Showing items 1 - 20 of 812

A blockchain-based distributed peer-to-peer ecosystem for energy trading

  • Authors: Islam, Mohammad
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: Blockchain technologies are revolutionising peer-to-peer (P2P) distributed energy trading. These technologies can leverage microgrid decentralisation and immutable data storage to provide efficient and secure trading to benefit prosumers. A double auction mechanism is best suited for energy trading in a P2P microgrid. This mechanism requires a solvent cryptocurrency reserve for payment settlement. Double auctions give rise to unspent auction reservations (UARs). Existing mechanisms can settle further auctions with UARs but need improvements to do this without affecting trading efficiency. Keeping a cryptocurrency reserve solvent also requires adaptations to existing mechanisms. Auction settlements within a microgrid leave UARs, meaning that other microgrids must join for further auction settlements, and this leads to security vulnerabilities. It is important to develop an ecosystem that can enhance trading efficiency, ensure the solvency of the cryptocurrency reserve and provide security for multi-microgrid energy trading. In distributed energy trading, an auctioneer passes UARs to the next auctioneer as specified by the passing mechanism. Traditional energy trading systems use simple passing mechanisms and basic pricing mechanisms, but this adversely affects trading efficiency and buyers’ economic surplus. Traditional P2P energy trading systems use passing mechanisms that only partially consider the auction capacity of the next auctioneer. We propose a blockchain-based energy trading mechanism using a smart passing mechanism (SPM) that uses an unspent reservation profile (URP) to represent the auctioneers’ capability to pass UARs within a P2P microgrid. We further propose an intelligent passing mechanism (iPass) that incorporates price information into URPs to enhance trading efficiency. We applied three metrics to measure trading efficiency: convergence time, auction settlements and the economic surplus of buyers and sellers. We simulated our mechanisms in Hyperledger Fabric, a permissioned blockchain framework that managed the data storage and smart contracts. Experiments showed that our SPM reduces the convergence time, increases auction settlements and increases the economic surplus of buyers compared with existing mechanisms. Experiments showed that iPass is even more efficient than other passing mechanisms, including SPM, further reducing the convergence time, increasing auction settlements and increasing the economic surplus of buyers and sellers. Settling payments in blockchain-based P2P energy trading requires maintaining the solvency of the cryptocurrency reserve to ensure a stable medium of exchange and reduce price volatility. Stablecoins, as a form of cryptocurrency—the most suitable medium of exchange—are gaining attention from central banks. A consortium of central banks has recommended compliance with capital and liquidity standards for high-quality liquid assets (HQLA). Stablecoins, as a form of HQLA, require the adaptation of these standards for P2P energy trading. We propose a mechanism (NF90) to control the inflow of stablecoins in response to the liquidity coverage ratio (LCR) for reserve resilience and to maintain solvency. The Basel III Accord recommends 100% LCR. We measured the effectiveness of NF90 using LCR as a metric simulating the mechanism in Hyperledger Fabric to manage deceni tralisation, data storage and smart contracts. NF90 was the most effective inflow control mechanism. The use of iPass for a P2P microgrid leaves UARs. Traditional trading mechanisms settle further auctions with UARs within a microgrid, which affects the economic surplus of prosumers. Auction settlements with neighbouring microgrids increase prosumers’ economic surplus, but the usual pricing of double auction mechanisms reduces their economic surplus. Other pricing mechanisms are needed in a multi-microgrid paradigm. Settling auctions for microgrids requires common computational resources that are close to microgrids. Edge computing technologies suit this need, and blockchain technology leverages immutable data storage in cloud servers. However, communication with a cloud server through proprietary edge computing devices exposes the ecosystem to security vulnerabilities. It is important to control access by prosumers and forensic users. Immutable data storage and the retrieval of data are essential. Two challenges in information security are incorporating reliable access control for users and devices while granting access to confidential data for relevant users and maintaining data persistence. This research used a blockchain structure for data persistence. We propose a framework of novel protocols to authenticate users (prosumers and auctioneers) by the edge server and of the edge server by the cloud server. Our framework also provides access to forensic users using immutable blockchain-based data storage with endpoint authentication and a role-based user access control system. We simulated the framework using the Automated Validation of Internet Security Protocols and Applications and showed that it can deal effectively with several security issues.
  • Description: Doctor of Philosophy

A case study to evaluate the effectiveness of chronic disease management plan on self-management among patients with diabetes mellitus at general practice settings

  • Authors: Ghasemi, Maryam
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: The chronic disease management (CDM) plan is designed to support people with chronic medical conditions. This plan provides a targeted, comprehensive approach, allowing individuals to receive the necessary care and support to manage their condition. With this plan, patients can access various health services, including visits to their general practitioner and allied health services. These provide the necessary support and care to manage their condition effectively. However, patients with chronic conditions are often poorly served by the current Australian healthcare system, which fails to coordinate care across different service providers. Aim The primary aim of this study was to examine whether the use of CDM plans can improve self-management among patients with diabetes. Method A mixed method collective case study was undertaken. It focused on identifying patients’ predisposing and biometric factors, exploring patients’ and health professionals’ perceptions of the CDM plan and examining the CDM plan’s clinical documentation. Semi-structured interviews were undertaken among patients with diabetes and healthcare professionals in a general practice setting in Victoria, Australia. Results Three main issues emerged from the study: the rigidity of the funding model, system and organisational constraints and the lack of person-centred care. Conclusion To enhance self-management support through CDM plans, it is crucial to understand interdisciplinary and intradisciplinary interactions between patients and healthcare professionals. Organisational structures can apply a powerful contextual influence on how patients and healthcare professionals interact, but individualised CDM plans with tailored allied health services, regular follow-up and review are essential for the sustainability of health outcomes. Primary care settings and services in Australia need to be reformed to meet the needs of high-cost health users with complex chronic conditions. Moving beyond fee-for-service funding can stimulate innovation in service delivery and configuring person-centred care. Payment and funding reform is needed, particularly for people with ongoing complex needs and comorbidities. The study has highlighted the lack of a clear coordination framework guiding CDM plans, lead to inconsistency and poorer patient outcomes. Keywords: Self-management, diabetes mellitus, diabetes self-management, chronic disease management plan, person-centred care plan, integrated care plan, multidisciplinary care plan approach, Andersen Behavioural Model of Health Service Use
  • Description: Doctor of Philosophy

