The degradation of components in the industry is a natural phenomenon resulting from various operating conditions. Therefore, it is crucial to develop models and strategies to anticipate failure events. Reliability-centred maintenance has been widely used and has proven to be effective. However, the reliability function alone does not provide information on how quickly a component will reach a critical level of degradation. For this reason, our work proposes the use of a piecewise, non-parametric hazard rate function (HRF) for more informed scheduling of maintenance actions, considering the possibility of imperfect maintenance actions. Specifically, the piecewise mean time between failures MTBF is used to estimate the maintenance period over the life cycle of a component. The component failure behaviour is modelled using the kernel density estimation (KDE) approach, which is a non-parametric method. The KDE is used to estimate both the reliability function and the piecewise HRF. A simulation analysis is performed using a component with historical failure data in terms of working cycles. Preventive and corrective maintenance costs are projected, taking into account the maintenance period estimated from the MTBF in sections. Our methodology seeks to improve the efficiency and effectiveness of maintenance actions by considering the degradation behaviour of the component and providing an optimal schedule based on the piecewise HRF and the estimated maintenance period.