7103 research outputs found
Sort by
Modulated electron beam produced by a thermionic cathode electron gun for particle accelerator applications
Get PDFElectron injectors used in radiation sources such as Free-electron lasers (FELs) and medical linear accelerators (LINACs) can generate high peak current and low emittance electron beams. There are different types of electron injectors depending on their cathode. These are thermionic cathodes, photocathodes, and field emission cathodes. Each one of them has its own advantages and disadvantages. In this thesis, considering the advantages of a long lifetime, large current density and being cost effective, a thermionic cathode gridded electron gun for a particle accelerator was designed and modelled. Both theoretical work and numerical simulations were carried out to explore the relationship between important parameters like the bunch charge and the bunch length of the modulated beam. Two 2D simulation packages, the DC electron trajectory solver TRAK and Particle-In-Cell code MAGIC were utilised initially to optimise the Pierce-type electron gun and to simulate the RF field applied to the grid. Similar to existing guns the electron energy, pulse duration and charge of the electron beam bunch were predicted but the significant deviation from existing guns was the pulse length as the function of the bunch charge. The beam dynamics simulations showed that a minimum pulse length of 106 ps could be achieved with a bunch charge of 33 pC when the driving RF frequency was 1.5 GHz. Simulations at ahigher RF frequency did not significantly reduce the micro-pulse length and the normalised emittance was measured to be 5.6 mmmrad obtained from the particle-in-cell simulations. The results obtained with 2D simulation packages were compared with 3D simulations using CST Particle Studio with similar values for the pulse length at a value of 102 ps for 35 pC bunch charge observed. Different designs of grids, the spider web grid and the pepper pot grid were simulated and their performance was examined and compared. The comparison resulted in similar values for the peak current and the bunch charge. Furthermore, other techniques for reduction of the bunch length like higher harmonics of the fundamental frequency, specifically the idealised case of a square wave, using CST Particle Studio are presented. Overall, the notable advance in science lies in obtaining the bunch length as the function of the bunch charge which enabled the calculation of the emittance as a function of time by postprocessing the output of the numerical simulations. This is vital information to be passed to the designers of the S-band (3GHz) LINAC as it enables the performance of the LINAC in terms of capture of electrons to be predicted.Electron injectors used in radiation sources such as Free-electron lasers (FELs) and medical linear accelerators (LINACs) can generate high peak current and low emittance electron beams. There are different types of electron injectors depending on their cathode. These are thermionic cathodes, photocathodes, and field emission cathodes. Each one of them has its own advantages and disadvantages. In this thesis, considering the advantages of a long lifetime, large current density and being cost effective, a thermionic cathode gridded electron gun for a particle accelerator was designed and modelled. Both theoretical work and numerical simulations were carried out to explore the relationship between important parameters like the bunch charge and the bunch length of the modulated beam. Two 2D simulation packages, the DC electron trajectory solver TRAK and Particle-In-Cell code MAGIC were utilised initially to optimise the Pierce-type electron gun and to simulate the RF field applied to the grid. Similar to existing guns the electron energy, pulse duration and charge of the electron beam bunch were predicted but the significant deviation from existing guns was the pulse length as the function of the bunch charge. The beam dynamics simulations showed that a minimum pulse length of 106 ps could be achieved with a bunch charge of 33 pC when the driving RF frequency was 1.5 GHz. Simulations at ahigher RF frequency did not significantly reduce the micro-pulse length and the normalised emittance was measured to be 5.6 mmmrad obtained from the particle-in-cell simulations. The results obtained with 2D simulation packages were compared with 3D simulations using CST Particle Studio with similar values for the pulse length at a value of 102 ps for 35 pC bunch charge observed. Different designs of grids, the spider web grid and the pepper pot grid were simulated and their performance was examined and compared. The comparison resulted in similar values for the peak current and the bunch charge. Furthermore, other techniques for reduction of the bunch length like higher harmonics of the fundamental frequency, specifically the idealised case of a square wave, using CST Particle Studio are presented. Overall, the notable advance in science lies in obtaining the bunch length as the function of the bunch charge which enabled the calculation of the emittance as a function of time by postprocessing the output of the numerical simulations. This is vital information to be passed to the designers of the S-band (3GHz) LINAC as it enables the performance of the LINAC in terms of capture of electrons to be predicted
Acquirers' gains, implications of information asymmetry, value of cash and capital controls