A qualitative study of Australian indigenous perceptions of success

  • Authors: Hamilton, Ian
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: This study explored Australian Indigenous perspectives of the concept of success. Qualitative audio interviews with semi-structured questions and a conversational approach adhered to the Indigenous research paradigm. Participants were provided with an opportunity to share their views. The welcoming nature of Indigenous communities as well as other rewarding experiences for myself were echoed during this study and examples were reported. A local Indigenous individual noticed that Indigenous perspectives of success often differed from many common ideas within a mainstream Western society. This was confirmed by the comments of eleven research participants. The literature review highlighted an earlier literary emphasis on problems rather than the positive aspects of Indigenous communities. Investigating stories related to success was an attempt to avoid adding more unpleasant material that was present in the literature. The literature review empathised with Dennis Foley researching Indigenous success in entrepreneurism and Jeannie Herbert examining success in higher education, followed by an examination of the meaning of success. This study sought generalised community views of success, rather than success in one field. Most data for this research were collected in the central Gippsland region, supplemented with three interviews at Circular Quay in Sydney. All participants indicated they lived a suburban lifestyle; 29 qualitative interviews were conducted, with 26 rated as useful and transcribed verbatim. All interviews satisfied the ethical clearance requirements and followed the advice of key authors arguing for decolonisation when completing modern research. Interviews were analysed by employing grounded theory thematic analysis, using an interpretive method with elements of deductive reasoning. Five observed themes were classified as general comments about success: success is different, success is available to all, difficulty discussing success, determination, and creativity or spirituality. Five themes were classed as definition of success: family support, contribution to community, achieving goals, happiness and continuing culture. A discussion of links between themes argued the broad range of views of participants evinced communal implications for success as well as a devaluation of materialism. Other links included the support for cultural traditions and the inadequacy of common Western definitions of the concept of success. The analysis acknowledged the overlaps between cultures that can exist when comparing Indigenous with mainstream cultures, especially when recognising individual differences. Literature identified concurred with findings from this study. Research limits included sampling restrictions and a limited time frame. Suggestions for further research included extra research be completed in more regions in Australia with more women, a broader age range and more cultural groups. Differences were observed regarding the concept of success for Indigenous individuals when compared with more common mainstream Western views as well appreciating the value of continuing the culture were two examples of the claims made by most participants. Therefore, the findings of the research resulted in recommendations for policymakers, educationists, academic researchers and the general community to converse with Indigenous groups and increase the understanding of the views for the meaning of the concept of success.
  • Description: Doctor of Philosophy

A voice for nature : a history of the Field Naturalists Club of Ballarat

  • Authors: Kruss, Susan
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: This thesis examines the Field Naturalists Club of Ballarat (FNCB) and its predecessors in a longitudinal study to understand how Ballarat’s naturalists have provided a voice for nature. The Ballarat Field Club and Science Society was formed in 1882 at the School of Mines Ballarat. It was discontinued around 1890 and then reconstituted in 1915 as the School of Mines Science and Field Naturalists’ Club, which closed in 1918. FNCB, established in 1952, regards itself as a reconstitution of the earlier Ballarat field clubs. Drawing on archival records, newspaper sources, government reports, oral histories and field excursions, this thesis uses narrative to describe the ways in which FNCB has provided advocacy for nature since its inception. This study locates FNCB in changing social and political contexts and argues that the club has provided a quietly determined, reasoned voice over many decades. A story emerges of continuity and resilience through different eras, which has continued into the present. This research is inspired by American historian Roderick Nash and his theory of environmental history as a “history from below” that speaks for the exploited natural world. It examines the achievements of a small organisation at the local level but places its operations in a global context. It argues that global policies and the environment movement formed the context for the club’s representations on behalf of the natural environment and examines how broader movements, ideas and policies affect ground-level nature conservation. It concludes that FNCB, by providing a voice for nature, made an important contribution to protection of the natural environment
  • Description: Doctor of Philosophy

Aboriginal people in the Northern Mallee Backcountry seasonal visitors? Active land managers?

  • Authors: Burch, John
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: This research uses previously unexamined sources and original methodologies to propose a new paradigm for Aboriginal peoples’ occupation of the Victorian Mallee backcountry. It refutes an historiography which saw Aboriginal people as simply ephemeral visitors to this area. In its place it demonstrates that Aboriginal people regularly occupied parts of the backcountry and posits that in those areas they practiced a unique Mallee Economy. In 1878 Robert Brough Smyth proposed that Aboriginal people were simply brief seasonal visitors to the backcountry and hence that the area was not permanently occupied. This has remained the dominant historical understanding and though it has been recently challenged, there is no agreement that it has been discredited. This thesis reviews that proposition and two research questions were formulated from the historiography as the basis for this investigation: 1. To what extent were Aboriginal people in the Northern Mallee ‘seasonal visitors’ and to what extent were they ‘permanent occupiers’? 2. What evidence is there that Aboriginal people in the Northern Mallee were ‘active land managers’? In the absence of an adequate written archive, this research adopted an original and relatively untested strategy of collecting evidence of Indigenous occupation and management of the land from cartographic sources. In the Northern Mallee this meant placing a heavy reliance on maps made to facilitate agricultural settlement, which occurred at least half a century after colonial occupation of the land. Information that was, or might be, indicative of Indigenous land use was gathered from over two hundred maps and archaeological sources and entered into a Geographic Information System (GIS) to allow its interpretation and analysis. Analysis in the GIS was supported by layers of already available data. The boundaries of major bioregional areas, vegetation groups, water features, colonial land-ownership, modern roads and towns, parish boundaries and the mapping of Indigenous land- ownership in North-Western Victoria were all added into the GIS. This research process produced varied but powerful findings that contradicted the conventional historiography. Hundreds of pieces of ‘track’ were collected from maps and when these were analysed a picture of extensive use of the backcountry was constructed. Long pathways provided avenues for trade and communication between the riverine corridors of the Murray, Darling, Murrumbidgee and Wimmera rivers. Other pathways provided connections into South Australia, whilst Wirrengren emerged as a major focus of travel into the Mallee. These ‘highways’ across the land were joined by ‘local roads’ reaching out from the numerous campsites along the densely settled Murray River. The second group of findings centred on the issue of seasonal visiting. The historiography of this understanding was subject to critical review. Brough Smyth’s proposition was placed in the context of ‘pastoral gentlemen’ seeking a rationale to remove Aboriginal people from desirable land, and the absence of early evidence of occupation was related to mapping conventions that afforded no recognition to perceived primitive hunter-gatherers. More recent pressures to conceal Indigenous heritage are also discussed. Demonstrating occupation was predicated on showing the presence of intermittently but regularly used campsites on estates similar to those occupied by Aboriginal people in arid areas. Cartographic sources proved of little value, but information was drawn from archaeological research and local histories, sufficient to show that parts of the backcountry were, within these criteria, permanently occupied. The research methodology proved more suited to the second research question, exploring Indigenous land management. Over 150 small ‘grassy plains’ were mapped in the study area and the GIS showed the location of these plains was apparently not random but appeared weighted towards a specific bioregional area. When these plains were seen in relationship to other non-grassy plains and patches of mallee eucalypts a sense of parts of the mallee as a mosaic of grassed areas and old mallee emerged. It was then posited that this mosaic created the conditions for a distinctive Mallee Economy. The small grassy plains were ideally suited to medium size mammals such as the boodie, bilby and bettongs, and also provided grain, murnong and other tuberous vegetables. The old mallee was, amongst other resources, the home to an iconic food source, Mallee Fowl eggs. Other food sources, some reptiles and the eggs of emus, were probably generally available across the mosaic. Due to the limitations in dating archaeological evidence no relationship could automatically be assumed to exist between the regularly used campsites and the Mallee Economy. But there were suggestions to support this. Most powerfully the GIS showed a strong geographical relationship between the campsites and the grassy plains. In another instance ethnohistorical evidence connected the two. Unable to be definitively answered was the question of whether the unique mosaic of grass and old mallee which allowed the Mallee Economy arose from ‘active land management’. Was the landscape created or maintained by the practice of ‘fire’ and ‘no fire’? There is only very limited direct evidence of landscape management in the backcountry, but the inference is compelling. Inference of active land management is strengthened by emerging evidence that if a regime of ‘fire’ and ‘no fire’ was not practiced the grasslands would have become senescent, if not totally overrun by mallee, and the productivity of the Mallee Economy seriously compromised. The broad findings of the research were that Aboriginal people permanently occupied parts of the Northern Mallee and may have used active land management to practice a unique Mallee Economy. The more specific finding was a map detailing the areas in which it is posited a Mallee Economy was being practiced, those areas in which it is clear it was not being practiced and other areas about which conclusions were not able to be made.
  • Description: Doctor of Philosophy

