Joint optimization of condition-based production and maintenance with mutual production-deterioration dependencies

Production control and maintenance scheduling are the two major cost-effective methodologies in the manufacturing systems. The existing literature considers these processes in isolation or with unidirectional dependencies. To fill this gap, we propose an integrated control model of maintenance and production process considering the mutual influence between production and deterioration in a single-machine system. By leveraging condition monitoring data, we enhance decision-making effectiveness and develop a comprehensive framework based on an infinite-horizon Markov decision process. We establish and prove the monotonic structural properties of total profit and maintenance actions relative to machine state, and develop an improved value iteration algorithm to reduce the computational space. Numerical experiments demonstrate the performance of our model in maximizing profits compared to traditional condition-based maintenance and condition-based production approaches. Sensitivity analysis further highlights key parameters influencing optimal policies, providing valuable insights for practical applications.