Driven by the need to reduce business risk, the next decade will see increasing pressure on vehicle and component manufacturers, not only to improve reliability and reduce costs, but also to provide formal demonstrations of product reliability and safety achievement in advance of product launch. In addition, assessment of product reliability early in the design stage has the benefit of shortening lead times and reducing the need for expensive validation testing. Current tools to support such demonstrations are strongly dependent on the availability of failure data generated in service or from extensive testing. Such data are not always available or easily adapted to meet the needs of the designer, especially when the product is new or contains a degree of novelty.
The research described in this paper is funded through the IMI Foresight Vehicle Programme and is focused on the development of reliability tools applicable at the design stage. The core goal for the research (code named Prescient) is the integration of a set of tools to enable the designer to demonstrate failure knowledge and reliability improvement expertise, and to forecast the impact of this capability on failure rate or probability of failure at component and system level during design.
The paper will describe the Prescient tool box and some of its key facilities for: failure visualization; performance of FMECA; component reliability forecasting using historical or test data or stress strength interference models; systems reliability analysis; construction of lessons learned knowledge databases and reliability improvement. Links to external software, e.g. influence diagrams, Bayesian belief nets and commercial reliability tools, will be demonstrated. The paper will illustrate the use of the tool with examples drawn from the design of the TRW Electrically Assisted Steering System.