American protectionist thought: the economic philosophy and theory of the 19th century American protectionists

  • Authors: Frith, Mathew
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: This study contributes to economic knowledge by recovering the system of economic philosophy and theory developed by the 19th century American Protectionists. The account provided in this thesis is comprehensive and covers all major aspects of their theoretical system. Put simply, this study provides an explication and clarification of the economic philosophy and theory of this important, yet neglected, school of economic thought. In doing so, the approach undertaken is one which goes beyond the writings of Alexander Hamilton, Friedrich List, and Henry Charles Carey, and instead draws upon the full corpus of American Protectionist literature. By drawing upon the writings of roughly seventy economists and statesmen from within this neglected tradition, this study has been able to consolidate and distil their ideas into a general and coherent system of economic theory. This thesis therefore recovers an important and original lens through which to interpret and understand the workings of the economy. This is important for both economists and historians of economic thought alike. Whereas most literature in the history of economic thought relegate the American Protectionists to a position of insignificance, this study demonstrates that their system is a refined and sophisticated doctrine of economic thought which rivalled, and even surpassed, other 19th century schools of economic thought. Indeed, against the Whig theory of history, which presupposes that economic knowledge progresses in a linear and upward manner, this thesis contends that superior theories can often be displaced by inferior ones, with knowledge being lost as a consequence. This thesis therefore advances the view that the system of thought developed by the American Protectionists is an example of where important and relevant economic knowledge has been lost. Rediscovering this lost knowledge is thus important not only for purely historical reasons, but for its ability to aid modern economic analysis.
  • Description: Doctor of Philosophy

Changing inner narratives : exploring the influence of transformative learning on coercive control survivors enrolled in Clemente in several locations in Australia

  • Authors: Cooper, Lesley
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: Family violence is pervasive in our communities, nationally and internationally, with the emergence of coercive control as a distinct form of abuse that impacts the daily lives of victims-survivors. Coercive control permeates victims’-survivors’ lives through tactics including but not limited to isolation, manipulation, threats of violence and surveillance, reducing their identity, autonomy, and freedom to participate in society, including the ability to work or study. Abusers often utilise an incident of physical violence to instil fear in the victim-survivor, laying a foundation for coercive controlling behaviour that is invisible. This alters victims’-survivors’ cognitive processes in developing their inner narratives, which underpin their beliefs and value systems. The transformative learning process foundational to the Clemente Course in the Humanities enables students to deconstruct their inner narratives and then analyse and reconstruct new inner narratives. Clemente is a community-based, free tertiary-level course offered to the disadvantaged based on the belief that personal transformation can be achieved through fostering critical reflection. This research explores whether an interconnection exists between Clemente, transformative learning, and coercive control within a higher education context. Victim-survivor voices are central to this research, creating a unique opportunity to gain an understanding of their experiences of Clemente and coercive control. Ten participants aged between 18 and 65 who attended Clemente in New South Wales, Australian Capital Territory, and Victoria participated in this research. The research aims to answer three main questions: (1) Does the Clemente course contribute to transformative learning that empowers students who experience coercive control in a family violence context? (2) How do participants describe the transition from coercive control to education and employment? (3) ‘In what ways and to what extent, if any, has participation in the Clemente course empowered students to regain control over their lives and future life trajectories?’ Data for this narrative analysis was collected by (1) a participant intake form gathering demographic student characteristics; (2) a coercive control self-assessment tool that measured students’ perceptions of coercive control; (3) a coercive control measurement tool measuring students changed views on coercive control; (4) the transformative learning tool enabling students self-assessment of their transformational journey from commencing to finishing and (5) semi-structured interviews encouraging expansion of measurement of change responses. The coercive control self-assessment measurement tool was amended from the tools developed by Kelly et al. (2014). This research adds to knowledge in understanding victims’-survivors’ adjustment to life after coercive control. Victims-survivors appreciated studying the humanities within a social and intellectual community where they felt valued and heard. The humanities curriculum, coupled with the intensive Clemente support, appeared to positively contribute to critical thinking skills development, increased confidence, and an ability to exercise more control over their lives, and engage with the community. Victims-survivors who were parents reported feeling like role models for their children indicating the course assisted in more meaningful engagement with their children. It demonstrates that a rigorous educational opportunity that is not vocationally orientated can assist personal development for victims-survivors who aspire to careers not just employment for the sake of earning an income.
  • Description: Doctor of Philosophy, Partial

Characterisation of novel Y chromosome-linked circular RNA TTTY15 in the context of coronary artery disease

  • Authors: McInerney, Molly-Rose
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: Coronary artery disease (CAD) is the leading cause of death worldwide. Characterised as narrowing of the coronary arteries that results obstruction, lack of blood flow and ischemic injury. CAD progression is regulated by a host of complex molecular pathways pertaining to inflammation, lipid disturbance and cellular dysfunction. Recently, circular RNAs (circRNAs) have been implicated in the regulation of these CAD-associated molecular pathways. CircRNAs represent a relatively recent discovery in the realm of non-coding RNAs, challenging the initial misconception that they are ‘by-products’ of mRNA splicing. Instead, these circRNA molecules have emerged as critical regulators of gene expression in various diseases, spanning from cancers to heart disease. CircRNAs exert their regulatory functions through several mechanisms. They can act as 'miRNA sponges,' sequestering microRNAs (miRNAs) and preventing them from regulating their target genes. Additionally, circRNAs interact with RNA-binding proteins (RBPs), modulating RNA stability and localisation. Some circRNAs also function as 'protein sponges,' influencing protein expression and gene regulation. While circRNAs have gained recognition in the context of obstructive cardiovascular diseases like CAD and atherosclerosis, no circRNAs derived from the Y chromosome have been experimentally characterised in this context. The Y chromosome, typically associated with male-specific traits and sexual development, remains relatively unexplored but recently has been implicated in the regulation of human diseases, including cardiovascular diseases. The exploration of Y-linked circRNAs represents a promising avenue for advancing our understanding of male-specific diseases and may open new possibilities for personalised health interventions.
  • Description: Doctor of Philosophy

Circular economy : role of municipal waste in creating value through local processing of household glass waste

