Joint optimization of production and condition-based maintenance scheduling for make-to-order manufacturing systems

Abstract

In make-to-order manufacturing systems, the production plan is made according to randomly fluctuating customer orders, therefore, the workload distributed to production machines fluctuates with customer orders. The workload that a production machine undertakes influences its deterioration process and maintenance. Besides, different types of dependencies among production machines also influence the production and maintenance scheduling of manufacturing systems. Existing studies have not paid enough attention to the joint optimization of production and maintenance for make-to-order manufacturing systems considering different types of dependencies. In this paper, first, customer orders are predicted using Autoregressive Integrated Moving Average (ARIMA) or Seasonal ARIMA (SARIMA) according to their characteristics. Second, the model of daily working hours is proposed to represent the influence of customer orders on the deterioration process of production machines. Third, the reliability and imperfect maintenance models of production machines considering three types of dependencies are proposed, and the optimal maintenance time window (MTW) for opportunistic maintenance is obtained in order to minimize the maintenance cost. Fourth, the mathematical model of the entire cost of a new production line construction, covering maintenance cost, operators’ salary, and construction cost, is proposed, and the optimal construction time is obtained in order to minimize the entire cost. Finally, a real use case of the manufacturing system for the shell of air-conditioning compressor is given to illustrate the feasibility and applicability of the proposed model and method. The originality of this work lies in the consideration of the influence of customer orders on the deterioration process of production machines and the modelling of different types of dependencies among production machines in a single model.