Get PDFMerger and acquisition (M&A) is one of such crucial investment decisions and draws the attention not only of shareholders but also of other stakeholders. Although there are voluminous studies on the wealth creations through corporate acquisitions, the results regarding whether engaging in such activities can enhance the value of the acquirer's shareholders remain inconclusive as the outcome depends on several factors with multi-directional influences. This thesis focuses on three issues that have received relatively little or no attention, namely (i) information asymmetry, (ii) marginal value of cash balance, and (iii) the capital control policy of the acquirers' domicile. The broad research objective of this thesis is to examine whether these three issues can affect the value of acquirers and if so how value can be created.;In the first empirical chapter (Chapter 2), we examine how information asymmetry between acquiring firms' corporate insiders and the market affects the acquirers' gains and takeover premium in the US market. Controlling for the information asymmetry of target firms and other determinants, we find a negative correlation between acquirer's information asymmetry and takeover premium in both stock and cash deals. There is no evidence to support that acquirers can benefit from exchanging their overvalued stocks for target firms' assets in the short- and long term. Instead, overvalued acquirers suffer greater loss, particularly when they have high levels of information asymmetry. However, our findings suggest that acquirers with high information asymmetry and knowledge of high takeover synergies can enhance their shareholders' value if they engage only in stock-financed acquisitions. Lastly, when the effect of self-revaluation and equity financing are alleviated, we can observe the improvement in the true gain of acquisitions.;The second empirical chapter (Chapter 3) posits that the marginal value of cash to the acquirer should be a better measure, relative to the nominal value of the cash balance, in explaining the acquirer's choice of method of payment and the value implications of corporate cash holdings. The results of US M&A confirm that the payment method choices of bidder managers are related to the marginal value of cash held by acquirers one year prior to the bid announcement. Acquirers engaging in stock bids have a relatively high marginal value of cash than those who choose to pay with cash. Further, we document that value of cash held by acquirers is not static across time and managers are good at timing the market of value cash. Lastly, we report that with the right payment decision corresponding to the value of cash, bidder managers can create value for their shareholders.;The third empirical chapter (Chapter 4) explores the effect of capital control of acquirers' domicile on their gains from cross-border deals. By following the springboard strategy, we predict that managers can enhance their shareholders' wealth and receive higher announcement period returns. We find evidence to support our prediction that cross-border flows and gains are driven by purchasing target firms resided in relatively low capital control countries. These gains can be further explained with the accessibility of cheaper capital and better risk diversification.Merger and acquisition (M&A) is one of such crucial investment decisions and draws the attention not only of shareholders but also of other stakeholders. Although there are voluminous studies on the wealth creations through corporate acquisitions, the results regarding whether engaging in such activities can enhance the value of the acquirer's shareholders remain inconclusive as the outcome depends on several factors with multi-directional influences. This thesis focuses on three issues that have received relatively little or no attention, namely (i) information asymmetry, (ii) marginal value of cash balance, and (iii) the capital control policy of the acquirers' domicile. The broad research objective of this thesis is to examine whether these three issues can affect the value of acquirers and if so how value can be created.;In the first empirical chapter (Chapter 2), we examine how information asymmetry between acquiring firms' corporate insiders and the market affects the acquirers' gains and takeover premium in the US market. Controlling for the information asymmetry of target firms and other determinants, we find a negative correlation between acquirer's information asymmetry and takeover premium in both stock and cash deals. There is no evidence to support that acquirers can benefit from exchanging their overvalued stocks for target firms' assets in the short- and long term. Instead, overvalued acquirers suffer greater loss, particularly when they have high levels of information asymmetry. However, our findings suggest that acquirers with high information asymmetry and knowledge of high takeover synergies can enhance their shareholders' value if they engage only in stock-financed acquisitions. Lastly, when the effect of self-revaluation and equity financing are alleviated, we can observe the improvement in the true gain of acquisitions.;The second empirical chapter (Chapter 3) posits that the marginal value of cash to the acquirer should be a better measure, relative to the nominal value of the cash balance, in explaining the acquirer's choice of method