Heriot-Watt University
ROS: The Research Output Service. Heriot-Watt University EdinburghNot a member yet
4689 research outputs found
Sort by
Site optimisation for native oyster (Ostrea edulis) reintroduction and its realistic application to modern day site suitability analyses
Full text linkOyster populations have been globally depleted by 85% and historical Ostrea edulis
beds may no longer be suitable due to anthropogenic and natural changes. Robust
site suitability assessment must occur in prospective restoration sites to determine if
restoration is feasible in the current environment. Studies on the environmental
preferences of the species are mostly based on lab studies and expert observation
and opinion. The relationship between oyster density and environmental conditions in
the last natural oyster fishery in Loch Ryan, Scotland, were tested in a unique field
study. The percentage of hard substrate and presence of gravel significantly
influenced oyster density across the loch. The optimum environmental conditions
determined from Loch Ryan informed a MCDA-GIS based site suitability model in the
Firth of Forth, Scotland. The most suitable sites for O. edulis restoration in the Forth
were identified around Inchkeith, Leven/Kirkcaldy and North Berwick with varying
confidence. The outcome of the suitability analyses varied based on the importance
applied to sub-models in response to different restoration scenarios. Application of
this type of modelling in similar environments must carefully assess and weigh the
parameters according to the species needs within the specific environment and its
associated pressures
Assisted living - an agent based architecture for automated cognitive assessment
Full text linkDementia is a broad category for chronic and progressive diseases affecting the brain,
which are defined as a decline in cognitive abilities severe enough to affect performance
of activities of daily living. Accurate diagnosis is essential to support pro-active and
personalised care practices.
Current assessments of Dementia are designed to measure cognitive abilities in isolated and artificial situations. The benefits of these tests are limited as testing may not
translate well to real world situations.
The vision is for assessment and assistive systems that can be kept consistently in tune
with the needs of their users, by more easily incorporating the input from human experts
and by using an explicit user model.
A customised agent architecture was implemented to enable discrete event simulations of ADLs, and to track human actions in a sensorised kitchen environment. A cognitive user model was developed based on declarative and semantic memory, and was
evaluated on a common cognitive assessment with a data set of 149 older adults.
An automated cognitive assessment protocol was proposed and implemented based
on these observations of kitchen ADLs, and evaluated with a user study of 30 participants. A system architecture linking automated assessment and automated assistance
was proposed
Study on degradation of plastics using cold plasma
Full text linkRecycling plastic wastes is challenging as the global plastic production and consumption
has been rapidly increasing in recent years. This problem has been exacerbated due to the
long-time scales required for polymer degradation in the natural environment. According
to Royal Statistical Society report in 2018, 90.5% of plastic waste has never been
recycled. The aim of this project is to rapidly reduce the time frame for polymer
degradation by using cold plasma technology for depolymerisation of waste plastics to
produce value-added chemicals in a simple, scalable fashion with enhanced energy
efficiency. Five of the most common types of plastics polyethylene (PE), polystyrene
(PS), polyethylene terephthalate (PET), polyvinyl chloride (PVC), polypropylene (PP); a
surgical 3-ply mask; four bioplastics polylactic acid (PLA), poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) (PHBH), polybutylene adipate terephthalate (PBAT) and
polycaprolactone (PCL); and two fibre reinforced polymers (FRP); fibre glass reinforced
polymer (FGRP) and carbon fibre reinforced polymer (CFRP) have been chosen in this
project to observe the plasma-surface interaction and depolymerisation processes. Results
indicate that all the plastics studied underwent mass loss to varying degrees after plasma
irradiation. In addition, PE, PS, PET, PVC, PP and the 3-ply surgical mask all
incorporated oxygen- and nitrogen-containing groups, the bioplastics incorporated the
hydroxyl (OH) group while the FRPs did not show any functionalisation. Moreover, for
PE, PS, PET, PVC, PP three different gas sources, ambient air, nitrogen and a mixture of
80% nitrogen/20% hydrogen were studied. Air was the most active while the 80%
nitrogen/20% hydrogen was the least active.Engineering and Physical Sciences Research Council (EPSRC) funde
Operating regimes of injection-locked quantum-dot lasers and applications in high-speed optical sampling
No full textAbstract currently unavailable. Restricted access until 31.08.2024. Please refer to PDF
Investigating phraseology in new speakers and first speakers of English in academia
Full text linkThis study is a corpus-based study investigating the phraseology of new speakers and first
speakers of English. As a consequence of globalisation and the movement of populations,
many people have become new speakers of a language they use in everyday social or
workplace situations but which they did not learn by conventional inter-generational
transmission. Furthermore, the language that has become the dominant language of
business, the media and academia is English. Consequently, there are now more new
speakers of English than first speakers and this brings into question issues of linguistic
capital, language ownership, the legitimacy of speaker-ness and the Chomskyan notion
that the native speaker is the authority on their mother tongue. This study is corpus based,
investigating the phraseology, in the form of n-grams, of new speakers and first speakers
of English in both the written and spoken language in the academic context in the UK.
