Sydney eScholarship
Not a member yet
31878 research outputs found
Sort by
Computational case-based redesign for people with ability impairment: Rethinking, reuse and redesign learning for home modification practice
Full text linkHome modification practice for people with impairments of ability involves redesigning existing residential environments as distinct from the creation of a new dwelling. A redesigner alters existing structures, fittings and fixtures to better meet the occupant's ability requirements. While research on case-based design reasoning and healthcare informatics are well documented, the reasoning and process of redesign and its integration with individual human functional abilities remains poorly understood. Developing a means of capturing redesign knowledge in the form of case documentation online provides a means for integrating and learning from individual case-based redesign episodes where assessment and interventions are naturally linked. A key aim of the research outlined in this thesis was to gain a better understanding of the redesign of spaces for individual human ability with the view to computational modelling. Consequently, the foundational knowledge underpinning the model development includes design, redesign, case-based building design and human functional ability. Case-based redesign as proposed within the thesis, is a method for capturing the redesign context, the residential environment, the modification and the transformational knowledge involved in the redesign. Computational simulation methods are traditionally field dependent. Consequently, part of the research undertaken within this thesis involved the development of a framework for analysing cases within an online case-studies library to validate redesign for individuals and a method of acquiring reuse information so as to be able to estimate the redesign needs of a given population based on either their environment or ability profile. As home modification for people with functional impairments was a novel application field, an explorative action-based methodological approach using computational modelling was needed to underpin a case-based reasoning method. The action-based method involved a process of articulating and examining existing knowledge, suggesting new case-based computational practices, and evaluating the results. This cyclic process led to an improvement cycle that included theory, computational tool development and practical application. The rapid explosion of protocols and online redesign communities that utilise Web technologies meant that a web-based prototype capable of acquiring cases directly from home modification practitioners online and in context was both desirable and achievable. The first online version in 1998-99, encoded home modification redesigns using static WebPages and hyperlinks. This motivated the full-scale more dynamic and robust HMMinfo casestudies prototype whose action-based development is detailed within this thesis. The home modification casestudies library results from the development and integration of a novel case-based redesign model in combination with a Human- Activity-Space computational ontology. These two models are then integrated into a relational database design to enable online case acquisition, browsing, case reuse and redesign learning. The application of the redesign ontology illustrates case reuse and learning, and presents some of the implementation issues and their resolution. Original contributions resulting from this work include: extending case-based design theory to encompass redesign and redesign models, distinguishing the importance of human ability in redesign and the development of the Human-Activity-Space ontology. Additionally all data models were combined and their associated inter-relationships evaluated within a prototype made available to redesign practitioners. v Reflective and practitioner based evaluation contributed enhanced understanding of redesign case contribution dynamics in an online environment. Feedback from redesign practitioners indicated that gaining informed consent to share cases from consumers of home modification and maintenance services, in combination with the additional time required to document a case online, and reticence to go public for fear of critical feedback, all contributed to a less than expected case library growth. This is despite considerable interest in the HMMinfo casestudies website as evidenced by web usage statistics. Additionally the redesign model described in this thesis has practical implications for all design practitioners and educators who seek to create new work by reinterpreting, reconstructing and redesigning spaces
Designing Affective External Human-Machine Interfaces for Autonomous Vehicles to Interact with Pedestrians in Shared Spaces
No full textPedestrians are increasingly encountering autonomous agents in modern cities powered by artificial intelligence and robotics. This includes autonomous vehicles (AVs) that interact with pedestrians in various traffic situations. Significant research has explored how AVs can communicate, cooperate, and coexist with pedestrians without human drivers. In particular, external human-machine interfaces (eHMIs) have shown promise in facilitating interactions between AVs and pedestrians. Although many communication strategies for eHMIs have been studied, the potential of affective interfaces -- those designed to convey or influence emotions -- remains open for exploration. Affective interfaces, known for their abilities to enhance the perception towards intelligent systems and the experience of users, have the potential to assist AV-pedestrian interactions.
