Manufacturing system degradation can damage its reliability, resulting in decreased product quality and delayed deliveries. These challenges are characteristic of imperfect manufacturing systems. Moreover, the propagation of delay time across task periods may reduce the operational availability in future periods. Condition-based maintenance is an effective method for mitigating system degradation and enhancing reliability. However, existing condition-based maintenance studies often overlook the impact of delay propagation on operational availability. To address this issue, this paper proposes a condition-based maintenance model based on a Markov decision process. By introducing delay time as a state variable to capture changes in operational availability and incorporating it into the reward model, the proposed strategy aims to maximize enterprise profit. A case study and comparative analysis using data from a manufacturing enterprise validate the effectiveness and superiority of the proposed model in improving economic performance.