Electromechanical and mechanical equipment reliability databases do not seem as widespread as those of electronic which are well established and commonly used by the whole industrial actors. This statement explains itself by the nature of its which do not propose prediction reliability model but present some similar categories of equipment. Indeed, no method was still commonly allowed for the prediction of the reliability of (electro-)mechanical equipment and the dispersion of failure rates for apparently similar system is significant. In order to assess the failure rates, we propose a method taking into account influential factors and ageing. A small application is presented in this article on fictitious data in order to clarify the method and show the impact of these contributors on failure rates.