48080 research outputs found
Sort by
Intrinsically Safe Materials for Impact-Resistant Lithium-ion Batteries
No full textLithium-ion batteries (LIBs) are extensively employed in various applications, including powering electronics, mobile phones, laptops, and tablets. They are also used in electric vehicles, and in the storage of renewable energy produced from intermittent sources like solar and wind powers. However, they face a significant safety challenge: thermal runaway triggered by mechanical abuse, which poses a risk of catastrophic consequences. It highlights the urgent need to develop intrinsically safe materials to prevent thermal runaway when subjected to mechanical abuse. In this thesis, three impact-resistant components have been developed to reinforce the safety of LIBs. They include shear thickening electrolyte, Kevlar electrodes, and Kevlar separator, all of which enhance the battery’s resilience to mechanical abuse while maintaining electrochemical performance.In the first work, a novel shear thickening electrolyte (STE) has been produced for impact-resistant LIBs. It is realised by integrating poly(ethylene oxide) (PEO) with silica nanoparticles within a common liquid electrolyte. This STE features low viscosity, high ionic conductivity, long-term stability, and impact resistance. Silica particles interact with polymer chains to facilitate the formation of hydroclusters for shear thickening effect at a relative low weight percentage (2.2 wt%) of filler. This low viscosity STE, with the added benefit of lightweight, holds the potential to realise impact-resistant and lightweight LIBs.In the second work, a highly impact-resistant Kevlar-based LIBs has been developed to mitigate thermal runaway against mechanical abuse. Kevlar electrodes are fabricated by coating active materials onto metallic Kevlar fabrics. Their compatibility with liquid electrolyte and STE enables them to achieve comparable electrochemical performance to foil-based electrodes. These electrodes can effectively dissipate impact energy by inheriting the strong mechanical properties of Kevlar fabrics. Their integration with shear thickening electrolytes further improves the impact resistance by leveraging the induced shear thickening effect and increased yarn to yarn friction between Kevlar fabrics. This integration resembles the configuration as liquid body armour.In the third work, a porous Kevlar separator has been developed for impact-resistant LIBs. This is achieved by introducing a porogen into the aqueous precursor solution through a non-solvent evaporation and vacuum-assisted filtration technique. The porogen has a high boiling point which allows sufficient time for the formation of a porous structure within the cross-linked polymer network when evaporating the solvent. The resulting Kevlar separator exhibits high ionic conductivity, excellent thermal stability, flame retardancy, and good electrolyte wettability. Its efficiency is evidenced by the delivered excellent electrochemical performance and good impact resistance.In summary, three key components including shear thickening electrolyte, the mechanically strong Kevlar electrodes and Kevlar separator, have been explored to reinforce the impact resistance of LIBs. This research work systematically investigates their impact resistance and electrochemical performance to pave a new path in the development of safe LIBs.</p
Developing interactive animations as a learning tool for work integrated learning
No full textThe role of field education for social work students is fundamental to linking students' theory to practice. Field education refers to the practical component of educational programmes where students gain hands-on experience in their field of study. In Australia, field education is 1000 hours and is an essential part of social work training, allowing students to apply theoretical knowledge in real-world settings under supervision. This practical experience is crucial for developing the skills and competencies required for professional practice. Integral to the success of the field education placement is the relationship between the student and their supervisor; however, challenges have been identified in the difficulty students face when commencing placement and developing these relationships. Recent advances in technology have afforded educators the opportunity to engage students in new ways to develop the skills and competencies they require for job readiness. This study used a four-stage mixed-method approach to conceptualise and design a suite of interactive animations (IA) in consultation with students, social work practitioners and animators to investigate how IA could impact work integrated learning (WIL) success. Mixed-method data collected from 113 undergraduate and postgraduate social work students found that using IA supported students' understanding about field education relationships and engaged students in developing the ongoing practices that will support them into their future social work careers. The design features that were well regarded by students included interactive questions, self-paced learning options and embedded hyperlinks. Conversely, the element that was less supported included the use of animal characters to represent diversity.</p
Literature survey on machine learning techniques for enhancing accuracy of myoelectric hand gesture recognition in real-world prosthetic hand control
