This paper investigates the issue of integrated decision-making in production, maintenance, repair, and product quality planning in a single-machine, single-product production system with a single final product inventory. A production machine wears out from the surface, and due to its operation, it is affected by breakdowns and causes them to break down. Performing partial repairs returns the production machine's wear level to its previous state, while complete repairs return it to its initial condition. Unlike previous research, in this study, the possibility of partial maintenance repairs is limited, and reaching the maximum allowable level necessitates unavoidable complete repairs. Additionally, the quality of the final product depends on the production machine's wear and tear level; thus, as the level of machine wear increases, the likelihood of producing low-quality products also increases. At the end of each day, demand enters the system following a Poisson process, and if the final product is available, the demand is met; otherwise, orders are backlogged up to a particular ceiling, and the remaining backlog is considered lost orders. The production system under investigation was modeled in the first step using an agent-based simulation approach to extract an optimal decision combination. In the second step, by employing a simulation-optimization approach, the connection between the agent-based simulation model of the production system and metaheuristic methods was established to extract an optimal policy. The goal is to find a policy that leads to the integrated optimization of the system and minimizes production costs, maintenance and repair costs, inventory holding costs, backlogged orders, machinery breakdowns, and low-quality product production. Four scenarios were designed to evaluate the proposed method. Finally, a comparison was made between the results obtained from the proposed method and production scenarios up to the time of failure and random decision-making. The results showed that the simulation-optimization approach performs up to 30% better than other policies.