The process of designing and producing mechanical and structural systems consists of several stages, starting from defining the requirements and the demands throughout the life cycle, that must be supported to determine the capacity or resistance needed to fulfil the equipment mission. The reliability is the one of the most important goals that manufacturers seek, while the economical aspect is a key factor and it has a great deal influence on this process. Therefore, the best design has to be carried out, in order to achieve the paradox of reliable products with minimal costs. This implies careful and exact investigation along the product life-cycle, from birth to death. This study encompasses the different phases of product life cycle, starting from the necessity to demonstrate the product reliability before starting the mass production under the constraints of economy and time. This work aims to answer the following three questions: * How can we demonstrate the product reliability on the basis of few tests? Among the four approaches considered in this study, the method of compound uncertainties shows its robustness to demonstrate the product reliability, without implying unnecessary over-design. * What is the robust design criterion under repetitive load for time-independent resistance? In the useful phase of product life cycle, failure is assigned to load variability under the assumption of non-degraded resistance. Stress-resistance model considers failure probability as a design target to be achieved; however, for the case of repetitive loading, this criterion is sensitive to the number of load applications. The present works shows that hazard is almost constant and gives a robust design criterion. * Is there a general approach that can cope with all degradation mechanisms? In the wear out phase, modelling degradation is mandatory for several reasons such as industrial risks control and life cycle management. The hazard function gives an appropriate indicator for the prediction of the degradation state and consequently, the estimation of residual product life.
