For small and medium sized enterprises, portfolio management can be a difficult exercise; especially when the available paths forwards are shrouded in mystery. To overcome this, they are required to research and decide upon the path to follow, but this can be difficult when the information to gather is unclear, leading to ad-hoc processes and inconsistency. The motivation for this project originated from a small and medium sized enterprise experiencing this problem of unknown critical information and a technological portfolio with significant potential. With no procedure or method in place, they conducted the research in an ad-hoc way resulting in uncertainty and low repeatability of decisions. To tackle this problem, innovation structuring frameworks were synthesised with appropriate risks and context to pose a new structure for information capture. The resulting structure was tested within the small and medium sized enterprise to investigate the repeatability of the information capture and the subsequent ranking. The proposed structure was also analysed by — process experts for its wider applicability. For the small and medium sized enterprise, this led to a consistent and repeatable method that delivered increased confidence about selecting a path forwards for its portfolio. In addition, it was found to be applicable to external organisations increasing the model’s worth and applicability. The implications of this work have led to a change in the operational procedure of the small and medium sized enterprise to utilise the defined process for researching new ideas for their portfolio which can lead to repeatable and trustworthy decisions. This also has applicability to other similar companies and could lead them through a repeatable process as presented here.

When companies engage in innovation, the appropriate selection of projects to invest resource in is paramount. In order to do this effectively, they need to research appropriate opportunities to create sufficient understanding. The various opportunities available need to be rationalised to match with the resource available. There are several rationalisation methods available, including Portfolio Management, Scoring Methods and Decision Support Systems. However, there are few that combine to be utilised by Small and Medium Sized Enterprises effectively. This work adds to the field of Small and Medium Sized Enterprise Decision Support by proposing an approach combining opportunity investigation, review and recommendation such that the most appropriate candidate innovation can be selected and taken forwards for development.

To achieve competitive advantage, many companies need to engage and invest in Research and Development. For this investment to be effective, resources need to be allocated appropriately across all projects. However, when the portfolio of the company is diverse or large, this assignment can be challenging. Portfolio Management has been created as a method for companies to effectively manage new, existing and potential projects. Yet, these methods can introduce bias and subjectivity without being flexible to the pieces of information, or attributes that are important to the company. This work adds to the field by proposing three scoring methods that convert any attribute into a numerical representation that can then be used for comparison. For managers, it means that they can select any attributes of importance to them to allow their portfolio to be prioritised and have the resource allocated appropriately to the projects that offer the greatest promise.

A variety of full-field displacement and strain measurement methods (1) have been applied to the testing of composite materials and structures. These include methods based on the use of video cameras via normalised correlation approaches; the use of grid methods based on moiré fringe methods; the use of laser speckle interferometry methods and the use of photoelasticity . The great advantage that all these techniques have is that the output is in terms of the distribution of displacement and, with additional processing, strain across the surface of the sample. This allows the direct visualisation of strain fields, for example for comparison with FEA outputs and may be useful in visualising changes in surface strain caused by sub-surface events. The application of these methods has been very valuable in demonstrating how non-uniform the stress fields may be in conventional materials test specimens, especially with composite materials. Having said that, the evidence that such methods provide an adequately accurate and reliable method of determining materials properties is somewhat mixed, with rather variable results being obtained when measuring the moduli of standard engineering materials. For the majority of these methods the measurements that are made are arrived by off-line computations, and in some cases such as ESPI the individual snapshots that are used to capture a changing load and strain environment are taken with the specimens in a fixed position, potentially allowing creep and strain redistribution to occur. The work on the Video Gauge that is reported here uses similar algorithms to those used in normalised correlation approaches to track and measure the position of targets in a video field, and then applies additional and novel algorithms to further refine the estimate of position. However, rather than effectively mapping the sample with an array of targets that covers the whole surface (which usually requires a speckle pattern to be applied) a set of specific targets are identified on or added to the surface and the displacements of and strains between these targets are tracked and measured. This has several advantages. One major advantage is that with a reduced computational load the displacements and/or strains can be tracked in real time and hence used directly to provide feedback control to the test machine.

Clinical diagnosis of pathological conditions is accomplished regularly via the recording and subsequent analysis of a physiological variable from a subject. Problems with current common practice centre around the obtrusive and rigid nature of this process. These include the length, timing and location of the diagnostic recording session, transfer of data to clinical staff, liaison between clinical staff and subjects and the integration of such diagnostic check-ups into the overall health care process. We have designed a modular diagnostic monitor that is centered around a wearable computer system which, when integrated into a suitable computer network and database architecture, is capable of addressing the above problems. The system is modular, allowing researchers and practitioners to utilise various sensor modules, reconfigure the unit in terms of its on-board storage and wireless telemetry capabilities, select the appropriate level of data preprocessing (before archiving data) and choose the appropriate level and nature of feedback to the subject. The system is GRID enabled, supporting e-clinical-trials. GRID clients can display live data, historical data, or perform data mining.

In this paper we show how we have used and adapted GT3 to support scalable and flexible remote medical monitoring applications on the Grid. We use two lightweight monitoring devices (a java phone and a wearable computer), which monitor blood glucose levels and ECG/SpO2 activity. We have connected those devices to the Grid by means of proxies, allowing those devices to be intermittently connected. The data from the devices is collected in a database on the Grid, and practitioners can obtain real time data or observe the patients historical data.

Current techniques for road-traffic monitoring rely on sensors which have limited capabilities, are inflexible and often, both costly and disruptive to install. The use of video cameras (many of which are already installed to survey road net works), coupled with computer vision techniques offers an attractive alternative to current sensors. Vision based sensors have the potential to measure a far greater variety of traffic parameters compared to conventional sensors. This thesis presents two vision based traffic-monitoring systems. The first is a number-plate recognition system. This is capable of monitoring the output from a video camera and detecting when a vehicle passes by. At this moment an image is captured and the vehicle''s number-plate is located and deciphered. The second system is a generic road-traffic monitoring sensor which utilises model based techniques to track vehicles as they manoeuvre through complex road scenes. The position of the vehicle in the image is transformed to the vehicle''s position in the real world enabling, among other things, vehicle speed and path to be easily measured. The development of each system is described in detail and results from testing the systems on images from real traffic scenes are presented.