This paper offers an alternative to determine reliability-centered maintenance (RCM) schemes for replaceable systems, when replacement times are censored and only the information that maintenance technicians, from the subjectivity of their experience, is available. Using differential entropy in information theory, and exploiting Lagrangian optimization algorithms, a Generalized Probability Density of Maximum Entropy (GPDME) is extracted. Lagrangian techniques provide a set of parameters that characterize the GPDME, the estimation of the parameters is done by first order perturbation of the integral of non-central moments, with which, the GPDME is typically built. In the emerging industry, RCM maintenance plans are not a common standard, in an attempt to put into practice, the benefits of RCM to this industrial segment, a case study, where the presented methodology was applied is provided. In the discussion and conclusions section, the areas of opportunity that are observed in the methodology presented in this work are adressed.