No full textThe human hand, essential for performing daily tasks and facilitating social interaction, is indispensable to everyday life. Millions worldwide experience varying levels of amputation, profoundly affecting their physical, emotional, and psychological well-being, limiting independence, and reducing quality of life. Myoelectric prosthetics, the most advanced active prosthetic hands, use surface electromyography (sEMG) signals and pattern recognition to translate user intentions into control signals. Despite these advancements, high rejection rates persist due to the non-stationarity of sEMG signals, leading to inconsistent and often frustrating user experiences. As a result, clinical and academic research has increasingly focused on improving myoelectric hand gesture recognition under real-world conditions to reduce rejection rates and enhance user acceptance of myoelectric prostheses. Given the vast and diverse range of methods applied in previous research, this survey aims to systematically highlight key studies and provide an overview of the field's current achievements. Furthermore, research on machine learning for myoelectric hand gesture recognition has been largely influenced by unrelated fields of computer science, such as computer vision and natural language processing. However, myoelectric hand gesture recognition presents unique challenges, particularly severe and unpredictable covariate shifts in sEMG signals, which require specialized approaches. To address these challenges, we propose a new taxonomy for categorizing machine learning models based on feature extraction methods and decision boundary strategies. Additionally, this paper highlights the need for benchmark datasets that accurately reflect real-world conditions and emphasizes the importance of re-evaluating real-time performance, particularly when using long temporal contextual windows. This study concludes with research challenges and future research directions to enhance the accuracy of myoelectric hand gesture recognition using machine learning techniques.</p
Recursive Convolutional Blocks: Analysis and Insights for Future Neural Network Development
No full textA common approach to improving the performance of Convolutional Neural Networks (CNNs) is by adding Convolutional Layers (CLs). This increases their physical depth enabling CNNs to model a larger receptive field and to leverage from a hierarchical representation of the features. However, this approach has several disadvantages. Not only does it lead to an increased number of parameters, computational overhead and risk of overfitting, this phenomenon also diverges from our understanding of the humans’ visual system which relies on few cortical areas and includes feedback connections. In contrast, creating a Recursive Convolutional Layer (RCL) by adding a feedback loop over a CL offers a compelling alternative. By relying on weight sharing, these RCLs can unfold into an arbitrarily long sequence of CLs without additional parameters, effectively increasing the number of feature interactions, which we refer to as the apparent depth. As a consequence, earlier works showed that RCLs perform well in shallow architectures but face challenges in deeper networks, giving rise to the opinion that they are better when combined with a gating mechanism.The current literature on RCLs lacks consistency in their experimental settings, making it challenging to consolidate and unify the existing findings. This inconsistency stems from a predominant focus on the optimisation of the performances which obscured the understanding of RCLs and the trade-offs that are involved. Our research addresses this limitation by focusing on a systematic investigation of the underlying mechanisms responsible for the behaviour of RCLs. Specifically, our study relies on non-gated RCLs with a particular emphasis on apparent depth with weight sharing, providing a more comprehensive and fair comparison of feedforward and recursive models. This is achieved through a novel approach to transform feedforward models into recursive alternatives: rather than replacing individual CLs, we substitute entire stages of CNNs consisting of repeated CLs blocks with a single non-gated RCL while preserving the apparent depth. Through a combination of empirical results and feature analysis, we investigate the behaviour of recursive models in order to achieve a better understanding and consequently to derive conditions under which RCLs can be beneficial when it is introduced in feedforward CNNs.Our findings indicate that RCLs effectively simulate the behaviour of their physically much deeper feedforward counterpart. Specifically, increasing the unfolding depth expands the effective receptive field which is combined with a gradually more apparent reliance on centralised and global patterns to assert the prediction. This process mirrors the effect of sequences of CLs in the evolution of the feature space, progressively refining their representations despite using weight sharing. Thus, when the apparent depth and the number of parameters is matched, weight sharing capture more features per layer, which makes RCLs more flexible for diverse tasks. As a consequence, we find that in a large majority of circumstances, the recursive non-dense models maintain or improve the performances, particularly in settings with limited number of parameters over that of feedforward models. Despite these advantages, we identify several cases where recursion may degrade performances. While RCLs are particularly beneficial when used with lightweight convolutional blocks, RCLs leverage less from dense connections and from multiscale feature extraction when the model is large. Furthermore, when compared to their feedforward counterpart, they are found to be more sensitive to overfitting, and excessive unfolding can degrade performance and unnecessarily increase computational overhead.From the results of this study, we develop an algorithm to automate the process of transforming a feedforward network into an efficient recursive variant. The algorithm exploits the strengths of RCLs, mitigates their limitations, and provides a foundation for future studies. By demonstrating the potential of recursive models to maintain performance with fewer parameters in controlled settings, this research opens new avenues for developing reliable and more efficient alternative to feedforward CNNs, offering a promising solution for resource-constrained environments and numerous other applications.</p
Navigating blue colonialism: the case of large-scale marine protected areas in overseas territories