  • Authors: Aashcharya, Hanwellage
  • Date: 2024
  • Type: Text , Thesis , Masters
  • Full Text: false
  • Description: The transition to a circular economy paradigm is critical for addressing the growing challenges of waste management and resource depletion. Hence, this thesis examines a way to expand circular economy practices in Australia, with a special emphasis on the role of municipal waste management in generating value through the local processing of household glass waste in Ararat Rural City, Victoria. It integrates insights from two independent but interconnected studies to provide a multidimensional study on the circular economy and value creation through such local processing, and on glass remanufacturing feasibility, thus highlighting their synergistic potential for sustainable development. First, this thesis explores Australia’s circular economy and waste management landscape, focusing on challenges and opportunities associated with the household waste crisis. It analyses data from 520 municipalities across the six Australian states to understand waste generation, disposal methods and circular economy initiatives at the local government level. Despite initiatives such as the National Waste Policy Action Plan, disparities persist, which calls for sophisticated policy interventions and collaborative solutions. Using publicly available data, the study establishes a detailed database on waste management practices and circular economy projects. While most councils offer general waste bins, fewer councils provide bins for food organic and garden organic (FOGO) waste and glass waste. Engagement in waste management and circular economy projects varies across councils, indicating the necessity for targeted interventions and incentives to promote consistent policies nationwide. Of the 520 councils surveyed, 481 offer general waste bins, 246 provide FOGO bins and 406 supply mixed recyclable bins. However, only 14 councils, all from Victoria, have separate bins for glass waste. Nationally, 171 councils actively participate in waste management and circular economy projects, whereas 80 do not engage in either. Addressing these disparities requires targeted interventions, financial support and incentives to foster formal waste management and circular economy policies across Australia. Then, this thesis focuses on the feasibility of, and stakeholder perspectives about, creating social, environmental and economic value through establishing a glass remanufacturing plant in Ararat, Victoria. Using a qualitative approach, comprising 16 semi-structured interviews with representatives from local governments, academia, state government waste management authorities and the recycling industry, this thesis explores the economic, social and environmental benefits of glass remanufacturing. Recognising the critical significance of glass remanufacturing in promoting energy efficiency, reducing carbon emissions and ensuring environmental sustainability, this thesis assesses the feasibility of establishing a glass remanufacturing business in Ararat Rural City. Stakeholder engagement and empirical study reveal the subtle aspects of glass remanufacturing sustainability, such as resource availability, regulatory challenges and community engagement methods, and environmental advantages, such as reduction in carbon emission and landfill. The study also explores the short-term and long-term viability of establishing a glass remanufacturing business in Ararat, Victoria. It finds that in the short term (>5 years), although this business has economic feasibility, it will face challenges. Initial investments in technology, infrastructure and human resources are crucial for market penetration, yet short-term profitability relies heavily on factors such as glass waste feedstock availability, regulatory stability and consumer demand. Addressing these challenges requires coordination with local authorities and innovative waste management initiatives to ensure a consistent supply of at least 50,000 tonnes per year. Moreover, regulatory concerns and the need to attract qualified personnel underscore the importance of complementary services and amenities for workforce retention. Looking ahead to 2030, the long-term viability appears more promising, with anticipated improvements in waste management methods, technological advancements and strategic investments positioning the business for sustainable growth and competitiveness, as identified from stakeholder analysis in this study. This thesis advocates for a comprehensive strategy for circular economy advancement, with municipal waste management serving as a key driver in unlocking value from household glass waste in Ararat. By bridging the gap between policy imperatives and on-the-ground execution, local governments may lead transformative projects that not only reduce waste but also promote socioeconomic development and environmental stewardship. Keywords: Circular economy, municipal waste management, household glass waste, remanufacturing feasibility, stakeholder perspectives, sustainable development
  • Description: Masters by Research

Comparison between real and simulated driving for training and assessment

  • Authors: Thang, Nguyen
  • Date: 2024
  • Type: Text , Thesis , Masters
  • Full Text:
  • Description: Driving simulators have emerged as instrumental tools, providing secure, regulated environments for scholarly investigation, driving appraisal, and training. These simulators negate the inherent risks associated with real-world driving experiments and offer a platform for methodical, cost-efficient research. Despite these advantages, the efficacy of driving simulators remains a contentious issue in academia, primarily regarding the transferability of acquired skills to actual driving conditions and the veracity of training and evaluation results. The current study aimed to contribute to this ongoing discourse by comparing four driving experiences—three driving simulator setups (Three-monitor based, video-based, and Virtual Reality-based) and one real on-road driving scenario. The research focused on various aspects such as immersion, performance, physiology, emotion, and simulator sickness. Data were collected from a small but intensively studied sample size of two participants who engaged in all four driving conditions. Metrics such as heart rate, breath rate, speed, acceleration, as well as responses to presence, simulator sickness, and mood questionnaires were amassed. The results revealed that pre-driving mood exerted a marginal influence on participants' physiological responses in this research context. In terms of presence, the three-monitor setup received the highest ratings, followed by Virtual Reality (VR) and 3D video configurations. Intriguingly, VR was implicated in eliciting the most substantial symptoms of simulator sickness. The study also observed individual disparities in baseline physiological measurements and cognitive tasks, elucidating the intricate nature of human interaction within simulated environments. Moreover, no clear relationship was established among immersion, simulator sickness, emotion, physiology, and performance across the four driving conditions, which included three different simulation setups and one real on-road experience. While no statistically significant distinctions in performance were observed among participants, notable variances manifested across distinct speed limit zones and simulator configurations. Given the methodological limitation of only assessing mood pre-experience, the study highlights the imperative to incorporate additional contextual factors, such as mood oscillations during the driving experience, in future research endeavours to enhance our understanding of their consequent impact on performance metrics.
  • Description: Masters of Research

Data-driven flexibility assessment for demand response in wastewater treatment plant

  • Authors: Yasmin, Roksana
  • Date: 2024
  • Type: Text , Thesis , Masters
  • Full Text:
  • Description: Today’s energy system is undergoing a significant transformation, moving from conventional energy sources to renewable energy (RE) systems on both national and global scales. This transition is expected to reduce global carbon emissions and meet the increased energy demand. However, the intermittent nature of renewable energy sources (RESs) can affect power system stability which poses the need for more flexibility in the power system. Supply-side flexibility through peaking power plants is expensive, and relying on conventional generation is undesirable, which shows the urgency of flexibility from the demand-side. Demand-side flexibility through demand response (DR) is a well-accepted mechanism in which consumers change their energy consumption patterns by responding to any power system issue or need and can receive financial benefits. Unlike residential consumers commercial and industrial (C&I) consumers are extensive energy users and potential DR candidates. However, a lack of knowledge about DR implementation and benefits hinders DR participation by C&I consumers having complex industrial processes. Therefore, further research on the applicability of DR for C&I consumers including benefits evaluation is significant. Wastewater treatment plant (WWTP) is one of the C&I consumers in the water industry which accounts for about 2-3% of global electricity use. DR participation by WWTPs can deliver benefits in multiple ways: it can reduce energy costs for the plant; provide sustained stability to the power system and decarbonisation for the wider community. However, WWTP loads and processes are usually interconnected and complicated and might not be interrupted frequently to provide DR as required, which has not been extensively reviewed in past studies. Further, WWTP processes are guided by different control parameters which should be strictly maintained to provide the quality requirements. Besides, the energy consumption of a system can be associated with different parameters and variables that are linked with the loads and processes. A systematic analysis of the energy data to extract information about the key parameters and variables can give an understanding of the potential flexible loads for DR participation. Moreover, the feasibility of DR application while securing techno-economic benefits is vital to encourage and realize DR participation of WWTPs. This research initially performed a literature survey to gain an understanding of the possible means to increase flexibility by C&I consumers which can apply to WWTPs. Consumers with inflexible or restricted loads can participate in DR program with the aid of an energy storage system (ESS), which is useful in storing energy for later use. Besides, onsite renewable generation (ORG) allows consumers to use RES generation during peak demand periods, avoid high energy prices, and respond to grid pressure relief. The literature survey addresses the gaps in both C&I consumers and WWTP-focused DR surveys which can lessen the knowledge barrier for WWTP DR participation. The survey analysis exhibits that, utilising ESS and ORG C&I consumers with inflexible loads can participate in DR programs, which can be applied to WWTPs. Several recommendations are provided which are deemed critical for fruitful DR implementation using appropriate ESS and DR strategy. Afterward, a data analysis is performed using WWTP real energy data which identifies the association of energy consumption with key parameters and variables based on the correlation analysis and ANN model. A systematic approach is developed to evaluate the influence of key features including wastewater inflow and weather parameters on the WWTP energy consumption. This data analysis can provide identifying energy consumption patterns of loads/processes which is useful in assessing DR flexibility. Finally, data-driven flexibility assessment is performed using 5%, 10%, and 15% load shifting flexibility potential on hybrid energy systems (HES). Four HES combining PV, battery, and fuel cells are analysed using HOMER software. The techno-economic assessment shows that up to 29.5% energy cost savings can be obtained with load-shifting flexibility integrated into HES which provides the economic feasibility of DR-ESS-ORG integrated energy system for WWTPs. Besides, the reduction of carbon emissions by 28.3% ensured environmental benefits and explored the promising role of hydrogen-based FC. The sensitivity analysis conducted integrating bioenergy from the WWTP results in additional economic and environmental benefits. In brief, this thesis focuses on WWTP DR participation with both flexible and inflexible loads. The data characterisation analysis can assist in identifying flexible loads for DR and the application of ESS and ORG can assist in DR participation without interrupting any inflexible loads/ processes. Hence, this study provides a comprehensive approach to the DR flexibility assessment of WWTPs. Further, it contributed to addressing the contemporary issues with energy transition and provided promising solutions to respond to the energy transition challenges through WWTP DR flexibility. The application of DR in HES including PV, battery and hydrogen-based FC provides a novel energy system model for WWTPs which can result in energy cost reduction and play a key role in decarbonisation.
  • Description: Masters by Research

