AbstractThis dissertation identified a gap in research for an intelligent monitoring system to monitor various indicators within complex engineering industries, in order to predict the potential situations that may lead to catastrophic failures. The accuracy of prediction was based upon lessons learnt from historic catastrophic incidents. These incidents are normally attributed to combinations of several minor errors or failures, and seldom occur through single point failures. The new system to monitor, identify and predict the conditions likely to cause a catastrophic failure could improve safety, reduce down time and prioritise funding.
This novel approach involved the classification of ten common traits that are known to lead to catastrophe, based on six headings used by the Health and Safety Executive and four headings used in Engineering Governance. These were weight averaged to provide a ‘state’ condition for each asset, and amalgamated with a qualitative fault tree representation of a known catastrophic failure type. The information on current ‘state’ was plotted onto a coloured 2D surface graph over a period of time to demonstrate one particular visual tool. The research demonstrated that it was possible to create the monitoring system within Microsoft Excel and to run Visual Basic programs alongside Boolean logic calculations for the fault tree and the predictive tools, based upon the trend analysis of historic data. Another significant research success was the development of a standardised approach to the investigation of incidents and the dissemination of information.
|Date of Award||Jan 2014|
|Supervisor||David Sanders (Supervisor), Nick Capon (Supervisor) & Giles Tewkesbury (Supervisor)|