of payment and the value implications of corporate cash holdings. The results of US M&A confirm that the payment method choices of bidder managers are related to the marginal value of cash held by acquirers one year prior to the bid announcement. Acquirers engaging in stock bids have a relatively high marginal value of cash than those who choose to pay with cash. Further, we document that value of cash held by acquirers is not static across time and managers are good at timing the market of value cash. Lastly, we report that with the right payment decision corresponding to the value of cash, bidder managers can create value for their shareholders.;The third empirical chapter (Chapter 4) explores the effect of capital control of acquirers' domicile on their gains from cross-border deals. By following the springboard strategy, we predict that managers can enhance their shareholders' wealth and receive higher announcement period returns. We find evidence to support our prediction that cross-border flows and gains are driven by purchasing target firms resided in relatively low capital control countries. These gains can be further explained with the accessibility of cheaper capital and better risk diversification
The influence of station keeping systems on tidal turbine structural performance when operating in combined-wave current sea states
Get PDFThis thesis reports on the performance and interactions of a tidal turbine and station keeping systems based on the adoption of a tension mooring system in different sea states. The capabilities of introducing damping are being investigated to reduce the peak loads that tidal turbines experience during operational life in high energy wave-current environments and extreme sea states. A neutrally buoyant turbine is supported from a tension cable based mooring system, where tension is introduced by a buoy fully submersed in water. The loading on the turbine rotor blades and buoy are calculated using a wave and current coupled BEMT. The modeling algorithm developed is based on an inverted triple pendulum, responding to different sea state conditions to understand the system response behavior and the blade load in different sea states, including extreme conditions. The results show the tension mooring system reduce speak thrust loading on the turbine, but it was found that there are certain limitation when using this design in extreme waves conditions.This thesis reports on the performance and interactions of a tidal turbine and station keeping systems based on the adoption of a tension mooring system in different sea states. The capabilities of introducing damping are being investigated to reduce the peak loads that tidal turbines experience during operational life in high energy wave-current environments and extreme sea states. A neutrally buoyant turbine is supported from a tension cable based mooring system, where tension is introduced by a buoy fully submersed in water. The loading on the turbine rotor blades and buoy are calculated using a wave and current coupled BEMT. The modeling algorithm developed is based on an inverted triple pendulum, responding to different sea state conditions to understand the system response behavior and the blade load in different sea states, including extreme conditions. The results show the tension mooring system reduce speak thrust loading on the turbine, but it was found that there are certain limitation when using this design in extreme waves conditions
Happiness sells : the impact of emotional pareidolic face configurations on cognition, product perception and consumer attitudes
No full textSeeing facial configurations in non-face objects – i.e. face pareidolia – is a ubiquitous psychological experience that commonly manifests itself within product design. Yet, there has been little investigation into the detection of emotional pareidolic face configurations in everyday products, as well as consumer attitudes and individual differences regarding these. To explore this, a series of image rating tasks and a visual cueing paradigm were formulated to quantitatively assess the effect of pareidolia on cognition, product perception and consumer attitudes. Like real faces, consumers (N = 37) could detect and label core emotions within pareidolic products, with happy faces being most accurately detected. Moreover, products with happy, surprised, and angry pareidolic content were most likely to capture the interest of and intrigue consumers (N = 102), though only happy products retained this advantage for intention to purchase. A general aversion was also shown towards products in which disgust was perceived, for all attitudinal measures. These trends were consistent across consumer types, with limited evidence for the role of psychological variables in consumers’response to pareidolia; only low mood (negatively) and social isolation (positively) predicted consumers’ interest in pareidolic products. However, no differences were found in response times when comparing consumers’ (N = 22) implicit attentional capture of pareidolic versus non-pareidolic product images. Integrated findings provide evidence that emotion plays a salient yet complex role in the interpretation of pareidolic products – consumers’ appraisals were impacted by emotion perception, as opposed to a general preference for products with faces of any kind. Implications for product designers are explored; there is potential for pareidolia