Elsevier

Journal of Manufacturing Systems

Volume 61, October 2021, Pages 423-449
Journal of Manufacturing Systems

Real-time integrated production-scheduling and maintenance-planning in a flexible job shop with machine deterioration and condition-based maintenance

https://doi.org/10.1016/j.jmsy.2021.09.018Get rights and content

Highlights

  • A real-time optimization-based approach is proposed to jointly optimize production schedules and maintenance plans.

  • Integrated plans created considering machines’ conditions are more effective.

  • Baseline plans with high quality are critical to the quality of the realized plans.

  • Increasing limits on rescheduling times will increase cost-saving.

  • Investing in Industry 4.0 concepts to gather real-time information for decision-making is beneficial.

Abstract

The introduction of modern technologies in manufacturing is contributing to the emergence of smart (and data-driven) manufacturing systems, known as Industry 4.0. The benefits of adopting such technologies can be fully utilized by presenting optimization models in every step of the decision-making process. This includes the optimization of maintenance plans and production schedules, which are two essential aspects of any manufacturing process. In this paper, we consider the real-time joint optimization of maintenance planning and production scheduling in smart manufacturing systems. We have considered a flexible job shop production layout and addressed several issues that usually take place in practice. The addressed issues are: new job arrivals, unexpected due date changes, machine degradation, random breakdowns, minimal repairs, and condition-based maintenance (CBM). We have proposed a real-time optimization-based system that utilizes a modified hybrid genetic algorithm, an integrated proactive-reactive optimization model, and hybrid rescheduling policies. A set of modified benchmark problems is used to test the proposed system by comparing its performance to several other optimization algorithms and methods used in practice. The results show the superiority of the proposed system for solving the problem under study. The results also emphasize the importance of the quality of the generated baseline plans (i.e., initial integrated plans), the use of hybrid rescheduling policies, and the importance of rescheduling times (i.e., reaction times) for cost savings.

Introduction

In the context of production decisions, the term “real-time” refers to the system’s instantaneous reaction to disruptions and unforeseen events without interrupting the production processes [1]. A reaction can include assigning an operation to a different machine, delaying or accelerating an operation, rerouting a job, or shifting the starting time of a maintenance activity [[1], [2], [3], [4]]. An event can be an order cancellation, a due date change, the arrival of a new order, a random machine breakdown, or a shortage of material [[1], [2], [3], [4]]. Real-time optimization models are in high demand in practice, as manufacturing systems are highly dynamic and subject to unexpected events and disruptions. Real-time optimization models can help manufacturing systems react and adapt to such events and requirements by continuously updating disrupted plans/schedules using real-time information about the status of the machines and operations on their shop floors. The availability of this real-time data is achieved by adopting innovative manufacturing technologies, such as Industry 4.0 [[5], [6], [7], [8]]. With Industry 4.0 concepts, such as the internet of things (IoT), cyber-physical systems (CPSs), detailed information about machines, orders (or jobs), operations, equipment, work-in-progress (WIP), inventory, and shipments can be collected in real-time. These concepts can also be integrated into the company’s enterprise resource planning/material requirements planning/computerized maintenance management (ERP/MRP/CMM) system, making all the information mentioned earlier instantly available for decision making. This level of real-time connectivity will empower Industry 4.0-based manufacturing systems to utilize real-time updates to manage, monitor, and control their operations. This includes production scheduling and maintenance planning, which are two of the core components of any manufacturing process. For more details about how such technologies (i.e., Industry 4.0 concepts, such as IoT and CPS) can be utilized for shop floor real-time data collection, processing, and analysis, readers are referred to [9]. In this paper, we intend to focus on utilizing real-time data to jointly optimize production schedules and maintenance plans in a dynamic and stochastic flexible job shop production environment. The flexible job shop scheduling problem (FJSSP) is a classical combinational optimization problem in real-world manufacturing systems. FJSSP is a generalization of the job shop and parallel machine scheduling problems [10]. In FJSSP, each job has a predetermined number of operations to be processed by more than one machine (i.e., alternative routings are allowed). As a result, FJSSP should deal with two problems: the machine assignment problem (which determines the best machine for each operation) and the sequencing problem (which searches for the best sequences for processing the jobs on each machine). FJSSP seeks to find an optimal assignment and sequence by optimizing a single (e.g., makespan [11]) or multi-objective (makespan and maintenance costs [12]) function. FJSSP is proved to be strongly NP-hard due to its complexity [13,14].

