This thesis addresses the issue as to how to model a student’s motivation when using a learning environment. Currently, this is considered as an important topic since a) motivation is regarded as key to successful learning by educationalists, b) those building computer-based learning environments are attempting to design in features that will assist in motivating learners while using their systems, c) researchers in the areas of intelligent learning environments (ILEs) and intelligent tutoring systems (ITSs) are looking for methods to detect a student’s motivation during their interaction with an adaptive system, and d) researchers in ITSs are looking for strategies that a learning environment could use to motivate a student or to scaffold his/her motivation. A well-founded model that partially explains the way in which motivation changes during an interaction with an ILE is potentially very valuable to these different goals. In particular, the model should prove to be of great benefit when creating learning environments that take into account the motivational aspects of the learners since it represents the process of how they are motivated through the use of an ILE. In this thesis the focus is on the construction of a learner’s motivational structure for an ILE grounded in the context of an educational game. The thesis reports the development of a qualitative model of the motivation of learners during their interaction with the ILE. One of the main issues is the specification of the context as the motivational structure of learners cannot be assumed to be the same in different contexts. A preliminary causal model showing the relationships between a learner’s motivational characteristics and features of the ILE was developed. Two computer-based research instruments were then built: a game prototype aiming to teach some concepts in databases (Alex’s Adventure) and a computer program (MoRes). Alex’s Adventure was developed in such a way that it included all features of the ILE presented in the model whereas the other parts of the model were implemented through MoRes. Six detailed case studies were performed in order to validate the preliminary model. The validation resulted in changes to the preliminary model which are believed to produce an improved model. The analysis also resulted in a number of key points and condition–action rules which are considered to be of use for the implementation of future adaptive computer systems that could manage and support learners. The thesis makes an original contribution to the Artificial Intelligence in Education (AIED) community, particularly in terms of the method of case study analysis and the new model of motivation.

This thesis is concerned with the realisation of leg cycling by means of FES in SCI individuals with complete paraplegia. FES lower-limb cycling can be safely performed by paraplegics on static ergometers or recumbent tricycles. In this work, different FES cycling systems were developed for clinical and home use. Two design approaches have been followed. The first is based on the adaptation of commercially available recumbent tricycles. This results in devices which can be used as static trainers or for mobile cycling. The second design approach utilises a commercially available motorised ergometer which can be operated while sitting in a wheelchair. The developed FES cycling systems can be operated in isotonic (constant cycling resistance) or isokinetic mode (constant cadence) when used as static trainers. This represents a novelty compared to existing FES cycling systems. In order to realise isokinetic cycling, an electric motor is needed to assist or resist the cycling movement to maintain a constant cadence. Repetitive control technology is applied to the motor in this context to virtually eliminate disturbance caused by the FES activated musculature which are periodic with respect to the cadence. Furthermore, new methods for feedback control of the patient’s work rate have been introduced. A one year pilot study on FES cycling with paraplegic subjects has been carried out. Effective indoor cycling on a trainer setup could be achieved for long periods up to an hour, and mobile outdoor cycling was performed over useful distances. Power output of FES cycling was in the range of 15 to 20 W for two of the three subjects at the end of the pilot study. A muscle strengthening programme was carried out prior and concurrent to the FES cycling. Feedback control of FES assisted weight lifting exercises by quadriceps stimulation has been studied in this context.

This thesis presents research carried out for the visualisation of surface anatomy data stored as large range images such as those produced by stereo-photogrammetric, and other triangulation-based capture devices. As part of this research, I explored the use of points as a rendering primitive as opposed to polygons, and the use of range images as the native data representation. Using points as a display primitive as opposed to polygons required the creation of a pipeline that solved problems associated with point-based rendering. The problems inves tigated were scattered-data interpolation (a common problem with point-based rendering), multi-view rendering, multi-resolution representations, anti-aliasing, and hidden-point re- moval. In addition, an efficient real-time implementation on the GPU was carried out.

In this work, some developments in the theory of modelling integrated optical devices are discussed. The theory of the Beam Propagation Method (BPM) to analyse longitudinal optical waveguides is established. The BPM is then formulated and implemented numerically to study both two and three-dimensional optical waveguides using several Finite-Difference (FD) techniques. For the 2-D analysis, comparisons between the performance of the implicit Crank Nicholson (CN), the explicit Real Space (RS) and the Explicit Finite-Difference (EFD) are made through systematic tests on slab waveguide geometries. For three-dimensional applications, two explicit highly-parallel three-dimensional FD-BPMs (the RS and the EFD) have been implemented on two different parallel computers, namely a transputer array (MIMD type) and a Connection Machine (SIMD type). To assess the performance of parallel computers in this context, serial computer codes for the two methods have been implemented and a comparison between the speed of the serial and parallel codes has been made. Large gains in the speed of the parallel FD-BPMs have been obtained compared to the serial implementations; both methods, in their parallel form, can execute, per propagational step, a large problem containing 106 discretisation points in a few seconds. In addition, a comparison between the performance of the transputer array and the Connection Machine in executing the two FD-BPMs has been discussed. To assess and compare the two methods, three different rib waveguides and three different directional couplers have been analysed and the results compared with published results. It has been concluded from testing these methods that the parallel EFD-BPM is more efficient than the parallel RS-BPM. Then, the linear parallel EFD-BPM was extended to model nonlinear second harmonic generation process in three-dimensional waveguides, where the source field is allowed to deplete, using the transputer array and the Connection Machine.

