issues posed by mechanical gears and was able to translate the same torque as that of mechanical gears. This study will cover on which of the two permanent magnet arrangements: Halbach or Conventional, will contribute to a greater torque output at two different GRs for RCMGs in a turboprop reduction gearbox of an ATR aircraft.Bachelor's degre
Enhancing security and efficiency for the internet of vehicles systems
No full textThe Internet of Vehicles (IoV) is an improvement and integration over Vehicular
Ad-hoc Networks (VANET) and the Internet of Things (IoT) [1].
By harnessing data insights from a range of devices and sources, IoV holds
immense potential to revolutionize urban transportation systems by enhancing
road safety, alleviating traffic congestion, and enriching user experiences
through onboard infotainment applications. However, the extensive connectivity
of IoV is a double-edged sword. While it significantly enhances the
data intelligence of IoVs, it also increases potential attack surfaces, putting
their security and privacy at risk. Traditional security mechanisms designed
to protect the confidentiality, integrity, and availability of individual devices
or communication channels are insufficient for the heterogeneous communication
environments IoV encounters. This issue is exacerbated by the varying
capabilities and priorities of devices, where devices may lack the resources to
implement traditional security measures, such as enforcing common encryption
standards. Furthermore, new security designs are required to safeguard
new types of interactions and security requirements that emerge from IoV interactions.
For example, maintaining the security of real-time, time-sensitive
traffic data in IoVs while guaranteeing the quality of onboard infotainment
services is a major challenge. In this thesis, we aim to study the security
requirements of the IoV ecosystem and design data-centric security solutions
that enhance the security and efficiency of the IoV systems.
Firstly, we present our study on the context-specific security requirements of
IoVs and the potential security challenges arising from their interaction with
the urban ecosystem. Instead of focusing on individual device or communication
channel protections, our research adopts a data-centric perspective
that encompasses the entire data lifecycle to achieve end-to-end security. By
doing so, we reduce the security risks posed by new system interactions or
unanticipated user behaviors while avoiding the need for frequent security
model upgrades.
Secondly, we propose a differentiated security architecture for infotainment
data communication in IoV networks from a data-centric perspective. The
security mechanisms used to protect each communication are determined by
the type of data exchanged and the security focus of that communication,
referred to as Quality of Security Services (QoSS) for each communication.
Specifically, we utilize Named Data Networking (NDN) for file distribution
and introduce a time-sensitive Key-Policy Attribute-Based Encryption (KPABE)
scheme for sharing subscription-based infotainment data. This approach
facilitates the easy distribution of files and ensures that only authorized
users can decrypt files, thus preventing piracy of digital media and
protecting the overall health of content creation.
In addition, to enhance the efficiency of our differentiated security architecture,
we model the IoV ecosystem as a data trading market within our
framework. Building on this framework, we propose a decentralized datasharing
incentive mechanism based on multi-agent reinforcement learning
to learn the supply-demand balance in markets and minimize transmission
latency. Furthermore, our proposed mechanism accounts for the dynamic
nature of IoV markets, which often experience frequent fluctuations in supply
and demand. By finding the supply-demand equilibrium, this mechanism
improves the efficiency and sustainability of the data trading markets.
Lastly, we present our conclusions and provide insights into potential directions
for future research.Doctor of Philosoph
A web-page-based software program for technology-enhanced learning (TEL) of finite state machines (FSMs) (A)
No full textThis project aims to develop a web-based learning tool to facilitate the learning of
the principles and applications of Finite State Machines (FSMs). My objective for
this project is to address the transition of State diagrams to Circuit form, the other
way around, which is the Circuit to State transition, will be done by another student
working on this project as well. The main focus of the project is to design an
application that is able to transition the data in state transition tables, which are
derived from finite state machines (FSMs), into related digital circuits with various
Flip-Flop types such as D, T, and JK. The project employs a user-friendly interface
such that users can input state transition data, which in turn gets processed for
generating excitation tables and simplified boolean equations. These equations are
critical in designing the logic circuits implementing the specified state transitions.
The project leverages modern web technologies, including React for the user
interface frameworks, JavaScript and CSS (Cascading Style Sheets) for visualizing
state diagrams and UI Design. Utilizing these technologies makes it easy for users
transitioning from theoretical state diagrams that we learnt from our modules to
practical circuit implementations. This project will not only be beneficial for
educational settings for teaching digital electronics logic design but also possibly
assist engineers in prototyping and testing digital systems efficiently
The project demonstrates the possibility of combining interactive web applications
with digital logic design, paving the way for more intuitive and easily-accessible
tools in the field of EEE and Computer Engineering.Bachelor's degre