to be deployed as a design tool to aid social isolation, and it is recommended that designers include pareidolic configurations with a positive valence (i.e. happy) and avoid those with a negative valence (i.e. disgust) to maximise product engagement and purchase intention.Seeing facial configurations in non-face objects – i.e. face pareidolia – is a ubiquitous psychological experience that commonly manifests itself within product design. Yet, there has been little investigation into the detection of emotional pareidolic face configurations in everyday products, as well as consumer attitudes and individual differences regarding these. To explore this, a series of image rating tasks and a visual cueing paradigm were formulated to quantitatively assess the effect of pareidolia on cognition, product perception and consumer attitudes. Like real faces, consumers (N = 37) could detect and label core emotions within pareidolic products, with happy faces being most accurately detected. Moreover, products with happy, surprised, and angry pareidolic content were most likely to capture the interest of and intrigue consumers (N = 102), though only happy products retained this advantage for intention to purchase. A general aversion was also shown towards products in which disgust was perceived, for all attitudinal measures. These trends were consistent across consumer types, with limited evidence for the role of psychological variables in consumers’response to pareidolia; only low mood (negatively) and social isolation (positively) predicted consumers’ interest in pareidolic products. However, no differences were found in response times when comparing consumers’ (N = 22) implicit attentional capture of pareidolic versus non-pareidolic product images. Integrated findings provide evidence that emotion plays a salient yet complex role in the interpretation of pareidolic products – consumers’ appraisals were impacted by emotion perception, as opposed to a general preference for products with faces of any kind. Implications for product designers are explored; there is potential for pareidolia to be deployed as a design tool to aid social isolation, and it is recommended that designers include pareidolic configurations with a positive valence (i.e. happy) and avoid those with a negative valence (i.e. disgust) to maximise product engagement and purchase intention
Entrepreneurial ecosystem : Islamic leadership model - an exploration of the entrepreneurial ecosystem in Sultanate of Oman
Get PDFOman’s economic growth and development have recently shifted from the declining oil-dependent economy to other diversified sources of income generation, such as enhancing the performance of entrepreneurship as an economic growth driver. One of the ways through which such economic goals can be achieved is through the efficient management of entrepreneurial ecosystems. However, the review of existing literature features limited details in terms of how entrepreneurial ecosystems work in reality, as well as in terms of policy-related challenges in the management of entrepreneurial ecosystems. By conducting interviews on 36 participants consisting of 18 policy makers and 18 entrepreneurs, the present research established that even though there are some positive factors that support the growth of the entrepreneurial ecosystem (such as the existence of the support systems, networks, and connectivity), there are a lot of challenges that are hindering efficiency with regards to the entrepreneurial ecosystem in Oman. The most noted challenges to the effective management of entrepreneurial ecosystems are policy vulnerability and the government’s use of a top-to-bottom approach in policy and decision-making that do not involve or engage entrepreneurs, creating, in return, a disconnect and lack of balance in the ecosystem. Low stakeholder engagement and poor implementation of set policies also hinder entrepreneurship in the country. Thus, this study recommends an Islamic leadership management (ILM) approach to the management of entrepreneurial ecosystems. In this regard, effective management can be attained when skills are developed through inclusion and information sharing, incorporating religious principles such as fairness in policy making towards building institutional links, respecting cultural aspects such as diversity for an adaptive ecosystem, prioritizing strong relationships leading to effective networks, and viewing humans as custodians of earthly resources leading to equitable allocation of resources in the ecosystem. Additionally, an adaptive co-management approach can help reinforce the Islamic model as it emphasizes on learning-by-doing, relationships and the capacity of the communities and resource users.Oman’s economic growth and development have recently shifted from the declining oil-dependent economy to other diversified sources of income generation, such as enhancing the performance of entrepreneurship as an economic growth driver. One of the ways through which such economic goals can be achieved is through the efficient management of entrepreneurial ecosystems. However, the review of existing literature features limited details in terms of how entrepreneurial ecosystems work in reality, as well as in terms of policy-related challenges in the management of entrepreneurial ecosystems. By conducting interviews on 36 participants consisting of 18 policy makers and 18 entrepreneurs, the