Findings show that there are similar trends with little divergence, minimal differences and
no statistical significance, between the phraseology of new speakers and first speakers of
English. This suggests that any differences are due to chance or natural variation, thus
supporting the questioning of the hegemony of ‘native speaker’ English in academia
Design and analysis of hardware-friendly optimization algorithms
Full text linkThis thesis presents an in-depth study of advanced optimization algorithms, focusing on Approximate Proximal Gradient Descent (AxPGD), Approximate Alternating
Direction Method of Multipliers (ADMM), and the newly proposed Dual-Feedback
Generalized Proximal Gradient Descent (DFGPGD). The research investigates these
algorithms’ convergence, robustness, and efficiency in the context of computational
inaccuracies, such as limited precision and round-off errors. A key aspect of this
study is the analysis of error bounds in numerical optimization algorithms, particularly in scenarios where only approximate first-order information is available, and
the proximal operator is computed with finite precision using approximate computing
techniques.
The work establishes worst-case and probabilistic upper bounds on the distance
from the optimal solution in terms of objective value evaluations. By assuming
stochastic bounded error sequences, new probabilistically tight bounds are proposed,
enhancing our understanding of the algorithms’ performance under bounded error
conditions. The research extends beyond convex problems to nonconvex lower-C
2
problems, offering convergence guarantees for both basic and accelerated versions of
AxPGD. The introduction of DFGPGD marks a significant advancement, with convergence guarantees and numerical stability established through both analytical and
dynamical system approaches.
This comprehensive study not only provides theoretical insights into the behavior
of optimization algorithms under computational constraints but also bridges the gap
between theoretical optimization and practical applications in energy-aware systems,
such as FPGA-based Model Predictive Control of satellite systems and compressed
image reconstruction. The proposed new power cost function and optimal algorithm
design parameters are particularly noteworthy for their potential to enhance performance in energy-constrained computational environments.James Watt Scholarshi
An exploration of deaf interpreters’ experience of professional autonomy and the source text : a qualitative and autoethnographic study
Full text linkThis qualitative study focuses on the lived experience of ten deaf interpreters and their sense
of professional autonomy in interpreting situations and takes an autoethnographic approach
which supports the research. The study chose to focus on how deaf interpreters access source
text as a way to research this topic. The study addresses the significant gap in sign language
interpreting studies literature due to the scarcity of research on the experiences of deaf
interpreters. Most sign language interpreting studies tends to focus on hearing interpreters. This
study gathers the views and experiences of deaf interpreters using qualitative interviews.
Interview data provides narrative understanding of professional autonomy as expressed by the
participants as factors that influence interpreting decisions and professional status. The theory
of professional autonomy frames the research widening our understanding of deaf interpreting
as a profession. The aim is to address the central research question ‘can deaf interpreters be
considered as having professional autonomy?’ The results show that deaf interpreters have a
complex relationship with their professional autonomy and are often more aware of it when
they feel its absence. The lack of and content of existing professional interpreting courses for
deaf interpreters compounds the problem. The research notes the implications for the future of
deaf interpreters and discusses the limitations of the research. The study’s contributions to
knowledge of sign language interpreting are also acknowledged. Finally, the study concludes
with recommendations around professional training for deaf interpreters and that the
developing profession of deaf interpreting requires further research
Studying the effects of integrating private water tanks on the performance of water distribution networks
Full text linkPrivate water tanks in Water Distribution Networks are essential in providing additional storage
to satisfy the consumer needs. These tanks, that are located at the rooftop or underground area of
the building, can affect the empirical parameters of the network, including pressure, flowrate,
pump’s power consumption, water age and chlorine concentration. However, they are often
neglected in design stages for the sake of quicker simulations and simplified calculations. This
oversight compromises the accuracy of network’s simulation by omitting crucial tank parameters,
like orifice inlet and volume capacity, which in worst cases, can cause pipe burst or insufficient
supply. Hence, this prompts to investigate the impact of adding private tanks on the network’s
performance.
This study employed a quantative research using pressure-driven analysis to integrate private
tanks in network models. This integration is carried out by incoperating mass balance models at
the nodes to mimic the filling and emptying condition of the private tank, which is extended to
the hydraulic model of the network, by modifying the energy conservation and mass conservation
equations. In total, three real-time networks from Dubai and Abu Dhabi, United Arab Emirates,
along with five sample networks, were built within hydraulic tools, EPANET and WDnetXL,
using the data from local water distribution codes. As private tanks are mandatory only in
resedential buildings, data from Supervisory Control And Data Acquisition and water meter
readings were also used to build consumption pattern to model the filling and emptying behaviour
of these tanks during simulation.
The results show that incorporating private tanks in a water system affects the pressure and flow,
where having larger tank’s inlet diameter, the flowrate in the pipe increases, which increases the
head loss in the pipes. This raises the pump’s power consumption and maintenance costs, due to
a higher risk of pipe bursts because of more flow. Ultimately, it can further result in decrease in
the pressure at the downstream nodes by 50% due to higher head losses at the upstream.
