Preventive maintenance of a mechatronic fleet aims at avoiding unexpected failures due to degradation. This task requires the evaluation of degradation kinetics, which is the objective of  the Uptime physics of failure model-library. It was originally developed for product development, in particular validation. This knowledge-base on failure risks is re-used for fleet analytics to evaluate the remaining useful life once a deviation is detected. However, since damage kinetics strongly depend on the root cause a diagnosis method is required to identify the failure mechanism for selecting the proper damage model. This diagnosis task is also supported by the knowledge-base. It delivers indicators for degradation, which are used as input for the Uptime model-based reasoning engine. 

This knowledge-based analytic system is growing along with consultancy and software applications. It is modular and highly generic, thus quickly applicable to various mechantronic fleets. Examples from several industries illustrate the potential of this approach. A corresponding recommendation system demonstrates the benefit for a large-scale fleet maintenance process.