If a repairable system is partially repaired or renewed, its failure intensity can be modelled by general repair processes. Dorado et al. (1997) proposed a unified model, which is an extension of several general repair process models, and a non-parametric method for this parameter inference. Nevertheless, its application to real world data has seldom been shown. This paper provides an example of the statistical inference of the model and its application to a real-world dataset, collected from a repairable system. The model enables to identify the extent of modification of the intensity and age of the system after repairs