Furthermore, large tank volume would increase the overall flow from the pumps and water age
as large tanks tend to close the orifice for a longer duration, which increases the water age in pipes
and increases the risk of contamination in pipes. As such, this study advocates using the smallest
values for orifice diameter and tank volume sizes that achieves the network reliability.
Additionally, this research also improves the network performance by reducing the pump’s
running time by using water supplied from private tanks. This decreases the overall pump’s
carbon footprint by 40% along with leakage rates, when compared to networks without private
tank. Ultimately, the findings of this research would provide an accurate depiction during design,
minimising the risk of overdesign by avoiding unnecessary parameter changes by the designers
and prompting conservation strategies using private tanks
Analysis of the availability of Infrastructure-as-a-Service-based cloud computing
Full text linkCloud computing has become pervasive in organizations worldwide. A primary
concern with the Cloud is security, especially the availability posed by Denial of
Service (DoS) and Distributed Denial of Service attacks(DDoS). The evolving nature
of this attack using reflectors and amplifiers eliminates the need for the attacker to
have access to huge resources.
We created a Cloud environment using Openstack and examined the availability
of each building block component of the Cloud. Our experiments revealed vulnerabilities that led to availability issues in the Advanced Message Queuing Protocol
(AMQP) message broker service, and the default one for Openstack is RabbitMQ.
Our experimentation showed that it is possible to launch an amplification attack on
RabbitMQ, which crashed the Cloud infrastructure.
Our study showed that when the impact of a DoS attack is considered, only
the network variables are considered in datasets. Hence, the resulting solutions
to detect or prevent DoS attacks are built by testing them against these datasets.
Using our infrastructure, we created a dataset with a series of systematic attacks in
the Cloud that captured over 230 variables from seven different resource categories
like processing, memory, and others.
We studied the impact of DoS attacks (specifically TCP flood attacks) across
different resource categories. We found attacks more impactful when the victim
and attacker co-resides are in the same cloud. Additionally, our work allowed us
to understand combinations of packets with flag and payload size that an attacker
can devise for the maximum impact on the victim. Our results contain previously
unknown insights, such as the fact that relatively smaller DoS packets could result in
a larger impact on the victims. We also identified the most impacted system metrics,
which would allow Cybersecurity software developers to build better and optimal
(D)DoS monitoring and detection tools. We also proposed a metric to quantify
the impact of a DoS attack considering the Cloud’s infrastructure and the various
resource categories
Underwater imaging with single-photon avalanche diode detector arrays
Full text linkThe time-correlated single-photon counting (TCSPC) technique was introduced less than
30 years ago for acquiring range information and reconstructing three-dimensional
scenes. Since then, the research field has undergone significant advancements in timing
electronics, optical system configurations, image processing algorithms, and particularly
single-photon detectors. The detectors used are typically semiconductor single-photon
avalanche diode (SPAD) detectors, which can operate in the visible, near-infrared and
short-wave infrared. More recently, single-photon imaging approaches have been used
for high-depth resolution imaging in challenging environments where light scattering
occurs, like through fog or underwater.
This Thesis examines aspects of underwater single-photon imaging. Initially, a three-dimensional single-photon depth imaging system, was used for several experiments. A
bistatic optical configuration was used with a picosecond pulsed laser system with an
operational wavelength of 532 nm. In the receive channel a CMOS SPAD detector array
with 192 × 128 pixels was used, with each pixel of the detector having its own time-to-digital converter for time-tagging detected events. This system was housed in a
waterproof enclosure and submerged to a depth of approximately 1.8 m, recorded static
target scenes at various scattering levels in water, with a target stand-off distance of 3 m.
The three-dimensional imaging system demonstrated sub-centimetre depth resolution in
clear water, and the angular resolution of the imaging system was evaluated to be 84 µrad
(both horizontal and vertical). The experiment included the data acquisition of moving
targets illustrating the capability of different algorithms in the real-time processing of
single-photon data with low-latency in the presence of high levels of scattering. This
work represented the first example of a fully submerged underwater single-photon
imaging system.
This bistatic transceiver system was used in the laboratory to examine the time-domain
characterisation of scattering in water. A mathematical model was developed to
reconstruct a single-event scattering process of photons in water. Consequently, this was
used to establish the average scattering length in various underwater conditions.
A different transceiver was constructed to utilise a linear 16 × 1 SPAD detector array,
developed using a custom fabrication process. The use of the custom fabrication meant
that low jitter performance was obtained across the detector array. This transceiver
system recorded moving target scenes and was evaluated for high scattering levels,
demonstrating its exceptional performance in highly scattering water. Its average optical
illumination power was less than 15.5 mW with a depth resolution of better than 2.5 mm
Access Repository Dashboard
Do you manage ROS: The Research Output Service. Heriot-Watt University Edinburgh ? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!