No full textOverseas territories (OTs) are critical ocean stakeholders, encompassing some of the world’s largest exclusive economic zones and marine protected areas (MPAs). Despite their significance in global ocean governance and conservation, there has been minimal study and analysis of the establishment of large-scale marine protected areas (LSMPAs) in OTs, either in academic discourse or in policy debates. This gap is particularly salient given the paucity of studies examining the influence of colonialism in this context, despite it being a pervasive thread running through the past and present of OTs and protected areas alike. The objective of this research is to explore the relationships between colonialism and the implementation and governance of LSMPAs in OTs, along with their equity ramifications. This thesis addresses the gap in study and analysis noted above by asking the following question: To what extent do LSMPAs established in OTs perpetuate colonial power imbalances and inequities? To tackle this overarching question, the study examines: the motivations of metropolitan States in the creation of LSMPAs in their overseas territories; how colonial power dynamics are reflected in the governance structures and decision-making processes of LSMPAs in these territories; and the ways in which the establishment of LSMPAs in OTs may challenge or reinforce the empowerment of these territories and the self-determination of their local and indigenous populations.The thesis adopts a mixed qualitative methods approach, combining a literature review, an in-depth case study of the Parc Naturel de la Mer de Corail (PNMC) [Natural Park of the Coral Sea], an LSMPA in New Caledonia, and semi-structured interviews with key stakeholders involved in the creation and/or governance of the MPA. Collected data is interpreted using a thematic analysis, identifying patterns, trends, and relationships within the data. The working hypothesis of this thesis, based on reviewed literature, was that LSMPAs in OTs would predominantly reflect colonial power dynamics, reinforcing geopolitical interests of (ex)colonial powers and perpetuating local inequities embedded in the protected areas conservation model. This prompted the introduction of the concept of ‘blue colonialism’, to describe the persistence of metropolitan States’ colonial biases and agendas in marine conservation policies. Case study results revealed a more nuanced understanding of the intersection between marine conservation and colonial legacies in the context of the New Caledonian MPA.The findings of this thesis demonstrate that while LSMPAs often mirror colonial practices, such as the marginalisation of indigenous voices, they also present OTs and their coastal communities with opportunities to challenge such inequities and advance their local interests and priorities, in and through ocean conservation policies. As international targets like 30x30 gain momentum and LSMPAs expand, this thesis shows that although these tools are influenced by colonial legacies that need addressing to prevent local inequities, they also hold potential for local empowerment and more equitable marine governance.</p
Information and Communication Technologies for Fisheries Conservation and Management: Developing and Exploring a New Research Intersection
No full textAddressing the contemporary information and communication needs of formal marine capture fisheries conservation and management (FCM) is crucial for upholding international fisheries law and policy norms for responsible FCM practices. Using electronic information and communication technologies (ICT) could facilitate the production and flows of FCM-relevant information, thus enhancing FCM actor capabilities. However, knowledge about how ICT are being used to address FCM challenges, and about the socio-ecological impacts of these uses, remains highly fragmented. This knowledge fragmentation exists at a time of heightened global interest in ocean resources, and it obstructs a balanced understanding of the practical opportunities, challenges and consequences of using ICT to achieve responsible FCM’s diverse objectives.This thesis research aims to expand knowledge on ICT uses and their marine socio-ecological impacts in applied FCM contexts. It also aims to highlight the practical and theoretical value of examining the complex relationships between ICT uses, FCM practices, and their outcomes. To achieve these aims, this thesis uses mixed methodological approaches to explore the relationship between ICT and legally-informed, responsible FCM practices. Identifying the need to develop a research space in which to begin consolidating, analysing, and discussing relevant knowledge, this research initiates novel conceptual and theoretical boundary work to build a working ‘identity space’ for a new research intersection that it calls ‘ICT4FCM’ research (i.e., information and communication technologies for fisheries conservation and management research). Undertaking this work also generates an interdisciplinary framework that guides the thesis research and has utility for future research.Using a global-scale consolidation and analysis of current scholarly literature, the research goes on to identify strengths and gaps in topical ‘ICT4FCM’ research knowledge. Based on findings, the research then addresses selected knowledge gaps by refining its focus to the Oceania region to generate new, contextually grounded, empirical insights about ICT’s current uses, roles, contributions and impacts in Oceanian FCM contexts.Research findings reveal that ICT, despite their decades-long integration into FCM processes, lack a widely recognised definition in the scholarly literature. However, they are broadly conceptualised as a dynamic group of electronic devices and resources that people use to facilitate the production and flows of information across FCM processes and to drive socio-ecological change. The global ‘ICT revolution’ has extended and scaled researcher and practitioner capabilities and insights, human oversight