This doctoral research proposes affective external human-machine interfaces (affective eHMIs) as a sub-category of eHMIs. The overarching aim is to understand how affective eHMIs can improve the interaction of AVs with pedestrians. This thesis achieves this aim by first reporting on background research, which includes a field observation of naturalistic pedestrian-vehicle interactions in urban shared spaces, a systematic review of emotionally expressive non-humanoid robots, and focus groups with domain experts to explore relevant use cases. This phase seeks to identify potential situations and methods for designing affective eHMIs. Following this, the thesis details a research-through-design phase where affective eHMIs were designed, prototyped, and evaluated through two empirical case studies related to urban shared spaces: one focusing on the emotional expressions of AVs and the other on the affective states of pedestrians. Furthermore, in advancing virtual reality test environments for eHMIs, this thesis reports on techniques and considerations for prototyping affective eHMIs in VR, validated by empirical studies
Surrogate Bodies: How contemporary art practice can communicate the lived experience of chronic pain.
No full textThis research project explores how contemporary art practice can deepen our understanding of embodied experiences of persistent pain and illness. Through installation-based practice, I aim to make visible my intangible experiences of living with endometriosis. Kate Seear identifies endometriosis as a disease characterised by uncertainty, a trait that permeates its aetiology, treatments, health outcomes, and broader discourse. Rather than attempting to define an essence of the disease, which risks losing sight of its vast and multifaceted scope, my research sits within the questioning of it, embracing its abstract and fluctuating qualities to more accurately articulate my embodied experience.
Informed by the philosophical and phenomenological perspectives of Drew Leder and S. Kay Toombs, among others, this research examines how pain and illness transform lived experience, reshaping not only one’s physicality but one’s identity and self-conception. Additionally, I turn to Elaine Scarry’s understandings of 'the weapon' and 'the wound' as primary metaphors for articulating embodied notions of pain. We can observe these metaphors playing out within the practices of Chiharu Shiota and Rebecca Horn, who further utilise the element of tension – which has come to play a central role in my own practice – employed as a structural tool, a conceptual and theoretical underpinning, and a conduit for conveying feeling itself. Lastly, insights from Eugenie Lee and Jill Bennett highlight the value of affect and heightened bodily awareness as a lasting outcome within the viewer, particularly within socially engaged practice.
In foregrounding the feelings that surround persistent pain and illness, my studio outcomes aim to hold space for difficult emotions and sensations, evoking a heightened bodily response in the viewer. The outcomes of this research aim to connect people in shared pain and illness journeys, whilst raising awareness and education within the broader community
Direct Numerical Simulation of Shear-Induced Transitional Mixing and its Interplay with Combustion
No full textTurbulent mixing is central to advanced energy and propulsion technologies, yet its complexity presents challenges for control and prediction. This complexity stems from a high degree of interdependence of the different underlying physical processes and scales of fluid motion and is further exacerbated when the mixing process is afflicted by rapid compositional and fluid property changes such as with combustion or nuclear fusion. Prominent advancements in the field were mostly through experimental investigations. However, with the highly chaotic nature of turbulence and the limited diagnostics available to collect comprehensive data sets from experiments, additional significant advancements in the field stagnated. The recent improvement of computing and storage facilities enable through numerical simulations, access to comprehensive flow field information and the opportunity to artificially isolate elements of interdependent physics.
This thesis leverages High-Performance Computing to investigate how fundamental shear instabilities, namely Kelvin-Helmholtz Instabilities (KHI), drive turbulence generation and mixing, offering new insights for turbulence and combustion control. Through high-fidelity simulations, the role of initial perturbation parameters on flow and mixing evolution in both inert and reacting flows are explored and demonstrated to exhibit wide-reaching implications on both the flow and the mixing development well beyond the initial stages of flow perturbation. The findings challenge the traditional sole reliance on the Reynolds number to characterize the extent of turbulence and reveals new sensitivities of the flow evolution that extend to the stabilised flame structure. These results pave the way for more accurate predictions and optimized designs to control turbulent mixing in cutting-edge energy and propulsion technologies
Radiological Markers of Sarcopenia in Colorectal Cancer Surgery – A Critical Appraisal
No full textThis thesis explores the field of sarcopenia, its association with overall survival in patients undergoing resection of colorectal cancer, challenges in diagnosis, and impediments translating research to diagnostic tools in clinical practice.