present research established that even though there are some positive factors that support the growth of the entrepreneurial ecosystem (such as the existence of the support systems, networks, and connectivity), there are a lot of challenges that are hindering efficiency with regards to the entrepreneurial ecosystem in Oman. The most noted challenges to the effective management of entrepreneurial ecosystems are policy vulnerability and the government’s use of a top-to-bottom approach in policy and decision-making that do not involve or engage entrepreneurs, creating, in return, a disconnect and lack of balance in the ecosystem. Low stakeholder engagement and poor implementation of set policies also hinder entrepreneurship in the country. Thus, this study recommends an Islamic leadership management (ILM) approach to the management of entrepreneurial ecosystems. In this regard, effective management can be attained when skills are developed through inclusion and information sharing, incorporating religious principles such as fairness in policy making towards building institutional links, respecting cultural aspects such as diversity for an adaptive ecosystem, prioritizing strong relationships leading to effective networks, and viewing humans as custodians of earthly resources leading to equitable allocation of resources in the ecosystem. Additionally, an adaptive co-management approach can help reinforce the Islamic model as it emphasizes on learning-by-doing, relationships and the capacity of the communities and resource users
Understanding RF gel formation : modelling fundamental nanoscale processes for enhanced material development
Get PDFThe wide-ranging application potential for porous materials has been of significant interest over the years, with a particular focus on those which possess attractive properties, such as low densities and high surface areas. Materials such as these have proven to be effective in a wide range of applications, many of which are imperative in reducing or eradicating detrimental environmental impacts of industry, heightening their pertinence to recent research. This work focuses on one such class of organic porous materials—resorcinol–formaldehyde (RF) gels—which are formed via a sol–gel process and subsequently dried, producing the lightweight, nanoporous structure of the final gel. Despite extensive research into these materials in recent years, a number of questions still remain around their formation mechanism and the impact of various parameters associated with their synthesis. As a resultof this, their application potential is yet to be fully realised, especially given the wide range of properties that can be achieved through fine tuning and optimisation of the RF gel synthesis process. In this work, the formation mechanism of RF gels is explored through both experimental and computational means. Through experimental analysis of the final textural properties of synthesised gels, the impact of variations in catalyst concentration and catalyst species are investigated, aiming to elucidate the specific role the catalyst compound plays within the RF reaction – something that, to date, has been widely debated. The importance of the metal cation within the catalyst is highlighted through the results presented here, its concentration decoupled with the initial solution pH, and the significance of both discussed in detail. The comparative efficacy of different solvents used within the solvent exchange step of gel synthesis is also investigated, measured in their ability to preserve the structure during drying, minimising the pore shrinkage and collapse that takes place. The implications of the results obtained are discussed in relation to process optimisation to achieve desirable properties applicable to specific uses.The synthesis and analysis of RF gels is time consuming, therefore, simulating these processes computationally in an efficient manner could be pivotal to facilitating their widespread use. In this work, a three-dimensional model is developed which captures the formation and growth of RF gels using lattice-based kinetic Monte Carlo theory, accounting for varying catalyst concentration and solids content – two parameters proven to control gel properties inexperimental work. The textural properties of the resulting simulated materials are analysed, including the accessible surface area and accessible porosity, the values of which reflect the increased structural density and inter-connected complexity associated with increasing solids content and catalyst concentration. Furthermore, the fractal properties of these materials are analysed through correlation dimension and Hurst exponent calculations, the results demonstrating that while fractal properties are not typically observed in scattering experiments for RF gels, they are possible to achieve with sufficiently low solids content and catalyst concentration. As the most commonly employed method of assessing properties of porous materialsexperimentally, adsorption analysis was carried out computationally for the simulated RF gel structures. The results indicated that both low catalyst concentrations and low solids contents resulted in structures with open transport pores that were larger in width, while high catalyst concentrations and solids contents resulted in structures with bottleneck pores that were narrower. Importantly, the computational isotherm data and pore size distributions