Although several studies have been dedicated to flexible job shop scheduling, most of these studies assumed that the production environment is static (i.e., all required data is available at time zero and is not subject to change with time) and machines are always available for production (i.e., never break or goes under planned maintenance). However, production environments are highly dynamic in practice, and changes such as new job arrivals [[15], [16], [17]], variable processing times [18], and due date changes [19] are expected during production. Also, machines are certainly subject to aging, deterioration, and some unavailability intervals due to scheduled preventive maintenance (PM) activities [12,20,21] as well as random machine breakdowns [15,22,23], which lead to corrective maintenance (CM) activities. PM activities can be broadly classified into two types: time-based maintenance (TBM) or condition-based maintenance (CBM). TBM tasks are performed in predetermined periods and are conducted to delay the deterioration (of machines or equipment) leading to failures and reduced production efficiency [20]. TBM assumes that a machine’s (or equipment’s) failure behavior under normal use can be estimated using historical failure data [24]. CBM tasks are performed based on information that is collected using condition monitoring. CBM actions are initiated only when there is evidence of unusual behaviors of machinery. CBM is supposed to avoid unnecessary maintenance tasks and therefore reduces maintenance costs [25]. CBM assumes that the existing indicative prognostic parameters can be detected, and the possibility of machine’s (or equipment’s) failure can be determined before its actual occurrence. Prognostic parameters indicate prospective issues and early defects that cause a machine (or a component) to depart from its expected performance level. Aging and deterioration of machines (or equipment) are common maintenance issues, among the main reasons for reduced production efficiency [26,27]. A trend analysis of the machine (or equipment) condition data can also reveal the degradation tendency of crucial components [28]. If all these issues and disruptions (e.g., aging, deterioration, random breakdowns, maintenance activities, new job arrivals, and due date changes) are neglected during production planning, they will result in production losses as well as a reduction in production efficiency and machines (or equipment) service life. Hence, integrating maintenance planning (by addressing aging, deterioration, and different maintenance activities) into flexible job shop scheduling with real-time dynamics will reduce maintenance and production costs and keep manufacturing systems competent, reliable, competitive, and efficient.

Although several researchers have addressed the issue of integrating maintenance planning and production scheduling, their studies did not address real-time dynamics and the proposed models considered TBM activities (e.g., [12,21,29,30]), TBM activities with fixed time intervals (e.g., [31]), different TBM activities in diverse intervals (e.g., [32]), opportunistic maintenance (e.g., [33]), or noncyclical TBM activities (e.g., [34]). Even for studies that addressed machine deterioration and CBM in flexible job shops, simulation-based optimization models (e.g., [20,35]), decomposition methods (e.g., [27,36]), and machine-learning-based health indicator models (e.g., [37]) were proposed, no closed-form formulation was developed before for machines reliability (including deterioration, random breakdown, minimal repairs, and CBM) and its impacts on machine condition, processing times, and maintenance costs. As a result, this work aims to solve the joint optimization problem of production schedules and CBM plans in a flexible job shop with real-time dynamics (which can be denoted as RTFJm). The issues and disruptions addressed in this paper are new job arrivals, due date changes, machines deterioration, random machine breakdowns, minimal repairs, and CBM actions. This new problem is solved by minimizing the expected total system cost including the expected total tardiness cost and the expected total maintenance costs. Due to the complexity of this new problem, this paper proposes the design and application of a metaheuristic-based approach to solve the problem. The genetic algorithm (GA) has been successfully applied to tackle combinatorial optimization problems [38]. Compared with the other metaheuristics, GA has more flexibility and does not need to embed knowledge [38]. Besides, GA has been successfully applied to tackle several production scheduling problems including flexible job shop scheduling [11,15,39]. Hence, we apply and adapt a modified version of a hybrid GA (denoted hereafter as mHGA) for the problem under investigation (more details about the proposed mHGA are provided in Section 4.4). Apart from determining operations assignments and sequences, this new problem needs to determine optimal CBM plans, and update created plans and schedules when needed (i.e., upon new job arrivals, due date changes, machines breakdowns, or CBM) using real-time updates. Thus, and to put it succinctly, the contribution of this paper can be summarized in the following two points:

  • The formulation of the new problem (i.e., the joint optimization of production scheduling and maintenance planning in a flexible job shop with new job arrivals, due date changes, machines deterioration, random machine breakdowns, minimal repairs, and CBM interventions) – (see Section 3 and Appendix A). To do that, we first modeled the reliability of machines as a multi-state system using continuous-time Markov chains. Then, a matrix-based approach is introduced to formulate the effects of machines’ deterioration, random breakdowns, and maintenance activities (both minimal repairs and CBM) on the processing times and the associated maintenance and production costs. Finally, the developed formulations are integrated into a mixed-integer programming model to jointly optimize production schedules and CBM plans by minimizing the expected total system cost, including the expected total cost of tardiness and the expected total maintenance cost.

  • The development of a proactive-reactive scheduling approach based on a modified hybrid GA (see Section 4). A hybrid rescheduling policy is included (see Section 4.2 for more details). A problem-specific solution representation, repair, and evaluation are embedded into the proposed mHGA. Finally, a new mechanism for generating updated solutions (i.e., production schedules and maintenance plans) is included in the proposed metaheuristic (see Section 4.4 for more details). The proposed approach can be utilized as a real-time optimization-based system to generate integrated plans and reacts to several production disruptions, including new job arrivals, random breakdowns, due date changes, minimal repairs, and CBM activities.

Several computational experiments are performed to test the performance of the developed model and algorithm, and also to investigate the following research questions:

  • RQ1: As frequent updates are to be conducted, is the quality of the initial plans still essential for the quality of the realized (i.e., final) plans?

  • RQ2: What is the best rescheduling policy (i.e., course of action) to deal with each of the newly addressed events (due date changes, minimal repairs, and CBM)?

  • RQ3: As decisions are required in real-time, what is the effect of reaction (or rescheduling) times on the efficiency of the generated plans?

  • RQ4: Despite the general belief that decision-making with real-time information will result in better decisions, to what extent can the availability of real-time information enhance the integrated decisions (in terms of total expected costs)?

The test instances are generated based on well-known benchmarks by including the new parameters of our problem (more details are provided in Section 5.1). The test instances are grouped into three sets: a calibration set (20 calibration instances) to tune the algorithm and find the best parameter setting, a pre-evaluation set (400 test instances) to test the performance of the proposed mHGA, and a scenario-based set (1800 test instances) to test the performance of the proposed mHGA -based proactive-reactive scheduling approach and to answer the considered research questions. The results of the conducted computational study indicate that the proposed mHGA outperforms a set of three heuristics (LFOH, LAPH, and LFO-LAP, which are developed by [15]) and two well-known metaheuristics (a hybrid GA proposed by [40] and a teaching-learning-based optimization (TLBO) proposed by [41]) in all instances. The results also show that the proposed mGHA -based proactive-reactive scheduling approach outperforms the commonly used approach in practice (i.e., the right-shift policy), and allows for about 27 % of cost-saving on average. Furthermore, additional cost-saving of about 30 % is achieved when the limits for rescheduling times are increased from 30 to 90 s. In addition, the results suggest several policies (i.e., courses of action) for each the newly addressed issues in this investigation (i.e., due date changes, minimal repairs, and CBM) based on the conducted experiments and the obtained results for different scenarios and strategies.