Since the conception and evolution of the term ‘sarcopenia’ in the 1980s and 90s, understanding of the pathophysiology and health outcomes related to the condition have deepened, and it is now recognised as a disease characterised by decreased muscle mass and quality, decreased strength, and poor physical performance. In recent years, sarcopenia has been associated with poor outcomes in patients treated for colorectal cancer. Research has focused on computed tomography as the primary diagnostic modality. However, despite the body of evidence promoting this modality, routine screening for sarcopenia in patients undergoing colorectal cancer surgery has not translated into clinical practice.
Chapter 2, investigates skeletal muscle index, the most widely researched marker of sarcopenia, finding a strong association with overall survival. In addition, this Chapter highlights difficulty finding a universally agreed threshold to enable prospective application of skeletal muscle index as a diagnostic tool.
In Chapter 3, alternative parameters of sarcopenia are promoted and strong associations with overall survival . This Chapter investigates the pragmatism of these alternative parameters, finding some to be less complex to measure while maintaining prognostic value; this may help facilitate screening for sarcopenia.
This thesis adds to the evidence that sarcopenia is associated with poorer overall survival in patients undergoing resection of colorectal cancer, and reiterates the challenges in identifying a clinically-applicable and universally accepted threshold to diagnose the disease. The results of this thesis are instrumental to further research on sarcopenia, its diagnosis and understanding its associated poor health outcomes
Improving the performance and sustainability of additively manufactured AlSi7Mg
No full textAdditive manufacturing represents an increasingly popular group of technologies used to manufacture components in a way that reduces material waste while allowing for a high degree of customisation. Powder bed fusion (PBF) is one such technology used to process metals, allowing significant potential for property control by the manipulation of processing parameters. The popularity of aluminium alloys in a wide range of applications has resulted in its adoption as one of the most popular metals for PBF. Given the excellent processability of aluminium-silicon alloys during PBF these have seen some of the most attention, with AlSi7Mg one of the more popular.
While PBF technology has seen significant development in recent times, its complex nature means that it has not yet reached full maturation. Many processing parameters are relevant to part properties, and great focus has already been placed on laser parameters. This leaves many parameters yet to be fully elucidated and therefore requiring significant investigative work, with the location of parts within the build volume one such parameter. Heat treatments are also commonly applied to additively manufactured parts, however, these are not often tailored for additively manufactured components. The nature of the powder spreading in PBF techniques also raises the question of powder reuse, as how this powder is reused will influence both the behaviour of the powder in the build chamber and the properties of parts manufactured in subsequent builds.
This work aims to improve the understanding of these factors on the properties of additively manufactured AlSi7Mg, by investigating the effects of part placement within the build volume, tailoring heat treatments for additively manufactured AlSi7Mg, and developing a robust powder reuse method justified by more than just ease of powder handling. In combination, these investigations will improve both the part performance and sustainability of this additively manufactured alloy
Efficient Deep Neural Architecture Design and Training
No full textDeep learning has revolutionized numerous fields, but its success is often hindered by computational inefficiency, reliance on vast labeled datasets, and challenges in designing optimal architectures. This thesis addresses these issues through contributions in four key areas: handcrafted efficient architecture design, automatic neural architecture evolution, effective knowledge distillation, and data-efficient training. First, we propose LightViT, a lightweight vision transformer, and LocalMamba, a visual state-space model, to advance handcrafted architecture design by balancing accuracy and efficiency. Second, we introduce GreedyNASv2, a method to optimize neural architecture search (NAS), and DyRep, a dynamic re-parameterization framework for evolving architectures during training. Third, our work on knowledge distillation includes DIST for improving logits-based distillation, MasKD for feature-level distillation via adaptive masks, and DiffKD, which unifies logit and feature distillation using diffusion models. Lastly, we tackle the challenge of data efficiency with ActGen, an active generation framework for synthesizing hard examples, and MI-MAE, a self-supervised method leveraging mutual information for masked image modeling. Together, these advancements form a cohesive framework for efficient deep learning, addressing computational, data, and architectural challenges to push the boundaries of scalable and practical machine learning systems
The Determinants of Research Culture: The Case of Indonesian Universities
No full textPursuing its ambition to create world class universities, Indonesia has adopted a publish or perish strategy, encouraging academics to write and publish extensively to institutionalize a research culture. While this approach may increase output, concerns arise about the potential diminishing returns for valuable knowledge, especially as many publications appear in peripheral or predatory journals. Historically, Indonesia's research culture has been insular, lacking the supportive, critical, and academically free environment necessary for maximizing the benefits of publishing incentives.