were also compared to those obtained experimentally, showing a promising agreement in trends between the two for varying catalyst concentrations, providing validation for the kinetic Monte Carlo model developed. Finally, the performance of RF gels in a specific application is tested, assessed in their ability to remove an endocrine disrupting pollutant from water through UV-Vis concentration measurements. Using the results obtained from both the experimental and computational analysis of the materials, the comparative efficacy of two RF gels synthesised under different catalyst concentrations is predicted and subsequently explored, and the properties required for optimal performance determined. This work not only highlights the potential for RF gels to be used for vital environmental applications, but also introduces the potential way in which a computational model could be used to predict and tailor the properties of these materials for maximum effectiveness in a given application.The wide-ranging application potential for porous materials has been of significant interest over the years, with a particular focus on those which possess attractive properties, such as low densities and high surface areas. Materials such as these have proven to be effective in a wide range of applications, many of which are imperative in reducing or eradicating detrimental environmental impacts of industry, heightening their pertinence to recent research. This work focuses on one such class of organic porous materials—resorcinol–formaldehyde (RF) gels—which are formed via a sol–gel process and subsequently dried, producing the lightweight, nanoporous structure of the final gel. Despite extensive research into these materials in recent years, a number of questions still remain around their formation mechanism and the impact of various parameters associated with their synthesis. As a resultof this, their application potential is yet to be fully realised, especially given the wide range of properties that can be achieved through fine tuning and optimisation of the RF gel synthesis process. In this work, the formation mechanism of RF gels is explored through both experimental and computational means. Through experimental analysis of the final textural properties of synthesised gels, the impact of variations in catalyst concentration and catalyst species are investigated, aiming to elucidate the specific role the catalyst compound plays within the RF reaction – something that, to date, has been widely debated. The importance of the metal cation within the catalyst is highlighted through the results presented here, its concentration decoupled with the initial solution pH, and the significance of both discussed in detail. The comparative efficacy of different solvents used within the solvent exchange step of gel synthesis is also investigated, measured in their ability to preserve the structure during drying, minimising the pore shrinkage and collapse that takes place. The implications of the results obtained are discussed in relation to process optimisation to achieve desirable properties applicable to specific uses.The synthesis and analysis of RF gels is time consuming, therefore, simulating these processes computationally in an efficient manner could be pivotal to facilitating their widespread use. In this work, a three-dimensional model is developed which captures the formation and growth of RF gels using lattice-based kinetic Monte Carlo theory, accounting for varying catalyst concentration and solids content – two parameters proven to control gel properties inexperimental work. The textural properties of the resulting simulated materials are analysed, including the accessible surface area and accessible porosity, the values of which reflect the increased structural density and inter-connected complexity associated with increasing solids content and catalyst concentration. Furthermore, the fractal properties of these materials are analysed through correlation dimension and Hurst exponent calculations, the results demonstrating that while fractal properties are not typically observed in scattering experiments for RF gels, they are possible to achieve with sufficiently low solids content and catalyst concentration. As the most commonly employed method of assessing properties of porous materialsexperimentally, adsorption analysis was carried out computationally for the simulated RF gel structures. The results indicated that both low catalyst concentrations and low solids contents resulted in structures with open transport pores that were larger in width, while high catalyst concentrations and solids contents resulted in structures with bottleneck pores that were narrower. Importantly, the computational isotherm data and pore size distributions were also compared to those obtained experimentally, showing a promising agreement in trends between the two for varying catalyst concentrations, providing validation for the kinetic Monte Carlo model developed. Finally, the performance of RF gels in a specific application is tested, assessed in their ability to remove an endocrine disrupting pollutant from water through UV-Vis concentration measurements. Using the results obtained from both the experimental and computational analysis of the materials, the comparative efficacy of two RF gels synthesised under different catalyst concentrations is predicted and subsequently explored, and the properties required for optimal performance determined. This work not only highlights the potential for RF gels to be used for vital environmental applications, but also introduces the potential way in which a computational model could be used to predict and tailor the properties of these materials for maximum effectiveness in a given application
