Optimized test planning to verify process reliability through Six Sigma

This paper presents a FIAT AUTO study aimed to improve the test planning methodology by reducing the standard car’s number (firstly over 7000) needed to validate the production process reliability, with a minimum loss in strength of detecting and fixing the faults. The consolidated reference method is Reliability Growth Testing concerning the infant mortality with most of tests covering distances not longer than 180 km. The activities have been developed in two sub-projects, both using Six Sigma approach: feasibility first and then main work. The reduction of the testing car’s number was obtained by combining different available types of test, taking into account that the chance of noticing defects rises depending on both the test distance and the sample size. The selected optimal solution needs less than 4000 cars.