Cost-oriented predictive maintenance based on mission reliability state for cyber manufacturing systems

Yihai He,Changchao Gu,Xiao Han,Zhaoxiang Chen

doi:10.1177/1687814017751467

Abstract

With the advent of Industry 4.0, maintenance strategy faces new demands to avoid the hysteresis of the conventional passive maintenance mode and the non-feasibility of the periodic preventive maintenance model. In view of the inherent polymorphism of manufacturing systems and with the objective of maximizing benefits, a novel cost-oriented predictive maintenance based on mission reliability state for manufacturing systems is proposed. First, the cyber-physical system is adopted to organize and analyze big data in the operational process of manufacturing systems in terms of predictive analytics in cyber manufacturing environment. Second, a new connotation of mission reliability is defined based on the big operational data to comprehensively characterize the dynamic state of the equipment health states and the qualified degree of the production task. Third, the predictive maintenance mode based on mission reliability state is quantified by the comprehensive cost, and the relationship between mission reliability and cost is established. Thereafter, cost-oriented dynamic predictive maintenance strategy is proposed. Finally, a case study on the maintenance decision-making problem of a cylinder head manufacturing system is presented. The final result shows that the comprehensive cost can be further reduced by the proposed method relative to the traditional periodic preventive maintenance strategy.

Highlights

With the rapid development of information technology and computing science and the integration of advanced analytics into cyber manufacturing today, many production activities are facing new opportunities and challenges.[1,2,3] Proper maintenance strategies are drawing increasing attention and facing new challenges as important methods in improving the reliability and availability of manufacturing systems in order to ensure the timely delivery of high-quality products to customers.[4,5] In many cases, maintenance costs can reach 15%–70% of the total production cost or even exceed the annual net profit.[6]
A novel approach for developing a dynamic predictive maintenance strategy based on the mission reliability of multi-state manufacturing systems has been presented in the context of cyber manufacturing
In accurately characterizing the polymorphism of manufacturing systems, mission reliability is put forward based on the equipment degradation state and production task demands

Summary

Introduction

With the rapid development of information technology and computing science and the integration of advanced analytics into cyber manufacturing today, many production activities are facing new opportunities and challenges.[1,2,3] Proper maintenance strategies are drawing increasing attention and facing new challenges as important methods in improving the reliability and availability of manufacturing systems in order to ensure the timely delivery of high-quality products to customers.[4,5] In many cases, maintenance costs can reach 15%–70% of the total production cost or even exceed the annual net profit.[6]. A major improvement is provided by making up for the traditional method’s deficiencies especially from the big data collection process and the transformation of big data into meaningful information through the cyber-physical system (CPS) development On this basis, a predictive maintenance decision-making method was proposed in this article based on mission reliability state for cyber manufacturing systems. The strategies of production operation and maintenance should be developed and optimized by analyzing massive data and by combining them with an effective model These data include the inherent information, such as equipment reliability, equipment capacity, and system structure, as well as a number of dynamic change information, such as the dynamic production task demands, equipment failure rates, and manufacturing pass rates. Ð14Þ where E is the number of predictive maintenance cycles in planning horizon T. e represents the residual time from the last predictive end of planning horizon mT.aiÐn0etlenE a+n1cdet activity until characterizes the the cumulative failure number during residual time e

Predictive maintenance cost

Production capacity loss cost

Indirect loss cost

Product quality loss cost

Background

Conclusion