Opportunistic maintenance policies for multi-machine production systems with quality and availability improvement
Introduction
Globalization and fierce market competition make manufacturing enterprises face great challenges in maintaining profits and competitive predominance. The machines in a production system undergo wear and degradation with age and usage. The degradation and wear can increase the risk of machine malfunction and deteriorate the product quality of the production system. Maintenance is used to mitigate the production system degradation and keep high product quality and production availability [1]. However, non-optimal maintenance activities can bring large maintenance costs and/or lead to long system downtime, resulting in low availability and large production losses. The maintenance problem is challenging as production system reliability and availability, and product quality are inherently conflicting objectives [2], which calls for a trade off by an effective maintenance policy with also lowest possible costs.
In the past few decades, the integration of two by two of the three key production system performances: reliability, product quality, and availability, has been extensively studied [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18]. Through the combination of statistical process control (SPC) and preventive maintenance (PM), some scholars have researched the combination of maintenance optimization and quality improvement from different aspects, such as control charts and acceptance sampling [9], [10], [11]. Yet, these studies do not address underlying mechanism of the impact of machine or system degradation on quality deviations. Jin and Shi [12] proposed the state-space model to establish the formation mechanism of product feature deviation in the production process. On this basis, Jin and Chen [13] proposed the QR-Co-Effect method to analyze the relationship between component reliability and product quality deviation. Then, based on the response surface model [19] and Taguchi’s loss function [20], Chen and Jin [14] further studied product quality degradation for the production system. However, these studies mainly focus on the relationship between product quality degradation and single machine system. The impact of the time required for maintenance actions that could exacerbate system unavailability and lead to further loss of profit, is often ignored.
As one of the key system performances, availability has been extensively studied in the reliability field [2], [15], [16], [17], [18], [21], [22], [23]. However, there is limited research on single-machine PM decision-making that integrates reliability, product quality degradation, and availability, simultaneously, let alone the multi-unit production system. It is clear that reliability and product quality, and availability are conflicting objectives: frequent maintenance activities can keep high reliability and product quality but would bring large maintenance costs and lead to frequent system downtime which results in low availability and thus more production losses; compromising reliability for cost efficiency would bring worse product quality. Besides, maintenance optimization integrating reliability, product quality degradation, and availability is a multi-objective optimization, whose methods of solution are mainly classified in two categories: traditional optimization methods [15], [17] and intelligent algorithms [2], [16], [21]. However, in practice for some traditional methods such as the weighted method, the determination of weights cannot be quantified, and for intelligent algorithms, it takes too much time to solve the target, which is not conducive to the practical application of maintenance policies. Therefore, a hybrid performance factor that combines traditional methods and intelligent algorithms is developed in this paper to achieve the multi-objective optimization.
Note that the studies mentioned above mainly involve single machine systems. A multi-machine production system (MMPS) is a system containing multiple machines working together to yield a final product. Such systems are quite common in the production industry, such as continuous flow manufacturing systems, ferrite phase shifting unit manufacturing systems and automatic stamping production lines [24], [25], [26]. For MMPS, an important issue that needs to be addressed is the maintenance dependence among units [27], [28]. To exploit the dependence, when maintenance activity on one machine in the MMPS leads to system downtime or partial downtime, this downtime opportunity can be used to schedule other machines for simultaneous maintenance. This strategy is called opportunistic maintenance (OM) strategy. Over the past decades, various OM strategies have been proposed for a wide variety of multi-component/unit systems with different maintenance dependence [29], [30], [31], [32], [33]. However, in these OM studies, cost rate is the only optimization objective. The OM policies simultaneously optimized by reliability, product quality degradation, and availability are rarely developed, which will lead to worse product quality and system availability decrease and, thus, considerable production losses.
Meanwhile, these OM policies can be generally divided into two categories according to the OM criterion. One category is short-term cost-saving grouping opportunity maintenance, that is, grouping all units in the system with all possible combinations and determining the optimal combinations by maximizing the cost-saving [30], [32]. The other category leads to implementing OM by defining an OM zone, which refers to the status threshold (i.e. hazard rate, age, etc.) or maintenance time window [29], [31]. In this category, when a unit in the multi-unit system is stopped for PM, other units that fall into the OM zone are maintained together. However, for the first category, when the number of units contained in the system increases, the computational complexity of the short-term cost-savings method experiences exponential growth, since the OM alternatives need to be enumerated. For the second category, there are usually only one or two common constant OM zones for all units in the system. In this way, some units may become a bottleneck to reduce system maintenance costs. Besides, although cost optimization at the system level may be achieved, at the unit level it may cause ‘over maintenance’ or ‘under maintenance’.
In this paper, a novel opportunistic PM policy with condition index (CI) and dynamic threshold functions (curves) is proposed for a serial MMPS, where PM is performed to improve the reliability, product quality, and availability of the individual machine and the MMPS, simultaneously. First, the CI and the PM threshold function are developed to evaluate the condition of machines in the MMPS and to implement PM decision-making of the individual machine. A hybrid performance factor is proposed to achieve the simultaneous optimization of the three objectives: reliability, product quality, and availability. The most effective maintenance schedule for the each individual machine is obtained by optimizing the corresponding hybrid performance factor. Then, the OM threshold function is defined to develop the OM zone (the OM criterion). The OM zone of each machine is variable and separately determined with the hybrid performance factor. Finally, based on the OM zone, the tentative PM schedule of all machines are further adjusted to obtain the optimal PM schedule of the whole MMPS.
The remainder of this paper is organized as follows. Section 2 gives the problem description and some assumptions. In Section 3, the PM model for each machine is formulated. Section 4 develops the OM zone and describes the opportunistic PM scheduling methodology. A case study and a comparison analysis are conducted in Section 5 to demonstrate the validity of the proposed model. Section 6 presents some discussions and conclusions.