Data-efficient graph learning for responsible prediction and recommendation

  • Authors: Tang, Tao
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: Graph learning offers a promising approach to uncover latent complex relationships within single or multiple graphs, thereby enhancing the performance of prediction and recommendation models. However, current graph learning methods often require significant computational resources and detailed training data for optimal performance. In real-world scenarios, graph-structured data are frequently sparse, with missing attributes and errors, particularly in distributed systems. Data heterogeneity can lead to Non-IID issues (e.g., imbalanced distribution) and limited computational resources. Additionally, ethical challenges in AI systems necessitate designing user-centered algorithms that consider privacy, transparency, and responsibility. These issues can degrade model performance, underscoring the need for user-centered, data-efficient graph learning models that enhance efficiency in both centralized and decentralized systems. Considering these challenges, this research investigates data-efficient graph learning for responsible prediction and recommendation in real-world applications. In this thesis, we propose effective and efficient graph learning algorithms for three sub-tasks: (1) Federated Graph Learning on Non-IID EHRs, (2) Multi-view Graph Learning on Sparse EHRs, and (3) Federated Graph Learning for Spatiotemporal Recommendation. Extracting latent disease patterns from Electronic Health Records (EHRs) is crucial for disease analysis and significantly facilitates healthcare decision-making. The first task, federated graph learning on Non-IID EHRs, aims to obtain complex disease graph representations with temporal dynamics from global imbalanced and locally insufficient Non-IID EHRs for downstream disease prediction tasks. We propose a personalized federated graph learning framework named PEARL, designed to avoid performance decreases in the global model on individual clients while enhancing the personalized capabilities of the learned global model. To further improve its effectiveness, we introduce a fine-tuning scheme to personalize the global model using local EHRs. Extensive experiments conducted on the real-world MIMIC-III dataset validate PEARL’s effectiveness, demonstrating significant improvement (10.25% on F1 scores) compared to baseline approaches. The second task, multi-view learning, offers a comprehensive exploration of both structured and unstructured EHRs. However, the intrinsic uncertainty among disease features presents a significant challenge for multi-view feature alignment. The sparsity of realworld EHRs further exacerbates this difficulty. To address these challenges, we introduce a novel fuzzy multi-view graph learning framework named FuzzyMVG, designed to mitigate the impacts of uncertainty in disease features derived from sparse EHRs. Extensive experiments on the real-world MIMIC-III dataset validate FuzzyMVG’s effectiveness. Results in the diagnosis prediction task show higher Precision (0.2991) that FuzzyMVG outperforms other state-of-the-art baselines. Finally, the third task addresses the challenges of limited computational resources, privacy leakage, and data silos in spatiotemporal Point-of-Interests (PoI) recommendation within distributed systems. We propose an efficient federated graph learning-based model to mine complex spatiotemporal features for generating recommendations. Experiments on the PoI recommendation task based on real-life check-in data validate the effectiveness of our proposed model. The results indicate that our recommendation model achieves competitive results (a higher accuracy, RMSE of 0.1096) with lower computational costs than the baselines.

Design, modelling, and optimization of inlet control valves for gas expanders for extended efficiency

  • Authors: Hossain, Md Shazzad
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: The Organic Rankine Cycle (ORC) is a promising process in energy systems due to its compact design and efficient operation with low-temperature heat sources. These attributes make it particularly suitable for small-scale power plants utilizing waste heat and renewable energy, contributing to the broader goal of sustainable energy solutions. A critical component in ORC systems is the gas expander, whose performance greatly influences the overall efficiency of the power plant. Among gas expanders, positive displacement machines, especially rotary type limaçon expanders, are highly advantageous for small-scale applications due to their ability to operate effectively at low speeds, low flow rates, and high-pressure ratios. These machines exhibit favourable pressure characteristics, offering significant efficiency and reliability advantages over conventional expander systems. Traditional gas expansion systems often rely on uncontrolled inlet ports or cam-operated valves that allow working fluid into the expander chamber but lack the ability to regulate the fluid flow effectively. This constraint leads to the wastage of high-quality working fluid and limited adaptability to varying loads. In contrast, a controlled inlet valve with an appropriate control scheme can address this issue by regulating fluid flow, thereby enhancing expander performance. This potential for enhancement, whilst it offers hope for the future of these energy conversion systems, underscores the significance of research projects focusing on designing, modelling, and controlling inlet valves to ensure desirable performance levels are achieved. This research initially proposes two direct-drive rotary valves actuated by a stepper motor and a rotary solenoid. Their performance is examined through mathematical models, and a comparative analysis of their impact on the limaçon gas expander’s performance is provided. The effects of temperature, friction, pressure, and leakage are also analyzed. The study finds that including the stepper motor valve can increase the isentropic and volumetric efficiencies of the expander considerably. However, this valve’s performance is sensitive to inlet pressure, which can degrade expander performance at higher pressures. Conversely, though less sensitive to pressure changes, the rotary solenoid valve yields improvements to these efficiencies to a lesser extent. A fast and accurate mathematical model is essential for optimizing and controlling complex systems such as a valved expander. Traditional analytical models, which are complex and time-consuming, are often computationally intensive and less suited for optimization and control applications. To address this limitation, an artificial neural network model is developed to predict the complex input-output relationships within the limaçon expander-stepper motor valve system. This model achieves high accuracy, with a normalized mean square error of 0.0014 and a coefficient of determination of 0.98, and is computationally efficient, outperforming simpler interpolation methods by 5.07%. The model’s efficiency and reduced computational loads make it suitable for optimizing and controlling the expander-valve system. The inlet valve must be optimized to achieve fast and accurate response characteristics. Thus, a push-pull solenoid valve is optimized using a Simultaneous perturbation stochastic approximation (SPSA) method. This optimization results in a 56-58% improvement in valve response speed. Analytically testing the optimized valve on a generic non-optimized limaçon expander shows that a faster valve alone can enhance the expander’s isentropic and volumetric efficiencies by 2.24% and 5.04% respectively. It is also observed that the optimized valve is robust and performs well even at lower pressures. However, future research will attempt to optimize a complex system that combines a valve and an expander. This thesis offers critical insights into the design and optimization of inlet valves for gas expanders, demonstrating their significant potential to enhance expander performance, particularly within ORC-based power generation. By addressing existing challenges and proposing innovative solutions, this work advances the prior understanding of limaçon expander technology and highlights the untapped potential of controlled inlet valves to improve performance. These findings lay the groundwork for future developments in small-scale renewable energy systems, significantly impacting the commercial viability of ORC-based power generation. However, further research is needed to translate these insights into practical, commercially viable applications.
  • Description: Doctor of Philosophy

