National University of Ireland, Maynooth
MURAL - Maynooth University Research Archive LibraryNot a member yet
18981 research outputs found
Sort by
A Causal Approach to Hard Constrained Control of Wave Energy Systems based on an Implicit Gaussian Differential Equation.
Full text linkThis paper introduces a novel method to causally
satisfy hard position and velocity constraints in wave energy
systems. The constraint mechanism is simple to implement,
computationally efficient, and does not require complex tuning
or optimisation techniques. The proposed strategy handles
system constraints by modulating a velocity reference with a
Gaussian-like envelope function that depends on both position
and velocity, which results in nonlinear closed-loop dynamics.
In this context, this paper focuses on the stability of the
constrained closed-loop dynamics, and it is proven that, for a
set of initial conditions within the constraint region, the system
trajectories remain within the prescribed limits. Finally, insilico evaluations demonstrate that the Gaussian-like function
effectively guarantees compliance with system constraints and
is broadly applicable to wave energy systems
Testing a dose‐response effect of the visuospatial game Tetris on intrusive memories
Full text linkTetris has been shown to reduce intrusions following exposure to experimentally induced and actual traumatic events. However, no study has systematically investigated whether multiple sessions of Tetris produce greater reductions in intrusions than a single session. In this study, 94 participants (58.5% female) watched a trauma film in the laboratory and were then randomly assigned to one of three groups: no Tetris (inactive control), a single session of Tetris (15 min on Day 1), or multiple sessions of Tetris (15 min per day on Days 1, 2, and 3). Participants recorded film‐related intrusions in a daily diary over 1 week. The results showed that the trauma film effectively induced intrusions. In terms of group differences, a single Tetris session was associated with a 22.0% reduction in intrusions compared to the control group, Exp( B ) = 0.78; and multiple Tetris sessions were associated with a 13.3% increase in intrusions compared to the control group, Exp( B ) = 1.13, and a 45.4% increase compared to the single‐session group, Exp( B ) = 1.45. However, none of these differences were statistically significant, p = .380. These findings may be partially explained by methodological factors, such as administering Tetris remotely via smartphones without researcher supervision and the repeated use of reminder cues. Alternatively, Tetris may not effectively reduce intrusions when played unsupervised in uncontrolled settings
Equidistant Landmarks Fail to Produce the Blocking Effect in Spatial Learning Using a Virtual Water Maze Task with Healthy Adults: A Role for Cognitive Mapping?
Full text linkBACKGROUND/OBJECTIVES
Cue competition is a feature of associative learning, whereby during learning, cues compete with each other, based on their relative salience, to influence subsequent performance. Blocking is a feature of cue competition where prior knowledge of a cue (X) will interfere with the subsequent learning of a second cue (XY). When tested with the second cue (Y) alone, participants show an impairment in responding. While blocking has been observed across many domains, including spatial learning, previous research has raised questions regarding replication and the conditions necessary for it to occur. Furthermore, two prominent spatial learning theories predict contrary results for blocking. Associative learning accounts predict that the addition of a cue will lead to a blocking effect and impaired performance upon testing. Whereas the cognitive map theory suggests that the novel cue will be integrated into a map with no subsequent impairment in performance.
METHODS
Using a virtual water maze task, we investigated the blocking effect in human participants.
RESULTS
Results indicated that the cue learned in phase 1 of the experiment did not interfere with learning of a subsequent cue introduced in phase 2.
CONCLUSIONS
This suggests that blocking did not occur and supports a cognitive mapping approach in human spatial learning. However, the relative location of the cues relative to the goal and how this might determine the learning strategy used by participants was discussed
Factorial Design and Optimization of Trimetallic CoNiFe-LDH/Graphene Composites for Enhanced Oxygen Evolution Reaction
Full text linkLayered double hydroxides (LDH) have exhibited promising applications
as electrocatalysts in oxygen evolution reactions (OER). In this work, trimetallic LDHs
(CoNiFe-LDH) were designed and grown on graphene (G) through a one-step
hydrothermal approach to obtain a structure that promotes efficient charge transfer. A 2-
level full-factorial design was utilized to evaluate the effects of varying the concentrations
of Co (1.5, 3, and 4.5 mmol) and graphene (10, 30, and 50 mg) on the OER activity.