Factors influencing trust, reliance, performance and cognitive workload in human-agent collaboration
Get PDFIncreasingly, automated systems are being incorporated in collaborative environments where they are used to alleviate the cognitive load of human operators while increasing task performance. Automated agents are present in a variety of domains, from safety critical environments to leisure-oriented activities, and more and more, they are being considered as a virtual teammate rather than simple decision-aid tools. Trust is a key factor that will determine how much a human operator is willing to take into account or rely on the help provided by an automated agent. Past research on trust in automation highlights key elements that will influence its development, such as how the automated agent is perceived, how reliable the agent appears to be and how transparent its actions are. However, most related work make use of turn-based tasks where trust is measured post-hoc, which does not entirely capture the evolving aspect of trust. This thesis presents the development and use of a real-time collaborative game where human operators can choose the extent to which they rely on the help of automated agents displaying different behaviours and various levels of performance. We used different levels of task difficulty as well as survey instruments and the logging of task-specific behavioural information to elicit and measure variables that are important to understand the human-agent relationship such as trust, reliance, task performance, cognitive load or situational awareness. We ran four user-studies using this apparatus. The first study tested the effects of different levels of agent reliability and predictability on the human-agent relationship while the second study experimented with different types of agent errors. The third study tested the impact of different types of environmental uncertainty on the human-agent relationship while the fourth and final study measured the benefits ofdifferent kinds of visualisation-based decision-aid systems. Overall, this work sheds lights on under-investigated issues in Human-Agent Collaboration scenarios by providing insights on factors that are most likely to harm thehuman-agent relationship and underline how the behaviour of agents as well as the context of interaction can drastically alter a person’s attitude toward an automated agent.Increasingly, automated systems are being incorporated in collaborative environments where they are used to alleviate the cognitive load of human operators while increasing task performance. Automated agents are present in a variety of domains, from safety critical environments to leisure-oriented activities, and more and more, they are being considered as a virtual teammate rather than simple decision-aid tools. Trust is a key factor that will determine how much a human operator is willing to take into account or rely on the help provided by an automated agent. Past research on trust in automation highlights key elements that will influence its development, such as how the automated agent is perceived, how reliable the agent appears to be and how transparent its actions are. However, most related work make use of turn-based tasks where trust is measured post-hoc, which does not entirely capture the evolving aspect of trust. This thesis presents the development and use of a real-time collaborative game where human operators can choose the extent to which they rely on the help of automated agents displaying different behaviours and various levels of performance. We used different levels of task difficulty as well as survey instruments and the logging of task-specific behavioural information to elicit and measure variables that are important to understand the human-agent relationship such as trust, reliance, task performance, cognitive load or situational awareness. We ran four user-studies using this apparatus. The first study tested the effects of different levels of agent reliability and predictability on the human-agent relationship while the second study experimented with different types of agent errors. The third study tested the impact of different types of environmental uncertainty on the human-agent relationship while the fourth and final study measured the benefits ofdifferent kinds of visualisation-based decision-aid systems. Overall, this work sheds lights on under-investigated issues in Human-Agent Collaboration scenarios by providing insights on factors that are most likely to harm thehuman-agent relationship and underline how the behaviour of agents as well as the context of interaction can drastically alter a person’s attitude toward an automated agent
Data science enabled rehabilitation