This paper is structured as follows: a summary of the related literature is presented in Section 2. In Section 3, the new problem is described and formulated. The proposed mHGA -based proactive-reactive scheduling approach is presented in Section 4. Details of the conducted computational study are addressed in Section 5. Finally, conclusions and directions for future research are presented in Section 6.

Section snippets

Related work

As the focus of this paper is on the joint optimization of production scheduling and maintenance planning in a flexible job shop with real-time dynamics, including new job arrivals, due date changes, machine deterioration, random machine breakdowns, minimal repairs, and CBM, the first sub-section reviews the current state-of-the-art on production scheduling with real-time dynamics (i.e., dynamic scheduling, rescheduling, or real-time scheduling) with a slight focus on flexible job shop

Problem description and formulation

In this section, we discuss the requirements, characteristics, and assumptions of the considered problem. The integrated optimization of production scheduling and maintenance planning in a flexible job shop production system using real-time information is addressed in this paper.

Flexible job shops are generalized versions of job shops such that for each operation; there is a set of available candidate machines to process it [10]. Therefore, compared to classical job shops, there are two

Proposed integrated real-time optimization-based system

The proposed integrated real-time optimization-based system is presented in Fig. 4. This system works in both modes (offline and online) and utilizes an integrated planning (i.e., maintenance-planning and production scheduling) module. The offline mode is triggered at time zero to generate the integrated baseline plans using the integrated maintenance-planning and production scheduling module. This mode requires initial information about the machines and the jobs (or orders), which can be

Computational study

In this section, details about the computational study are provided. All the developed codes are written in MATLAB and run on a PC with an Intel® Core™ i5 CPU processor at 3.10 GHz and 10 GB RAM.

Discussion

Considering the importance of production scheduling, maintenance panning, and real-time dynamics, in this paper, we addressed these aspects in a well-known complicated production environment, the flexible job shop. In addition, several issues that could be of high interest in real-world settings were considered, which are new job arrivals, due date changes, random machine breakdowns, machine deterioration, minimal repairs, and CBM. An integrated mixed integer programming model was proposed to

Conclusions and future work

This paper investigates the integration of production scheduling and maintenance planning in the context of Industry 4.0 (i.e., under real-time settings). The considered production layout is a flexible job shop in which machines are not identical and are subject to deterioration, random breakdowns, minimal repairs, and condition-based maintenance. New formulations are introduced to model each machine's condition and the effects on the processing times and the associated maintenance and

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.

Acknowledgments

The funding for this research was provided by the Canada Research Chairs (CRC) program. This work is also supported by the Natural Sciences and Engineering Research Council (NSERC) of Canada through the Discovery Grant (#RGPIN-2017-04434).

References (80)

  • M. Nouiri et al.

    Two stage particle swarm optimization to solve the flexible job shop predictive scheduling problem considering possible machine breakdowns

    Comput Ind Eng

    (2017)
  • J. Xiong et al.

    Robust scheduling for multi-objective flexible job-shop problems with random machine breakdowns

    Int J Prod Econ

    (2013)
  • A.K.S. Jardine et al.

    A review on machinery diagnostics and prognostics implementing condition-based maintenance

    Mech Syst Signal Process

    (2006)
  • P. Rokhforoz et al.

    Distributed joint dynamic maintenance and production scheduling in manufacturing systems: framework based on model predictive control and Benders decomposition

    J Manuf Syst

    (2021)
  • F. El Khoukhi et al.

    The “Dual-Ants Colony”: a novel hybrid approach for the flexible job shop scheduling problem with preventive maintenance

    Comput Ind Eng

    (2017)
  • E. Moradi et al.

    Bi-objective optimization research on integrated fixed time interval preventive maintenance and production for scheduling flexible job-shop problem

    Expert Syst Appl

    (2011)
  • X. Chen et al.

    An approximate nondominated sorting genetic algorithm to integrate optimization of production scheduling and accurate maintenance based on reliability intervals

    J Manuf Syst

    (2020)
  • T. Wu et al.

