Supervisors: Dr Sarah Dunnett (Department of Aeronautical and Automotive Engineering), Dr Paul Bell (Department of Computer Science)
During the design of any industrial system, an assessment of the systemâs reliability should be performed. Such an assessment uses mathematical models to determine the failure probability of the system. If results from such an assessment are obtained early on in the design process, then they can be used to investigate different design options and can influence the development of the system. This is the most beneficial way of applying the models. Generally however, results from these mathematical models are obtained too late to effectively influence the design process. This is due to the time spent on generating the mathematical models from a description of the engineering system, which is still done by hand. If this stage could be performed by an automated generation process then full advantage could be taken of the reliability assessment in the design process.Â
This project will address this area to develop algorithms to take the system description in a form currently used in industry (i.e. a CAD or PID) and translate it into a representation facilitating the generation of the appropriate reliability model automatically.Â The project will involve the derivation of suitable computer / automation techniques which encompass both qualitative reasoning and expert knowledge encapsulation, as well as the causal relationships of the components within the system.Â It is anticipated that a variety of artificial intelligence and reliability techniques will be used. With an automated procedure, the ability to use the derived procedures in decision-making will allow the diagnosis of system state throughout the lifecycle, enabling changing and dynamic system behaviour to be considered.Â This ability for intelligent interpretation of system behaviour and its impact on reliability issues is currently very limited and hence an area for this research.Â
If you are interested in the above project and require further information please contact Dr Sarah Dunnett: [email protected]Â
A studentship is available to fund the research for UK/EU nationals.
The Studentship covers tuition fees and provides a tax-free stipend of Â£13,590 p.a.for a three-year duration. Â Applicants must hold a good honours degree (first class or an upper second) in Engineering , Mathematics or Computer Science. The start date for the project is expected to be July 2013. Applications can be accessed on lineÂ by clicking theÂ Apply link below.