Get PDFPreviously held under moratorium from 10th June 2021 until 12th June 2023.Stroke is a main cause of impairment/disability. More stroke survivors undergounsupervised home rehabilitation. Autonomous self-rehabilitation systems usingsensing and machine learning are not tailored to patients’ needs.Based on a systematic narrative literature review, home-based rehabilitation systemswere taxonomized and new design criteria were formulated for increased patientengagement enhancement and individualism. No system that addresses all the criteriawas found in literature. An in-house low-cost home-based rehabilitation AmbientIntelligence (AmI) system was deployed meeting the criteria, and an accuracyevaluation method proposed, in line with medically approved tests. The Timed Up andGo (TUG) and Five Time Sit To Stand (FTSTS) tests evaluate daily living activityperformance in the presence/development of comorbidities. The AmI-driven systemcomplies with Accountability, Responsibility, and Transparency (ART) requirementsfor wider acceptability. A method is presented for generating synthetic datasetscomplementing experimental observations mitigating bias present due to practicallimitations. Also, an incremental hybrid machine learning algorithm is proposed. Itcombines ensemble learning and hybrid stacking using extreme gradient boosteddecision trees and k-nearest neighbours to meet individualisation, and ARTrequirements while maintaining low computation footprint.The proposed approach was based on the criteria: nonintrusive, nonwearable,motivation and engagement enhancing, individualized, supporting daily activities,cost-effective, simple, and transferable. The motivation method, suitability for elderly,and intended use were examined as supplementary criteria. Indicators of enhancedmotivation and engagement, through questionnaire responses, demonstrate that >83%of participants support the proposed system’s motivation and engagementenhancement. The system is fit for purpose with statistically significant (ϱc>0.99,R2>0.94, ICC>0.96) and unbiased correlation to the gold standard. The model reachesup to 100% accuracy for FTSTS and TUG in predicting associated patient medicalcondition, and 100% or 83.13%, respectively, in predicting area of difficulty in thesegments of the test. Results show an improvement of 5% and 15% for FTSTS andTUG, over previous intrusive approaches.Keywords:Home-based rehabilitation systems, Stroke rehabilitation, Telerehabilitation, Patientparticipation, Motivation, Comparative effectiveness research, Automated timed upand go test, Automated five time sit to stand test, Self-evaluation, Evaluation ofsensor systems, Non-intrusive sensing, Sensing for health, Accountable ArtificialIntelligence, Responsible Artificial Intelligence, Transparent Artificial Intelligence,Hybrid ensemble learning, Patient-centric individualised rehabilitationStroke is a main cause of impairment/disability. More stroke survivors undergounsupervised home rehabilitation. Autonomous self-rehabilitation systems usingsensing and machine learning are not tailored to patients’ needs.Based on a systematic narrative literature review, home-based rehabilitation systemswere taxonomized and new design criteria were formulated for increased patientengagement enhancement and individualism. No system that addresses all the criteriawas found in literature. An in-house low-cost home-based rehabilitation AmbientIntelligence (AmI) system was deployed meeting the criteria, and an accuracyevaluation method proposed, in line with medically approved tests. The Timed Up andGo (TUG) and Five Time Sit To Stand (FTSTS) tests evaluate daily living activityperformance in the presence/development of comorbidities. The AmI-driven systemcomplies with Accountability, Responsibility, and Transparency (ART) requirementsfor wider acceptability. A method is presented for generating synthetic datasetscomplementing experimental observations mitigating bias present due to practicallimitations. Also, an incremental hybrid machine learning algorithm is proposed. Itcombines ensemble learning and hybrid stacking using extreme gradient boosteddecision trees and k-nearest neighbours to meet individualisation, and ARTrequirements while maintaining low computation footprint.The proposed approach was based on the criteria: nonintrusive, nonwearable,motivation and engagement enhancing, individualized, supporting daily activities,cost-effective, simple, and transferable. The motivation method, suitability for elderly,and intended use were examined as supplementary criteria. Indicators of enhancedmotivation and engagement, through questionnaire responses, demonstrate that >83%of participants support the proposed system’s motivation and engagementenhancement. The system is fit for purpose with statistically significant (ϱc>0.99,R2>0.94, ICC>0.96) and unbiased correlation to the gold standard. The model reachesup to 100% accuracy for FTSTS and TUG in predicting associated patient medicalcondition, and 100% or 83.13%, respectively, in predicting area of difficulty in thesegments of the test. Results show an improvement of 5% and 15% for FTSTS andTUG, over previous intrusive approaches.Keywords:Home-based rehabilitation systems, Stroke rehabilitation, Telerehabilitation, Patientparticipation, Motivation, Comparative effectiveness research, Automated timed upand go test, Automated five time sit to stand test, Self-evaluation, Evaluation ofsensor systems, Non-intrusive sensing, Sensing for health, Accountable ArtificialIntelligence, Responsible Artificial Intelligence, Transparent Artificial Intelligence,Hybrid ensemble learning, Patient-centric individualised rehabilitatio
Access Repository Dashboard
Do you manage STAX (Strathclyde Repository)? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!