Development and evaluation of a driving clinical decision pathway for non–driver trained occupational therapists to guide return-to-driving practices for adults following a change-in-health status

  • Authors: Scott, Hayley
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: Driving is a meaningful occupation that provides adults of all ages with the independence and freedom to access and interact with their communities to fulfil life roles (Fristedt et al., 2011). However, as many health events can interfere with the physical, cognitive and perceptual skills needed for safe-driving performance (Austroads, 2022), all occupational therapists have a duty of care to address driving as an occupation (Occupational Therapy Australia, 2022). In Australia, some occupational therapists complete postgraduate training to specialise in driving assessment and rehabilitation. They are known as occupational therapy driver assessors (OTDAs) whose practice is well supported by clinical guidelines and competency standards (Unsworth, 2007). However, non–driver trained occupational therapists (non-OTDAs) lack similar evidence-based guidelines and standards to guide their practice. Clinical decision pathways (CDPs) have been used in health care to link evidence to practice while also providing a standardised method to guide clinical decision-making for specific health conditions and have been found to improve the consistency of care and patient outcomes (Rotter et al., 2010; Schrijvers et al., 2012). The purpose of the research program that is the subject of this thesis was to develop and evaluate the effectiveness of a driving CDP used by non-OTDAs to guide return-to-driving practices for adults following a change-in-health status. A convergent mixed-methods design was used within a pragmatist research paradigm (Creswell & Plano Clark, 2018), which included four separate research studies. In the first study, a descriptive survey and file audits were used to understand the practices of non-OTDAs when addressing driving. This study found functional observations were commonly being used; however, driving was not being consistently addressed in practice due to gaps in self-reported knowledge, skills and the confidence of non-OTDAs. The study also found that non-OTDAs lacked the confidence to interpret assessment findings and the potential impact on driving. The second study investigated whether a performance-based assessment tool, the Multiple Errands Test-Home (Burns et al., 2019), could, alone or in combination with other standardised cognitive tests, predict driving outcomes that could help inform non-OTDA decision-making. While this assessment tool is not associated with driving outcomes, non-OTDAs still require further resources to inform their decision-making regarding driving. As a result, a comprehensive driving CDP was developed. In the third study, OTDA and non-OTDA focus groups reviewed and identified content changes to a newly developed driving CDP for use by non-OTDAs. A high level of consensus on inclusion of content and level of clinical usefulness was obtained to validate the driving CDP for clinical practice. Finally, a before–and-after design was used to evaluate practice changes following the implementation of the driving CDP in non-OTDA practice. This study demonstrated that there had been an increase in the number and type of assessments used and recommendations provided by non-OTDAs as well as an increase in self-reported knowledge, skills and confidence when addressing driving following the implementation of the driving CDP in practice. The research presented in this thesis has made a substantial contribution to non-OTDA practice in Australia. An evidence-based driving CDP has been developed, validated and evaluated to guide non-OTDA practice when addressing driving for adults following a change-in-health status. Finally, this research has described a process for developing CDPs which could be used by occupational therapists in other health related practice areas.
  • Description: Doctor of Philosophy

Discriminating malware families using partitional clustering

  • Authors: Mishra, Pooja
  • Date: 2024
  • Type: Text , Thesis , Masters
  • Full Text: false
  • Description: Malware, malicious software designed to compromise device security, is crafted by expert software engineers and distributed through a specialized black markets. Identifying malware families within daily feeds remains a significant challenge for internet security firms. Industry-standard Yara rules, based on regular expressions, are prone to failure due to malware evolution. This thesis presents an alternative approach leveraging malware clustering. By clustering malware samples based on dynamic analysis features, Yara scans can efficiently pinpoint known families, but unrecognized samples signify potential new variants, earmarked for further scrutiny by analysis teams. This process diminishes the necessity for individual sample scans, thereby streamlining operations and lightening the analysis team’s workload. This research evaluates the partitional clustering algorithm for improved handling of sparse malware features, setting it against the following traditional algorithms K-Means, Agglomerative Clustering, DBSCAN, and Spectral Kmeans Clustering. Each algorithm is evaluated, with a focus on their efficacy clustering performance: KMeans optimizes for homogeneous variance across n groups; Agglomerative Clustering scales for large datasets via connectivity matrices; DBSCAN discriminates clusters based on density metrics; and Spectral K-means Clustering employs affinity matrix-based low-dimensional embedding prior to clustering. The contribution of this thesis include a comprehensive performance comparison of the partitional clustering algorithm against Hierarchical, Densitybased, Spectral K-means, and K-Means algorithms; enhancement of the partitional clustering algorithm for sparse data; an in-depth evaluation of features extracted from Application Programming Interface call parameters and Domain Name System queries executed by malware; and the development of countermeasures against malware’s anti-analysis tactics. The research utilizes a real-world malware dataset sourced from abuse.ch 1 [1]. Empirical results demonstrate the superior performance of the partitional clustering algorithm over traditional clustering techniques in the majority of tests conducted
  • Description: Masters of Research

Enablers and potential barriers of female participation in tertiary education in Afghanistan : an analysis of contemporary issues

  • Authors: Najibi, Parwaiz
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: Afghanistan’s education system has been devasted by more than four decades of sustained conflict (United Nations International Children’s Emergency Fund [UNICEF], 2019). Even completing primary school remains a distant dream for many children, especially those in rural areas, particularly females. According to UNICEF (2019), a key challenge for Afghanistan is that an estimated two-thirds of the female population do not attend school. Furthermore, as personal security in the country deteriorates, female enrolment in tertiary education is also declining (Pherali & Sahar, 2018). In addition, many families have fled their villages because of the ongoing unrest and are concentrated in cities where they live in poverty and have little access to educational services (Baiza, Nevertheless, according to the United Nations Educational, Scientific and Cultural Organization (2009), education is the most potent weapon for positive change in the world. Moreover, the increased participation of women in tertiary education improves economic growth and stability (McLean, 2020; UNICEF, 2011). This study examines problems concerning women’s access to tertiary education in Afghanistan and potential solutions to these problems. A mixed methods experimental sequential research design was used for this study (Creswell & Plano Clark, 2011). This study used an online survey of 120 undergraduate students and explored the lived experiences of 10 female undergraduates and 10 graduates through an online individual interview. Further, 10 lecturers were recruited from five disciplines and three universities through individual online interviews. Ethics approval was obtained to engage with only undergraduates/graduates currently living in Australia, because of safety considerations. As per Leighton et al. (2021), snowball sampling was also utilised for surveys and semi-structured interviews. This study aimed to understand the participants’ perspectives concerning the factors influencing their decision to pursue tertiary education and the obstacles preventing Afghan women from participating in tertiary education in Afghanistan. The findings from this study revealed that while women's education is universally acknowledged as essential for economic, cultural, social, and political development, its realisation in Afghanistan is impeded by entrenched cultural, societal, religious, and political factors. The study highlighted the benefits of tertiary education for Afghan women, including increased employment opportunities, higher income potential, and contributions to national development and community advancement. However, these are overshadowed by substantial challenges such as government policy and practices, university system and infrastructure and family and culture restrictions. As measures to increase female enrollment in tertiary education in Afghanistan, this study recommends addressing cultural and social norms through awareness campaigns challenging traditional gender roles. It also recommends engaging influential figures such as fathers, religious leaders, and community elders in advocating for women's education. It is imperative to provide women-friendly campuses, implement security measures, and provide gender-segregated classrooms (when culturally necessary) to ensure a safe and inclusive learning environment. Financial support can alleviate economic burdens, including scholarships and partnerships with international organisations.
  • Description: Doctor of Philosophy, Partial