of marine ecosystems and fishing peoples, and fuelled global optimism about the future of ICT-enabled FCM practices. Yet, major knowledge gaps remain, particularly concerning ICT’s applied socio-ecological outcomes. Existing ICT4FCM research remains predominantly scientifically and technically focussed, often assuming rather than critically assessing ICT’s broader impacts.In Oceania, where supporting socio-technical systems for ICT adoption and diffusion into FCM are still developing, these knowledge gaps have significant implications. Limited research on ICT4FCM in Global South contexts and a general disregard for ICT’s socio-political dimensions obscure understanding of how ICT’s integration is reshaping fisheries governance, stakeholder relationships, decision-making, and FCM outcomes in Oceania. Evidence of unequal access to socio-technical expertise and resources suggests that ICT-enabled shifts in the region’s FCM practices may disproportionately benefit more socio-technically ‘developed’ actors while limiting the agency of Pacific Island Countries and Territories (PICTs), notably Small Island Developing States (SIDS), to exercise independent control over their fisheries resources and governance narratives, to effectively advocate for their interests, and to derive long-term socio-economic fisheries benefits.This research concludes that a more interdisciplinary, context-sensitive approach to ICT4FCM research is needed to better capture the complexities of ICT’s role in FCM. By developing a more structured space for ICT4FCM research, this research provides a foundation for future scholarship while offering practical knowledge benefits for FCM practitioners in planning, decision-making, and policy evaluation.</p
Self-regulated learning, satisfaction with distance learning, and learning outcomes among female gifted high school students
No full textThis study investigates the relationships between self-regulated learning (SRL), satisfaction with distance learning, and learning outcomes among female gifted high school students in Saudi Arabia, addressing a notable gap in the literature on K-12 gifted learners in distance learning. While prior research has extensively explored SRL, its role specifically within gifted distance learners remained underexamined. Additionally, student satisfaction has been studied at the university level but has not been fully examined in relation to learning outcomes and SRL among gifted high school students. By integrating these three variables, this research offers new insights into the dynamics of distance learning for gifted students.The study sample included 311 gifted female students who completed a four-part questionnaire developed by the researcher from validated instruments. The questionnaire collected general information, assessed SRL, students’ perceived learning outcomes, and satisfaction with distance learning. Structural equation modelling (SEM) was employed to analyse the data. SEM, chosen for its confirmatory approach and ability to model complex relationships, enabled the assessment of both direct and indirect effects among variables.</p
Safe Adaptation of Foundation Large Language Models for Clinical Information Extraction in Aged Care: Evaluating Optimal Methods
No full textThe abstract for this item has not been populated.</p
Marking time: griefbots, world-time, and the phenomenology of bereavement
No full textBereavement has long been recognised as an experience of profound temporal disruption, unravelling and even severing the connections, meanings, and strategies we habitually use in navigating time. The technologies that are involved in our grieving and mourning processes can therefore also be temporal technologies; our encounters with them can both shape and be shaped by our entanglement in the temporalities of loss. The recent phenomenon of so-called ‘griefbots’ or ‘deathbots’ – Large Language Model-powered chatbots trained to imitate the dead – is no different. This paper explores some of the implications that interactions with such bots might have for a bereaved person’s experience of time by turning to Heidegger’s conceptualisation of world-time. This framework is particularly useful for thinking about both intimacy and loss, and can shed light on both beneficial and deleterious consequences of griefbot usage. In particular, such an approach can show that the dangers of griefbots usage lie less in the potential for the bereaved being deceived, and more in the reductive perspective that these bots can encourage. Usage of, and even temporary reliance upon, a griefbot can generate attitudes with a protective effect, similar to the function that Lisa Bortolotti ascribes to delusions. However, griefbots can also interfere with the temporal unfolding of grief and mourning by disrupting the way that loss tends to transform our embedding within the dynamics of world-time.</p
Extracting symptoms of agitation in dementia from free-text nursing notes using deep learning and transfer learning in natural language processing
No full textNursing staff record observations about older people under their care in free-text nursing notes. These notes contain older people's care needs, disease symptoms, frequency of symptom occurrence, nursing actions, etc. Therefore, it is vital to develop a technique to uncover the important data from these notes. This study developed and evaluated a deep learning and transfer learning-based named entity recognition (NER) model for extracting symptoms of agitation in dementia from the nursing notes. We employed a Clinical BioBERT model for word embedding. Then we applied bidirectional long-short-term memory (BiLSTM) and conditional random field (CRF) models for NER on nursing notes from the Australian residential aged care facilities. The proposed NER model achieves satisfactory performance in extracting symptoms of agitation in dementia with 75% F1 score and 78% accuracy. We will further develop machine learning models to recommend the optimal nursing actions to manage agitation.</p