This thesis examines whether the publish or perish policy is transforming Indonesian social science academic culture towards being more aspirational, competitive, and meritocratic. Through a critical realist approach, I use surveys and interviews with academics from three Indonesian state universities to gather insights on research and publishing practices. My findings reveal that international publication now holds greater structural and cultural significance for academic careers. Publish or perish rules align managerial and professional logics regarding excellence but also expose tensions between teaching-research priorities, international-national focus, and intrinsic-extrinsic motivations.
Academics navigate these challenges by optimizing, satisficing, gaming, or avoiding international publication based on their structural, cultural, and agential positions. The strategic choice to internationalize work is influenced by factors such as workload, low pay, and individual professional ambitions. Despite potential improvements in productivity and quality, structural and cultural contradictions may undermine the effectiveness of the publish or perish policy. Strengthening the intellectual commons and providing time and space for critical inquiry are crucial for maximizing the benefits of publishing incentives and fostering a diverse and competitive research culture in Indonesia
Trend-cycle decomposition in the presence of large shocks
No full textWe introduce some refinements of the Beveridge-Nelson filter to help address possible distortions from large shocks. We then compare how the Beveridge-Nelson filter and other popular univariate trend-cycle decomposition methods perform given the extreme outliers associated with the Covid recession. Real-time estimates of the output gap based on the Hodrick-Prescott filter are highly unreliable in the years just prior to the pandemic, although the revised estimates during the pandemic are similar to those of the more reliable Beveridge-Nelson filter. The Hamilton filter suffers from base effects that produce a mechanical spike in the estimated output gap exactly two years after the onset of the pandemic, in line with the filter horizon. Given projected data with a simulated Covid-like shock, both the Hodrick-Prescott and Hamilton filters overstate the true reduction in the output gap and fail to capture the implied movements in trend output. The Hodrick-Prescott filter generates a spurious transitory boom just prior to the simulated shock, while the Hamilton filter produces another mechanical spike exactly two years after the simulated shock, as well as an ongoing divergence in forecasted values of the output gap away from zero. Only the Beveridge-Nelson filter correctly forecasts trend and cycle movements when faced with a Covid-like shock
Life in the big smoke: terrestrial vertebrate assemblages and their drivers along an urbanisation gradient in Sydney, Australia
No full textContext. Urban areas are rapidly expanding, increasing anthropogenic pressure on global biodiversity. There are many threats associated with urbanisation,such as habitat loss and the spread of invasive species. Thus, to effectively manage urban greenspaces for native species, we need to understand species assemblages, and the factors that influence their diversity. Aims. We assessed how assemblages of terrestrial vertebrate species differ across urbanisation levels in northern Sydney, Australia, and whether habitat characteristics influenced them. We also investigated the role of invasive species in these areas by comparing daily activity patterns between invasive and
native species. Methods. Nine forest (dry sclerophyll) patches, surrounded by three urbanisation levels (high-urban, mid-urban and low-urban), each with five motion-sensor cameras, were used to survey terrestrial fauna in northern Sydney from May to July 2023. Five vegetation variables and eight spatial variables were also recorded at each patch. Key results. There were differences in terrestrial vertebrate assemblages across the three urbanisation levels, with more invasive species, including red foxes (Vulpes vulpes) and black rats (Rattus rattus), in high-urban areas. We also found native species assemblages (particularly superb lyrebirds, Menura novaehollandiae; long-nosed bandicoots, Perameles nasuta; and bush rats, Rattus fuscipes) to differ with tall and low shrub cover, whereas invasive species did not differ. Furthermore, mean activity peaks did not differ between red foxes and a critical weight range (CWR) mammal was recorded in high-urban sites. Last, we observed a significant negative relationship between invasive species richness and native species diversity. Conclusions. Urbanisation has a strong effect on the assemblages, activity and interactions of native and invasive terrestrial vertebrate species in Sydney. Whereas remnant patches may support native species, the prevalence of invasive fauna in areas that are more highly urbanised may compound other factors affecting biodiversity. Implications. This study has provided a baseline understanding of urban terrestrial vertebrate assemblages in northern Sydney, and has shown that mid-urban areas may be good candidates for reintroduction sites. Our findings can be used to guide management actions to support the conservation or re-establishment of native species in the region or other urban areas