Ensemble Approaches for Robust Reconstruction of Gene Regulatory Networks

  • Authors: Gamage, Hasini
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text:
  • Description: Gene regulatory networks (GRNs) are intricate control systems governing gene expression dynamics, playing a pivotal role in biological processes. The ongoing development of high-throughput microarray and sequencing technologies has greatly facilitated the acquisition of gene expression data, promoting an extensive body of research focused on unravelling the intricacies of GRNs. This endeavour involves deciphering how genes regulate each other, vital for understanding the molecular functions of cells and diseases, and for designing targeted therapies. However, GRN reconstruction is a formidable task due to the high dimensionality, limited sample size, and the presence of noise in gene expression data. Various reverse engineering approaches have been developed to grapple with these challenges. Each method exhibits certain method-specific issues. A recent trend in the field is the emergence of ensemble methods designed to yield robust reconstructions. This thesis presents a comprehensive exploration of GRN inference through the development and application of ensemble methods, offering both theoretical insights and practical tools. This research commenced with a thorough examination of the challenges involved in the reconstruction of GRNs through the utilization of existing individual inference methods, along with an assessment of their inherent limitations. While there is a plethora of GRN modelling approaches available, we focussed on three distinct modelling approaches: Boolean, regression, and information theory-based fuzzy methods. The selection of these methods was underpinned by established categorizations found in the existing literature and substantial empirical evidence of their remarkable performance in GRN inference. Each of these methods is conceptually different, offering diverse vantage points on the inference challenge. One of the primary objectives of this study was to apply ensemble techniques to each selected individual modelling approach, thereby enhancing the inner method diversity to further enhance the individual method performance. The first modelling approach involves a Boolean network. The initial version employed a simple Boolean network and yielded near-optimal results, while the second version enhanced accuracy by incorporating feature selection methods. The second modelling approach treated GRN inference as a regression problem and involved two variations. The first variation combines cross-validated Lasso (LassoCV) and cross-validated Ridge (RidgeCV), while the second variation addressed noisy gene expression data by combining quantile regression (QR) and RidgeCV for gene selection. The third modelling approach was a novel hybrid fuzzy method, a combination of information theory-based pre-processing stage and a subsequent fuzzy method, known as MICFuzzy, which brought substantial performance improvements. Our experiments showed that each of these individual methods produced performance improvements over state-of-the-art methods with regard to both accuracy and efficiency, by increasing the robustness of GRN reconstruction. In conjunction with the enhanced performance of each of these modelling approaches, our ultimate objective was the development of a novel, unique ensemble framework for robust GRN inference, obtained by combining the outcomes of the different modelling techniques. Thus, we developed a novel ensemble framework called GRAMP: A Gene Ranking And Model Prioritisation framework, to aggregate the inferred networks produced by the aforementioned modelling approaches. This framework addresses the need for a reliable approach for ensemble modelling in GRN inference, aggregating the outcome of diverse modelling approaches. GRAMP includes a novel network aggregation method based on gene scores. Gene scores are evaluated based on the performance of each inference method in a specific problem context and both local and global gene ranking. Experimental results using both simulated and real-world gene expression datasets confirmed the superior performance of this ensemble framework in inferring gene regulatory networks. To further enhance the practical application of these ensemble approaches we introduced a user-friendly desktop application that implements the GRAMP framework, allowing researchers to integrate multiple inference methods and datasets from various problem contexts. This tool fills a critical gap in the availability of an interactive software tool for ensemble model building. This application is freely accessible to the research community. This thesis demonstrates how the research offers a systematic exploration of diverse modelling techniques and the success of our ensemble approaches for GRN inference in the form of our published work. These contributions collectively pave the way for producing robust GRN inference in systems biology, with broad-reaching implications for biology and medicine.
  • Description: Doctor of Philosophy

Exercise response and exhaled volatile organic compounds

  • Authors: Bell, Leo
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text: false
  • Description: A growing body of research features the variability of responses to exercise training, frequently showing approximately 30-40% of participants do not experience meaningful improvements in fitness (also known as ‘nonresponse’). Hence, researchers have focused on investigating determinants of responsiveness to optimise exercise interventions at a personalised level and mitigate the prevalence of nonresponses. This PhD thesis aims to advance knowledge of these factors and provide future directions for research into potential biomarkers for monitoring exercise response. Interindividual variability in response to exercise is determined by a combination of biological (e.g. age, sex, genetics), environmental (e.g. lifestyle, diet, physical activity), and methodological factors (e.g. exercise dose, study design, statistical approach). While genetics are a large component of interindividual variability, studies comparing exercise response in twins indicate genetics may not play as substantive role as originally thought. Additionally, mRNA expression is tissue-specific, expensive, and invasive. Therefore, to improve the incidence of exercise fitness response it is imperative cost-effective, practical, and less invasive molecular predictors of exercise response are identified. This work highlights the potential of exerkines, circulating cell-free DNA, and metabolomic profile changes in blood, saliva, and breath as candidate biomarkers. Exercise studies mainly use reporting measures at the group level (e.g. mean, standard deviation). However, central tendency measures fail to highlight the interindividual variability and capture the proportion of nonresponses. This thesis demonstrates the merit of applying statistical frameworks for assessing interindividual variability and classifying individual responses to exercise training. A study examined whether progressively increasing treadmill run intensity would improve the proportion of positive cardiorespiratory fitness responders compared to constant intensity training. The findings show a significantly higher proportion of positive responders in the progressive overload group, however, a notable percentage of participants (~33%) did not exceed the technical error and smallest worthwhile change threshold (128.2mL). The following study examined the impact of different supervision types on individual responses to exercise training in tertiary education employees. Personal supervision led to the most significant improvements in cardiorespiratory fitness, muscular strength, and body composition compared to non-personal supervision and unsupervised training. Analysis of individual responses indicated a reduced incidence of non-responses for muscular strength and total fat loss, but not V̇O2peak, for personal supervision compared to other supervision types, suggesting that personal supervision can improve individual responses to exercise training. Additionally, the analysis supports evidence that outcome responses do not aggregate consistently within participants. These findings, in combination with the first study, highlight the need to develop measures capable of screening and predicting the likelihood of response to exercise. Breath analysis is a promising non-invasive method for monitoring physiological and metabolic adaptations. Consequently, a breath testing method was piloted to identify potential exhaled volatile organic compounds for monitoring exercise response. Using glass tubes and solid-phase microextraction fibres paired with gas chromatography-mass spectrometry, a non-targeted analytical approach was employed to speculate for volatile organic compounds of interest. The descriptive analysis highlights compounds from endogenous, exogenous, and mixed origins before, 10 minutes and 24 hours after a standardised 20-minute treadmill run. Metabolites of interest included monoterpenes, alkenes and dienes, ketones, aldehydes, alcohols, organosulfur, alkanes, amines, and amides. The presence of certain metabolites was transient and highlighted the confounding influence of the environment on breath sampling. In a follow-up study, the reliability of two commonly abundant compounds, acetone and isoprene, and their dynamic responses across time points are reported. The reliability analyses showed that acetone has poor reliability and consistency within individuals, while isoprene shows moderately good reliability and consistency. Isoprene levels significantly decreased 10 minutes after exercise and returned to baseline after 24 hours, aligning with existing theories on their dynamics. In summary, this thesis aims to highlight the prevalence of non-response to generic training regimens by reporting the relative importance of progressive increases in exercise intensity and the impact of supervision type on individual responses to exercise training. Additionally, this thesis reports on the potential of breath analysis to be a viable biomarker of trainability and suggests future directions for research in this area.
  • Description: Doctor of Philosophy

