Browsing by Author "Morineau, Julien"

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    Cost modelling for cloud computing utilisation in long term digital preservation
    (Cranfield University Press, 2013-09-19) Shehab, Essam; Sanya, Isaac; Badawy, Mohamed; Ocal, Fatma; Namiesnik, Bartosz; Morineau, Julien; Odika, Stephen; Fernandez Ortiz, Zaira
    The rapid increase in volume of digital information can cause concern among organisations regarding manageability, costs and security of their information in the long-term. As cloud computing technology is often used for digital preservation purposes and is still evolving, there is difficulty in determining its long-term costs. This paper presents the development of a generic cost model for public and private clouds utilisation in long term digital preservation (LTDP), considering the impact of uncertainties and obsolescence issues. The cost model consists of rules and assumptions and was built using a combination of activity based and parametric cost estimation techniques. After generation of cost breakdown structures for both clouds, uncertainties and obsolescence were categorised. To quantify impacts of uncertainties on cost, three-point estimate technique was employed and Monte Carlo simulation was applied to generate the probability distribution on each cost driver. A decision support cost estimation tool with dashboard representation of results was developed.
  • Thumbnail Image
    ItemOpen Access
    Precise vehicle location as a fundamental parameter for intelligent selfaware rail-track maintenance systems
    (Elsevier, 2014-10-31) Durazo-Cardenas, Isidro; Starr, Andrew G.; Tsourdos, Antonios; Bevilacqua, Maurizio; Morineau, Julien
    The rail industry in the UK is undergoing substantial changes in response to a modernisation vision for 2040. Development and implementation of these will lead to a highly automated and safe railway. Real-time regulation of traffic will optimise the performance of the network, with trains running in succession within an adjacent movable safety zone. Critically, maintenance will use intelligent trainborne and track-based systems. These will provide accurate and timely information for condition based intervention at precise track locations, reducing possession downtime and minimising the presence of workers in operating railways. Clearly, precise knowledge of trains’ real-time location is of paramount importance. The positional accuracy demand of the future railway is less than 2m. A critical consideration of this requirement is the capability to resolve train occupancy in adjacent tracks, with the highest degree of confidence. A finer resolution is required for locating faults such as damage or missing parts, precisely. Location of trains currently relies on track signalling technology. However, these systems mostly provide an indication of the presence of trains within discrete track sections. The standard Global Navigation Satellite Systems (GNSS), cannot precisely and reliably resolve location as required either. Within the context of the needs of the future railway, state of the art location technologies and systems were reviewed and critiqued. It was found that no current technology is able to resolve location as required. Uncertainty is a significant factor. A new integrated approach employing complimentary technologies and more efficient data fusion process, can potentially offer a more accurate and robust solution. Data fusion architectures enabling intelligent self-aware rail-track maintenance systems are proposed.