This article proposes a stochastic optimisation model in order to reduce the long-term total maintenance cost of complex systems. The proposed work is based on the following approaches: (i) optimisation of a cost model for complex multi-component systems consisting of preventive and corrective maintenance using reliability analysis , which faces two different maintenance policies (periodic block-type and age-based) and (ii) a clustering method for maintenance actions to decrease the total maintenance cost of the complex system. This work evaluates each maintenance policy and measures the effects on imperfect maintenance actions. Finally, the proposed optimisation model is applied to a numerical example which focuses on passenger urban aerial ro-peway transport systems, in which the current maintenance policy has been evaluated, considering the established by the international regulation of passenger aerial cable cars.