Deploying condition-based maintenance (CBM) techniques in a fleet of Cyber-physical systems (CPSs) is not always a trivial task because of the complexity of these systems. Over the years different techniques have been utilized in the diagnosis and prognosis of CPSs. This information has been used to facilitate maintenance decision making in long term and short term. From the existing literature, the authors have remarked that, most of the proposed methodologies deal with CBM for individual CPSs with less attention on the cases where CPSs form fleets, in which each CPS is comprised of similar types of subsystems or equipments. The authors therefore, address the fleet level of CPSs through which the existing diagnosis and decision making methodologies can be restudied. This article proposes an original way to support fleet maintainer's decision using the fuzzy logic methodology approach in diagnosis and Pareto optimization process for decision making. An application to the pantograph of a fleet of trains is proposed.