Section snippets
Problem description
In this study, a repairable series MMPS is considered, which is used to produce a specific type of product within a finite planning horizon. The productivity of the MMPS remains stable (except during PM activities). During production, the machines in the MMPS are subject to continuous degradation, which not only increases the failure probability of the machines but also causes product quality degradation. PM of individual machines results in downtime of the whole series MMPS, with considerable
PM model for single-machine
In this section, the degradation of the single-machine and the product quality are analyzed to build the quantitative relationship between machines and PQCDs. Then, the definition of the CI is proposed to reflect the condition of the machines in the MMPS based on monitoring data, and the corresponding PM threshold function is developed to formulate individual machine PM decision-making. Finally, based on the availability model, the total quality loss cost, and the cost rate model, the hybrid
Opportunistic PM model for the series MMPS based on dual dynamic thresholds
In this section, the tentative PM schedule of all machines obtained by the joint optimization of cost, product quality, and availability is further adjusted to obtain the optimal opportunistic PM schedule of the whole series MMPS.
Overview
In this section, a case study about a series three machine production systems which manufactures a kind of shaft sleeves is used to validate the opportunistic PM model proposed. The functions, variables of the machines, and the PQCDs are described in Table 1.
The parameters of the quantitative relationship model mainly refer to the Ref. [29], as shown in Eqs. (37) to (40):
Conclusion
This paper proposes a novel opportunistic PM model for a MMPS, where the reliability, product quality, and availability are jointly optimized. Based on monitoring data, the CI is developed to characterize the condition of the machines in the MMPS, and the corresponding threshold functions are developed to realize maintenance decision-making for individual machines. To make up for the shortcomings of current multi-objective optimization methods and facilitate the implementation of the proposed
Funding
This work was supported by the Graduate Scientific Research and Innovation Foundation of Chongqing, China under Grant [No. CYB190 08]; National Natural Science Foundation of China under Grant [No. 71801126].
CRediT authorship contribution statement
Haohao Shi: Writing – original draft, Methodology, Investigation, Formal analysis. Ji Zhang: Methodology, Investigation. Enrico Zio: Writing – review & editing. Xufeng Zhao: Writing – review & editing, Supervision.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References (41)
- et al.
Joint modeling of preventive maintenance and quality improvement for deteriorating single-machine manufacturing systems
Comput Ind Eng
(2016) - et al.
Maintenance policy for a system with a weighted linear combination of degradation processes
European J Oper Res
(2020) - et al.
Preventive replacement policies with time of operations, mission durations, minimal repairs and maintenance triggering approaches
J Manuf Syst
(2021) - et al.
The evolution of system reliability optimization
Reliab Eng Syst Saf
(2019) - et al.
Integrated optimization of quality and maintenance: A literature review
Comput Ind Eng
(2021) - et al.
An integrated maintenance and statistical process control model for a deteriorating production process
Reliab Eng Syst Saf
(2022) - et al.
A reliability based integrated model of maintenance planning with quality control and production decision for improving operational performance
Reliab Eng Syst Saf
(2022) - et al.
Availability-based reliability-centered maintenance planning for gas transmission pipelines
Int J Press Vessels Pip
(2020) - et al.
Multi-objective availability and cost optimization by pso and coa for series-parallel systems with subsystems failure dependencies
Microprocess Microsyst
(2022) - et al.
A methodology for equipment condition simulation and maintenance threshold optimization oriented to the influence of multiple events
Reliab Eng Syst Saf
(2023)
Multi-objective optimization in redundant system considering load sharing
Reliab Eng Syst Saf
Reliability and availability assessment and enhancement of water-cooled multi-chiller cooling systems for data centers
Reliab Eng Syst Saf
Evaluation of the availability and reliability of a standby repairable system incorporating imperfect switchovers and working breakdowns
Reliab Eng Syst Saf
Integrated predictive maintenance approach for multistate manufacturing system considering geometric and non-geometric defects of products
Reliab Eng Syst Saf
A review on maintenance optimization
European J Oper Res
Generalized condition-based maintenance optimization for multi-component systems considering stochastic dependency and imperfect maintenance
Reliab Eng Syst Saf
Opportunistic preventive maintenance scheduling for serialparallel multistage manufacturing systems with multiple streams of deterioration
Reliab Eng Syst Saf
Group maintenance optimization of subsea Xmas trees with stochastic dependency
Reliab Eng Syst Saf
Maintenance gravity window based opportunistic maintenance scheduling for multi-unit serial systems with stochastic production waits
Reliab Eng Syst Saf
Multi-level opportunistic predictive maintenance for multi-component systems with economic dependence and assembly/disassembly impacts
Reliab Eng Syst Saf
Cited by (11)
Joint optimization of maintenance and spares inventory policy for a series-parallel system considering dependent failure processes
2024, Reliability Engineering and System SafetyIntegrated mission reliability modeling for multistate manufacturing systems considering heterogeneous feedstocks based on extended stochastic flow manufacturing network
2024, Reliability Engineering and System SafetyMaintenance modeling for hot rolling production lines with constraint of auxiliary resources
2024, Computers and Industrial EngineeringReliability modeling and opportunistic maintenance optimization for a multicomponent system with structural dependence
2024, Reliability Engineering and System SafetyMission reliability-centered opportunistic maintenance approach for multistate manufacturing systems
2024, Reliability Engineering and System SafetyA hybrid inspection-replacement policy for multi-stage degradation considering imperfect inspection with variable probabilities
2024, Reliability Engineering and System Safety