The potential needed to deliver 10 mA cm−2 was chosen as the response parameter. The
independent and dependent parameters were fitted to a linear model equation through
ANOVA analysis. The computed p-values were below 0.05 signifying the statistical
significance of the concentrations of cobalt and graphene and their interaction,
suggesting a correlation with the OER activity. The OER experiments were conducted in
triplicate using the Co[3]Ni[3]Fe[3]-LDH/G[30] (central point) to estimate variability
(0.58%). Comparative analysis showed that Co[1.5]Ni[3]Fe[3]-LDH/G[10] achieved the
lowest onset potential (1.54 V), potential at 10 mA cm−2 (1.58 V), and Tafel slope (58.4 mV dec−1
), indicating that a low
concentration of cobalt and graphene make an efficient electrocatalyst for OER. Furthermore, the optimized composite
demonstrated favorable electronic properties, with a charge transfer resistance (RCT) of 188.1 Ω, and exhibited good stability,
maintaining its catalytic activity with no significant loss over a 24-h period
Isotopy and Concordance in Intermediate Ricci Curvatures
Full text linkIn this thesis we extend a result of M.Walsh that showed that, under reasonable conditions,
positive scalar curvature metrics which are Gromov-Lawson concordant are
in fact isotopic. This thesis generalises this result by proving that Gromov–Lawson
concordance implies isotopy in the space of Riemannian metrics on simply connected,
smooth, closed manifolds with positive Ricci-(k, n) curvature for certain k
at least 3 when n ≥ 5. To do this, we use a strengthening of the Gromov-Lawson
surgery technique for extending a positive scalar curvature metric over the trace of
a codimension ≥ 3 surgery to a positive scalar curvature metric which is a product
near the boundary. We extend this to positive Ricci-(k, n) curvature metrics making
use of theorems of Wolfson and Kordass. We also compute the Ricci-(k, n) curvature
on a variety of standard metrics on the sphere and disc including so-called mixed
torpedo metrics. In addition we give the conditions under which these standard
metrics are isotopic in the space of positive Ricci-(k, n) curvature metrics.
Moreover we extend a theorem of Carr to show that the space of positive Ricci-
(k, 4n−1) curvature metrics on a (4n−1)-dimensional, smooth, closed, spin manifold,
n ≥ 2, has infinitely many path components
Acoustic Machine Learning Tools and Analysis Software for Advancing Biodiversity Monitoring
Full text linkWhen investigating the ecological and behavioural patterns of wildlife through
sound, bioacoustic studies using machine learning (ML), such as convolutional
neural networks (CNNs), are key for analysing large acoustic datasets. For other
biodiversity monitoring methods like ecosystem accounting, practitioners often do
not have the required technical knowledge. This thesis presents a series of studies
focused on the application of ML methods to bioacoustic data, and providing tools
to assess ecological impacts.
We begin by conducting a systematic literature review of Passive Acoustic Monitoring
(PAM) to investigate how it’s used with ML methods and identify trends or
gaps present. We find increases in dataset size over time, and spectrograms being
the most popular representation of bioacoustic data for inference, but highlight
the need for standardized evaluation methods and broader use of open-source to
advance the field.
Our second contribution is NEAL, an open-source Shiny R application, which
enables granular annotation of audio data for use in training and evaluating species
classification models. Its no-code interface and modular design empower use by
non-programmers, improving annotation workflows and supporting machine learning
model development in bioacoustics.
Later, we use generative AI, specifically Stable Diffusion, to create synthetic
spectrograms for use in training bird species classification models. We use a dataset
annotated using NEAL to benchmark these models. We demonstrate that supplementing
training data with synthetic samples enhances classification performance
on the human-labelled test set.
Our next contribution is ExActR, an open-source Shiny R application which enables environmental project managers to quantify land cover changes using geospatial
datasets. It supports ecosystem extent accounting without requiring GIS expertise.
ExActR facilitates accessible and reproducible ecological assessments, with
potential to expand into a comprehensive tool for ecosystem accounting.
Our final contribution investigates the use of a lightweight CNN to classify
sparse vocalisations of the invasive small Indian mongoose in Okinawa using audio
from camera trap videos. The classifier was then applied to a large acoustic dataset
to gain insights into the distribution of the mongoose across Okinawa and aiding
conservation efforts
Machine Learning and Device’s Neighborhood-Enabled Fusion Algorithm for the Internet of Things
Full text linkIn the Internet of Things, information fusion is among the crucial problems and probably occurs due to the dense deployment of consumer electronic devices. In the literature, various methodologies have been developed to fine-tune raw data; however, consumer electronic devices’ neighborhood information has been completely ignored. In this manuscript, a machine learning and neighborhood-assisted fusion approach has been developed for consumer electronic devices to ensure that captured data values have been properly refined before onward processing at the respective edge. In this approach, every server accepts member request invitations from electronic devices deployed in its coverage area. It applies the well-known K-mean and supports vector machine (SVM) algorithms to refine captured data values by consumer electronic devices. Apart from that, the server module has the built-in intelligence to compare the captured data values of those electronic devices, which reside nearby and probably have a higher redundancy ratio. Simulation results have concluded that the proposed machine learning-assisted fusion approach is an ideal solution for the IoT in general and the Artificial Intelligent-enabled IoT in particular. Additionally, the proposed algorithm was thoroughly examined via various performance evaluation metrics such as lifetime, energy efficiency, and refinement ratio, where it has shown convincing results such as 30% improvement in the fusion ratio
Knee pain associated with bone–patellar tendon–bone autografts does not limit activity levels, sports participation or quality of life after ACL reconstruction
Full text linkPurpose: Bone–patellar tendon–bone (BPTB) and Hamstring (HT) autografts are commonly used for anterior cruciate ligament reconstruction (ACLR). Concerns exist regarding postoperative anterior knee pain (AKP) and kneeling discomfort with BPTB grafts. However, many studies solely report the presence/absence of anterior knee pain, without assessing its clinical significance in terms of functional limitation or impact on quality of life.