    Proactive maintenance scheduling in consideration of imperfect repairs and production wait time

    J Manuf Syst

    (2019)
  • M. Alimian et al.

    New integration of preventive maintenance and production planning with cell formation and group scheduling for dynamic cellular manufacturing systems

    J Manuf Syst

    (2020)
  • J. Hu et al.

    Joint optimization of job scheduling and maintenance planning for a two-machine flow shop considering job-dependent operating condition

    J Manuf Syst

    (2020)
  • F. Pezzella et al.

    A genetic algorithm for the flexible job-shop scheduling problem

    Comput Oper Res

    (2008)
  • N. Al-Hinai et al.

    Robust and stable flexible job shop scheduling with random machine breakdowns using a hybrid genetic algorithm

    Int J Prod Econ

    (2011)
  • Y.R. Shiue et al.

    Real-time scheduling for a smart factory using a reinforcement learning approach

    Comput Ind Eng

    (2018)
  • Z. Hammami et al.

    On-line self-adaptive framework for tailoring a neural-agent learning model addressing dynamic real-time scheduling problems

    J Manuf Syst

    (2017)
  • P. Fattahi et al.

    Dynamic scheduling in flexible job shop systems by considering simultaneously efficiency and stability

    CIRP J Manuf Sci Technol

    (2010)
  • D. Rahmani et al.

    A stable reactive approach in dynamic flexible flow shop scheduling with unexpected disruptions: a case study

    Comput Ind Eng

    (2016)
  • A. Bousdekis et al.

    Decision making in predictive maintenance: literature review and research agenda for industry 4.0

    IFAC-PapersOnLine

    (2019)
  • M.C. Garcia et al.

    SIMAP: Intelligent System for Predictive Maintenance: application to the health condition monitoring of a windturbine gearbox

    Comput Ind

    (2006)
  • B. Einabadi et al.

    Dynamic Predictive Maintenance in industry 4.0 based on real time information: case study in automotive industries

    IFAC-PapersOnLine

    (2019)
  • V.M. Dalfard et al.

    Two meta-heuristic algorithms for solving multi-objective flexible job-shop scheduling with parallel machine and maintenance constraints

    Comput Math Appl

    (2012)
  • M. Ghaleb et al.

    Integrated production and maintenance scheduling for a single degrading machine with deterioration-based failures

    Comput Ind Eng

    (2020)
  • E. Ahmadi et al.

    A multi objective optimization approach for flexible job shop scheduling problem under random machine breakdown by evolutionary algorithms

    Comput Oper Res

    (2016)
  • X. Qi et al.

    Disruption management for machine scheduling: the case of SPT schedules

    Int J Prod Econ

    (2006)
  • M. Nouiri et al.

    Towards energy efficient scheduling and rescheduling for dynamic flexible job shop problem

    IFAC-PapersOnLine

    (2018)
  • J. Shahrabi et al.

    A reinforcement learning approach to parameter estimation in dynamic job shop scheduling

    Comput Ind Eng

    (2017)
  • S.Q. Liu et al.

    Metaheuristics for minimizing the makespan of the dynamic shop scheduling problem

    Adv Eng Softw

    (2005)
  • M.C. Fitouhi et al.

    Performance evaluation of a two-machine line with a finite buffer and condition-based maintenance

    Reliab Eng Syst Saf

    (2017)
  • M. Sharifi et al.

    Optimal production and maintenance scheduling for a degrading multi-failure modes single-machine production environment

    Appl Soft Comput

    (2021)
  • C.M. Harmonosky et al.

    Real-time scheduling in computer integrated manufacturing: a review of recent research

    Int J Comput Integr Manuf

    (1991)
  • G.E. Vieira et al.

    Rescheduling manufacturing systems: a framework of strategies, policies, and methods

    J Sched

    (2003)
  • Cited by (49)

    View all citing articles on Scopus
    View full text