Reliability index - a method to predict failure rate and monitor maturity of mobile phones

This paper describes the development of a new measure ´reliability index´ to measure and monitor the product maturity and to estimate the product field failure rate during the R&D phase. Nokia a traditional reliability prediction method, British Telecom (BT) for a number of years, but with limited value as the prediction method principally addressed only electrical and electronic components, which contributed least to field returns. Hence, as mobile phones became more complex with new technologies, electromechanical devices, the deviation between the predicted and actual field failure rates increased. This created an increasing demand to develop a method to estimate the field failure rate of a product prior to its launch to the field. An estimate of the field failure rate of a product is a key requirement for a product development program, as this information is used to compare with field failure rate targets, make budgetary estimates, identify warranty support requirements, improve product competitiveness and to get an understanding that customer expectations are likely to be met. During development of a product several build samples are made prior to the product launch. These builds are aimed to improve the product maturity through improvements in; manufacturability, testability, performance, process capability, quality and reliability. These build samples are subjected to a variety of activities in different product development areas. In product testing a sample of each product build is subjected to several reliability, environmental and Electro magnetic compatibility (EMC) tests. Hence, a study was initiated to develop a product level reliability prediction tool and was carried out on recently launched products. For each test a ´weight´ was allocated based on the effectiveness of the test to simulate field typical failures. Due to the severe use conditions of mobile phones, often more severe and difficult to pass tests are used to simulate field typical- - failures and they are allocated higher test weights. For launched products, it is possible to obtain the list of applicable tests and their outcome. Hence the overall weighted test outcome at the product launch is obtained. This would include the sum test weights for all tests passed at the product launch, while the maximum possible test weight would be the sum of individual test weights for all applicable tests. The ratio of the achieved test outcome to the maximum possible test outcome as a percentage is named as the reliability index (RI)%. It can be assumed that products with higher RI% values would be more reliable than those with lower RI% values. This implies that the field failure rate of product with higher RI% values would be lower than those with lower RI% values. That is (100-RI)% would be proportional to the field failure rate. The methodology was studied on a number of products and correlation of (100-RI)% with the field failure rate was studied with different regression models. The findings are encouraging, a logarithmic regression model was found most suitable to describe the correlation with a regression coefficient in the order of 0.9. As tests are focused on hardware development, a better correlation with hardware field failure data was observed. The tool was effective in predicting the reliability of products at the launch and a very good correlation between predicted and actual field failure rates has been observed