The work described in this thesis is aimed at exploring the possibility of optically integrating an OTDM transmitter operating at 4X10Gb/s on an appropriate substrate. It has been shown that such an OTDM transmitter system could be integrated on III-V semiconductor quantum well (QW) substrates if the design of the substrate, the choice of fabrication techniques and the design of the devices are carefully considered. Suitable device structures for the three main kinds of devices involved in OTDM transmitters, namely light source, optical multiplexers (couplers) and optical modulators, have been discussed. Significant progress regarding these aspects, both theoretical and experimental, has been achieved. In this work, it has been intended to investigate all the devices from the integration point of view. This has been reflected in many aspects in the device design and fabrication process. Integration has always been a very important factor to consider in the determination of substrate material structure, device configuration, waveguide structure and fabrication techniques. As a result, the devices developed in this project are suitable for the proposed purpose of an integrated OTDM transmitter system. Investigation into integration techniques has also been carried out. The most important was to introduce bandgap difference on a semiconductor QW material. IFVD technique is studied and produced some encouraging results such as the extended cavity SRL, which integrates an active section with a passive MMI coupler. Vertically coupled waveguide structures have also been invstigated in an attempt to produce extended cavity lasers. The design considerations of extended cavity lasers employing this waveguide structure have been discussed.

The main aim of this study is to use electromyography to study muscle activation during natural, unrestrained movements. At total of fifty-seven male volunteers participated in the experiments. There were three main studies: vertical jumping, kicking a tethered football and isokinetic dynamometry. These experiments were designed to investigate the hamstrings activity during high-velocity knee extension movements in an attempt to relate the magnitude of hamstrings co-activation and the timing of hamstrings activity to the speed or power of the movements. It is clear that there is substantial co-activation of the three muscles in hamstrings and vastus lateralis in all three studies. Co-activation is present in almost all volunteers even in the slowest speed and lowest power movements studied. Co-activation has been reported by others during single knee extension movements on isokinetic dynamometers. This thesis reports for the first time that the extent of co-activation changes during repeated movements. Co-activation occurs during unrestrained vertical jumps across the whole range from the lowest power jumps in which the volunteer barely leaves the ground to maximum power jumps. In addition, it is commonly observed even in professional football players capable of producing very fast knee extensions during powerful kicks. The duration of co-activation of hamstrings during kicking was significantly shorter in the highly trained 15-year-old footballers that in their 11-year-old counterparts or in untrained adults. It is possible that this reflects changes in the way their kicking movements have developed with prolonged training. This is the first study of age related changes in co-activation. Interestingly, there is anecdotal evidence from the club coaches that injuries are far more frequent in the 15 year olds than in the younger teams. It may be that the increase in speed of movement achieved by reducing co-activation, places the limb at more risk.

This thesis examines information imperfections in asset markets and its impact on economic performance through technology adoption and innovation. In a rational setting, where equity market participants take into account common public information in addition to their private signals about fundamentals, equity prices are persistently biased towards the public signals. Chapter 2 investigates the real effect of such mis-pricing, when R&D producing firms rely on equity finance. Relating to the recent technology stocks boom, the model shows how market's optimism causes more innovations. Furthermore, such optimism can generate gains in aggregate consumption. Chapter 3 analyzes equity markets' role in facilitating ownership transfer from entrepreneurs investing in adopting technology to managers running these firms once technology is adopted. Information imperfections in equity market affect entrepreneurs' willingness to invest in frontier technology in two ways. First, uncertainty about equity price or lack of market liquidity discourages technology adoption. This can explain slow technology adoption and limited venture capitalists' participation in under-developed equity markets. Second, imperfectly informed market participants take fast adoption as a positive signal. The resulting increase of expected market value encourages technology adoption. Probability of fast technology adoption is highest at an intermediate number of informed investors. Chapter 4 looks more closely into the extent of asset mis-pricing by endogenizing the variance of investors' private signals. Better quality of freely available public information reduces incentives to invest in private information and can magnify the extent of asset mis-pricing. Furthermore, in a dynamic setting, investors' react more slowly on changes of the fundamentals because incentives to invest in research are low in early trading periods. The chapter also shows that availability of longer price history might not bring asset prices closer to the fundamentals, as investors choose to free-ride on other investors' research efforts.