High speed motion planning with practical constraints for robot manipulators

  • Authors: Le, Minh
  • Date: 2024
  • Type: Text , Thesis , Masters
  • Full Text:
  • Description: Robot motion planning is crucial for enabling robot manipulators to perform tasks efficiently and safely within complex environments. The incorporation of task-specific constraints further augments the capabilities of these robots, allowing them to navigate intricate spaces while fulfilling specialized objectives. This thesis pursues three primary research objectives. The first two objectives delve into robot motion planning with an emphasis on task-specific constraints. The first objective aims to achieve high-speed and safe transportation of a grasped object by constraining the end-effector’s spatial acceleration during the motion and integrating this constraint into a motion planning scheme for evaluation. The second objective focuses on preventing spillage when transporting a liquid-filled, open container by constraining the gripper’s pose during movement while maintaining a fast completion time. The third objective proposes a novel approach to collision avoidance in motion planning, utilizing a robust and intuitive mathematical framework. The motion planning framework employed in this thesis leverages the feasibility horizon of a given motion optimization problem to devise a high-speed resulting trajectory. For the first research objective, the recursive Newton-Euler algorithm is applied to formulate the constraints on end-effector spatial acceleration. The Jacobian matrix is used to formulate the task-specific constraint for the second objective concerning the gripper’s pose. For the third objective, the thesis explores the potential of geometric algebra as an alternative framework for robotics, developing a collision avoidance scheme that benefits from the geometrical robustness of geometric algebra. A comprehensive comparison with traditional matrix algebra approaches is also conducted to demonstrate the advantages of geometric algebra. Applying the proposed methods to robot motion planning scenarios demonstrated several improvements. The research produces motion paths exhibiting beneficial characteristics tailored to their respective tasks, such as safely constrained spatial acceleration of the end-effector, stability of the end-effector’s pose, and collision-free trajectories. Additionally, the use of the time-efficient motion planning framework contributes to the overall time efficiency of the resulting motions. This thesis makes significant contributions to the field of robot motion planning. Firstly, it highlights the critical role of task-specific constraints in shaping the trajectory planning process. Secondly, by proposing the adoption of geometric algebra, the research offers a novel approach to addressing the collision avoidance challenge in motion planning. The findings have important implications for the design and implementation of robotic systems, paving the way for more efficient and robust motion planning algorithms. Through this research, I aim to inspire further investigation into the integration of task-specific constraints and the adoption of geometric algebra in robot motion planning, informing future research directions and leading to advancements in robot manipulation and navigation capabilities.
  • Description: Master of Engineering Science

Image data understanding and preparation

  • Authors: Kaur, Roopdeep
  • Date: 2024
  • Type: Text , Thesis , PhD
  • Full Text: false
  • Description: Data understanding and preparation involves a process of analyzing, cleaning, transforming, and organizing the data in preparation for data mining. To improve the performance of applications that use image processing, verifying the quality of the images and image cleaning are crucial steps. However, diverse and complex environments have a great effect on images, directly affecting the decisions derived from image analysis and thereby limiting the acceleration of industrial automation. Environmental and camera impacts play a vital role in the quality of the photos captured in outdoor environments. Because of these impacts, the use of images captured in outdoor environments limits the effectiveness of an application in which image processing is involved. There are many techniques available in the current literature for analyzing the impact of the environment on Internet of Things (IoT) images. However, objectively assessing the effect of dynamic and complex environments on IoT images is challenging. To advance this research area, we present an innovative technique for evaluating the impact of environmental parameters on image quality compared with the quality affected by the Joint Photographic Experts Group (JPEG) image compression technique and the different levels of Gaussian noise. The quality values produced by the structural similarity index measure (SSIM) are consistent with the different levels of environmental impacts, JPEG image quality, and Gaussian noise, and can be used for image understanding and preparation. For camera impacts, there exist many approaches that assess the influence on the quality of the images. Analysis shows that none of the existing metrics produces quality values consistent with intuitively defined impact levels for lens blur, lens dirtiness, or barrel distortion. To address the loopholes in the existing metrics and to ensure that the quality assessment metrics are more reliable, we introduce a new image quality assessment metric that uses the Dempster–Shafer theory to fuse quality values from different metrics. Our proposed metric produces quality values that are more consistent and better aligned with perceptually defined camera parameter impact levels. Various noise reduction techniques are proposed in the literature for image data preparation, including the use of median, Gaussian, and bilateral filters. Convolutional neural networks (CNNs) have gained popularity in image denoising owing to their ability to extract complex patterns and features from data. CNNs are highly adaptable, making them effective tools for various image-denoising tasks. The drawback of CNN-based techniques is that they require an appropriate training dataset and all images to be resized. Another notable disadvantage of these filtering techniques is that they work for certain types of environmental and camera impacts. To bridge this research gap, we analyze the impact of denoising on CNN performance. First, we filter noise from images using traditional denoising methods before using them in the CNN model. Second, we embed a denoising layer within the CNN. We conduct extensive experiments on traffic sign and object recognition datasets to validate the performance of image denoising. We also present an approach using peak signal-to-noise ratio (PSNR) distribution to determine whether denoising should be adopted and which filter to use. Both CNN accuracy and PSNR distribution are used to determine the type of filter that needs to be used. The results vary by filter type, impact, and dataset, with traditional denoising showing better accuracy and embedded denoising offering shorter computational time in most cases. This comparative study provides insights into adopting denoising in various CNN-based image analyses. From this analysis, it is shown that traditional and embedding denoising techniques are efficient in reducing many impacts; however, these are not competent enough to reduce impact types such as salt and pepper, lens blur, and shadow. To address this problem, finally, for the first time, we introduce an approach to directly filter out poor-quality images for different environmental and camera impacts. Our approach assesses quality using an image quality metric and employs an optimal threshold to remove low-quality images while ensuring that an adequate number of images remain for deep learning model development. Results from real and simulated traffic and object recognition data showcase the superior performance of our approach compared with state-of-the-art approaches. The merit of our technique is that it works well for all environmental and camera impacts with comparable computational time.
  • Description: Doctor of Philosophy