Methods: This study prospectively analysed 1407 patients undergoing primary ACLR with BPTB or HT autografts. Knee pain prevalence, severity, and location were measured at 6 months, 1 year, 2 years, and 5 years postoperatively using a pain questionnaire. Patient‐reported measures (Knee Injury and Osteoarthritis Outcome Score [KOOS], Western Ontario and McMaster Universities Osteoarthritis Index [WOMAC], International Knee Documentation Committee [IKDC] and Marx) and return to play (RTP) rates were also collected to evaluate knee symptoms, function and activity levels. Multivariable regression identified factors associated with knee pain at each time point.
Results: The mean age was 24.5 ± 7.1 years, with 74.3% male. BPTB grafts were used in 81% ( n = 1145) and HT in 19% ( n = 262). At 6 months, the BPTB group reported a higher prevalence of AKP (26% vs. 6%, p < 0.001). There was no difference between graft types at 1 year and 2 years postoperatively. At 5 years, the BPTB group were 1.59 times more likely to report pain, although most pain was mild and there was no significant differences in KOOS, WOMAC, IKDC, Marx scores or RTP rates. Female patients (OR 1.41, p < 0.035) and BPTB grafts (OR 1.78, p < 0.004) were associated with knee pain at 6 months. At 5 years, older age (OR 1.06, p < 0.001), BPTB grafts (OR 1.59, p < 0.027), and medial femoral condyle chondral pathology (OR 1.7, p < 0.020) increased the odds of having pain.
Conclusion: BPTB grafts are associated with early AKP, which improves over time. AKP related to BPTB is mild and does not affect activity levels, sports participation or quality of life. Mild AKP should not deter surgeons from using BPTB autografts for ACLR, given the other advantages of this graft choice.
Level of Evidence: Level II, prospective study
Responsible research impact: Ethics for making a difference
Full text linkThe need for ethical guidelines that support and empower researchers who aim to enhance the societal impact of research has become critical. Recognizing the growing emphasis on research impact by governments and funding bodies worldwide, this article investigates the often overlooked ethical dimensions of generating and evaluating research impact. We focus on ethical issues and practices that are specific to the process of intentionally working to develop societal impacts from research. We highlight the complexities and ethical dilemmas encountered when researchers engage with non-academic groups, such as policymakers, industries, and local communities. Through a combination of literature review and insights from participatory workshops, the article identifies key issues and offers a new ethical framework for responsible research impact. This framework aims to guide researchers and institutions through the process of limiting potential harm while delivering societal benefits in a way that is realistic and balanced. The aim is to establish ethical practices for engagement and impact, without making the process so onerous that researchers are less likely to undertake such activities. The article concludes with actionable recommendations for policymakers, research funders, research performing organizations, institutional review boards and/or ethics committees, and individual researchers. Making use of such recommendations can foster an ethically responsible approach to research impact across academic disciplines
Weakly interacting species as drivers of ecological stability
Full text linkDetermining how individual species can act to moderate the stability of entire ecosystems is a pressing challenge in a world undergoing rapid environmental change. Here, we show that even very weakly interacting species with no discernible effect on ecological dynamics can contribute substantially to ecosystem stability. Further, the nature of this contribution depends on biotic context, and both the type and complexity of interspecific interactions in the community. By manipulating multitrophic aquatic microcosm communities experimentally, we found that the contributions of a bacteriophage parasite to overall system stability following a pulse perturbation were variously stabilizing, destabilizing and neutral, depending on the presence of competitor or predator species of its bacterial host. This was despite the phage itself having no detectable effect on the biomass or growth rates of its host. Our results demonstrate the pivotal importance of both weak and indirect interactions in moderating the stability of whole ecological networks, and have profound implications for our ability to predict the consequences of perturbations on ecosystems