This study explores the phenomenon of the social construction of IS investment evaluation in the financial services industry. The extensive literature on information systems (IS) evaluation stretches back more than thirty years and it is widely agreed that IS evaluation is an important and complex issue. Constructivist evaluation, which serves to develop the interpretivist strand of the literature, concentrates on the notion that evaluations are the outcomes of an interaction and argumentation process between various interested parties. The purpose of this research is to shed further light on the IS investment evaluation area. Explaining the practice of IS evaluation within the dynamic and technologically sophisticated financial services industry would benefit academics and financial services firms alike. This context is relatively neglected in the existing literature. The chosen case study reflects the demands in the financial services industry to upgrade their infrastructure in order to manage an ever increasing number of transactions, as well as increased regulation. The fieldwork produced extensive data concerning traditional trading, brokerage, high frequency algorithmic trading and transaction clearing. The rapid rise of derivative transactions during the last few years caused problems in terms of clearing and managing the transactions. At the same time, hedge funds‘ growing demand for low latency execution services and changes in regulations increased, pressuring firms in the industry to invest heavily. Actor-network theory (ANT) is used in order to describe and explain the formation of networks between human and non-human actors. The theory‘s specific vocabulary allows IS evaluation to be seen in a new light and the study uses ANT analysis to produce insight into the longstanding problem of IS evaluation, human and non-human interactions and roles within the context if IT investment appraisal. Building on ANT, this thesis is an in-depth case study of the employment process of an IT evaluation method at a global financial services company.

In children and newborn babies on intensive care, information regarding blood oxygen saturation (SpO2) is determined non-invasively by a device called a pulse oximeter. Sensors are usually placed on a hand or foot where their operation relies on the presence of pulsatile arterial blood. Light shines at two or more wavelengths (usually red and infrared) into the tissue where the pulsatile blood modulates, absorbs and scatters the different wavelengths of light in varying amounts and is detected by a photo-detector as a photoplethysmograph (PPG). The spectral information received is then processed electronically and digitally to determine the amount of haemoglobin present. In the sickest of children blood supply can become compromised to these sensor locations and the pulsatile component of the blood may diminish and pulse oximeter readings may become unreliable, especially at times when accurate blood oxygen information would be vital. Currently the alternative is to take blood from an arterial line and run a relatively lengthy analysis (pulse oximeters are near-instantaneous in their operation) that may be unnecessary if the pulse oximeter could be relied upon at these critical moments. In the smallest of babies invasive sampling of blood becomes even more of an issue as any blood loss could lead to hypovolaemia and introduce extra sites of infection plus it causes a lot of stress to the neonate. Since central blood flow may be preferentially preserved, the anterior fontanelle was investigated as an alternative monitoring site. Custom reflectance fontanelle and reference PPG sensors have been designed and built to investigate the fontanelle in those children at risk of peripheral supply compromise. Dedicated instrumentation and software has also been successfully developed for the control of the sensor electronics and the data-logging of PPG signals for retrospective analysis. Sixteen neonates were recruited for fontanelle monitoring; all were ASA 1 – 3 (ASA ranges from 1 to 5 where 1 is the least sick and 5 is the most critically ill). As part of the approved protocol the delivered oxygen to the patients was artificially altered to look for corresponding changes in PPG signal amplitudes. Amplitude results reveal strong correlations (R > 0.5) between the reference sensor (placed on the foot) and the fontanelle sensor. This suggests that the fontanelle sensor is sensitive to changes in amplitude when oxygen in the blood alters. Correlation of the health of the child, using the ASA score, and the difference in amplitudes of PPGs between the sensors reveals that the fontanelle sensor does detect increasing fontanelle PPG amplitudes when compared to the PPGs from the reference sensor the sicker the child is, confirming that pulsatile flow is being preferentially preserved at the fontanelle in those children who are the most at risk from peripheral supply compromise. SpO2 estimation at the fontanelle reveals a mean difference of 2.2 % to the SpO2 as read by the commercial device and a 1.7 % difference to the blood gas results. These results confirm that the anterior fontanelle may be used as an alternative location for SpO2 measurement in those who are at most risk of peripheral supply compromise.

The impact of uncertainty in the commercialization lifecycle of new technologies is a complex phenomenon. Technologies are research intensive and exposed to uncertainty regarding their successful development and functionality. Further these technologies have to be absorbed by volatile markets in order to be commercialized. These different forms of uncertainty are of primary importance for decision makers but have not been thoroughly studied in previous technology commercialization research and put under one theoretical framework. The main focus of this thesis is to comprehend the recently growing trend among universities and public research organizations to commercialize their research activities from an empirical and theoretical perspective. More particularly the thesis focuses on the life cycle of two main commercialization streams namely the entry and exit of university spinouts, which are companies that evolve from intellectual property developed within academic institutions as well as the licensing and licensing termination of inventions. The main focus of the thesis therefore analysesmarket and technological uncertainty and explains the conditions under which spinout formation, spinout failure, licensing and licensing failure occur by putting them under the theoretical framework of real option theory.