An integrated framework for reliability prediction during product development process.

摘要

The increasing pressure of global competition impels companies to develop more innovative and reliable products in a shorter development lead-time and cost. This requirement puts immense pressure on the product development process to help in achieving the organizational goals as well as meeting increasing customer expectations. Under such extreme pressure of competition companies are looking for more effective and efficient reliability assessment programs to reduce product development time and cost and to ensure product reliability. It necessitates that organizations focus on a proactive approach of “anticipate and prevent” rather than following the traditional “make and test” approach in order to bring reliability issues and concerns upfront at the product development stage.; The objective of this research is to bring the paradigm shift in the approach of reliability assessment during the product development process. This research provides a comprehensive framework of reliability assessment during the product development process that is capable of providing good reliability estimates for small failure/test data. The proposed framework incorporates all kinds of information available at any stage of the product development process and updates the estimates as and when new evidence is available.; This research provides a formal and structured mechanism to identify failure modes and quantify the impact of design changes and corrective actions as improvement index for each failure mode. It further introduces the application of fuzzy reasoning process to combine and quantify the subjective information (improvement) and maps their impact on product reliability. The proposed framework is based on the Bayesian approach. The integration of the fuzzy reasoning model enhances the capability of the Bayesian method to deal with probabilistic as well as subjective uncertainty in reliability evaluation. Finally, this research presents a very simple and practical approach to utilize reliability prediction numbers in locating critical source/cause and help in planning further actions for reliability demonstration and improvement.