Full length Article
Multi-phase integrated scheduling of hybrid tasks in cloud manufacturing environment

https://doi.org/10.1016/j.rcim.2019.101850Get rights and content

Highlights

  • Three problems including order assignment, supplier selection, and production line scheduling are jointly considered for hybrid manufacturing processes.

  • The task queuing time, resource idle time, and in-production transportation time are modeled to assess the overall time and cost for each production order.

  • Six multi-objective evolutionary algorithms are applied and compared to determine the one best suited for integrated problem.

  • Experimental results show that integrated scheduling is more suitable for hybrid task compared than traditional step-by-step decision.

Abstract

Cloud manufacturing paradigm aims at gathering distributed manufacturing resources and enterprises to serve for more customized production. Production order which involving several tasks can be taken by distributed suppliers collaboratively at lower cost. The cloud manufacturing platform is responsible for not only arranging reasonable priorities, suitable suppliers, and production processes to multiple orders, but also scheduling hybrid tasks from different orders to manufacturing resources. To maximize the production efficiency and balance the trade-off among different production orders, this paper studies multi-phase integrated scheduling of hybrid tasks in cloud manufacturing environment, which containing order priority assignment, supplier and production process selection, and production line scheduling. Five key objectives are taken into account to analyze the interconnections among different resources and production processes. Six representative multi-objective evolutionary algorithms are adopted to solve the integrated scheduling problem. Experimental results on six production cases show that integrated scheduling is more effective than the traditional step-by-step decision, leading to less production cost and time. In addition, a comparison among the six algorithms is carried out to determine the one best suited for the integrated scheduling problem in different circumstances.

Introduction

Cloud Manufacturing (CMfg) is a paradigm that uses Internet-of-Things and cloud computing technology to integrate distributed manufacturing resources as services [1], [2]. Customers can post their requirement and find suitable services for manufacturing products they required without purchasing additional equipment. It has made customized production easier and state-of-the-art machining technologies accessible for particularly middle and small-sized enterprise [3].

Since many manufacturing resources are organized in a production line with complex control programs, suppliers only provide simplified interface information of these resources [4]. Without an effective communication and pre-established trust between customer and suppliers, a production order, which contains a series of processing tasks, is more likely to be executed with low-quality or be delayed in delivery [5]. This phenomenon becomes more serious when hybrid production tasks from multiple orders are submitted to the platform. Therefore, CMfg platform is in charge of finding suitable schedule to execute these tasks in multiple production lines [6].

The above process is normally divided into three phases, demand negotiation, supplier selection, and resource allocation [7], [8], [9], [10], [11], [12], to assign the priority of different customer orders, find suitable suppliers, and allocate manufacturing resources with proper size to carry out the tasks. It is assumed that production orders in CMfg platform are handled one by one. The production process for each order is pre-determined without change. Each task of an order can find independent resources without preemption. If a production task is assigned to a group of resources, it does not need to queue for execution. But in fact, this is not the case.

Enterprise conducts their own resource pool and shares it to a public cloud platform. The resources can be either managed independently, or organized in a product line. Thus, a CMfg platform is established as a multi-centered resource pool in which each center is maintained by an enterprise [13]. Customer has a lot of options on the production process of a customized product. In this situation, a production order is firstly decomposed into multiple tasks. These tasks are allocated to a group of manufacturing suppliers. Resource preemption inside an enterprise and load imbalance among multiple centers become two common phenomena that hinder the efficiency and quality of distributed production when limited number of resources are shared for multiple orders [14].

Though existing researches of supplier selection, service composition and production line scheduling are carried out considering the characteristics of CMfg, there is still large cooperation barriers to be reduced between customer and enterprise. In this case, recent researchers put more emphasis on the hierarchical structure of CMfg. Some of them built hierarchical resource allocation models in CMfg environment with individual users, cloud manager, and enterprise providers [15], [16]. They demonstrated the interconnections among users, cloud manager, and enterprise providers in general production process. Researchers summarized the characteristic of different production and categorized them into order allocation, supplier selection, service composition, production line scheduling and so on [17], [18], [19], [20]. However, the interactions between different production phases and the resource constraints in distributed production lines have not been considered in the above research. If the CMfg platform selects the best suited supplier and process for each task, it can cause serious resources preemption and even production delay in the underlying production lines. On the contrary, better production line scheduling strategies would be hard to satisfy high production quality for each order simultaneously.

Motivated by the above discussion, a three-phase integrated scheduling model for CMfg is established in this paper. The three phases combine order priority assignment, supplier and process selection, and resource scheduling. In this way, the whole execution scheme for multiple production orders is optimized so as to maximize the benefit of manufacturing enterprises and increase customer satisfaction. After that, six typical multi-objective evolutionary algorithms (MOEAs) are used to solve the model in different phases. Through numerical analysis on the CMfg instances, one can observe that the integration of three-phase optimization for hybrid manufacturing orders is feasible and productive. The result and discussion drawn from this paper also show the performance of the six MOEAs for different scheduling requirements.

Section snippets

Related work

Since CMfg was introduced in 2010, more and more researchers and engineers contribute to the establishment of CMfg platform for manufacturing resource sharing. Xu et al. [21] outlined that CMfg mainly consists of three layers, user domain, enterprise domain, and provider domain. Based on these three layers, Liu et al. [22] summarized the life-cycle of CMfg service as resource encapsulation, resource perception and connection, resource provisioning and task execution.

To optimize the resource

The multi-phase integrated scheduling model

In this section, an integrated scheduling model for determining priorities, suppliers, processes, and production schedules of hybrid customer orders is established. The optimization process is originally divided into three decision phases, order priority assignment, supplier and production process selection, and production line scheduling. The definition and preliminaries of this integrated problem are provided. The interrelationship between every two phases is analyzed mathematically. The

Multi-objective evolutionary algorithm for multi-phase integrated scheduling

Since the inter-relationship among the three scheduling phases are embodied in different objectives, the multi-phase integrated scheduling model is a multi-objective optimization problem. In this study, multi-objective evolutionary algorithms (MOEAs) are introduced to solving the integrated problem. MOEA is developed to balance the trade-off among different objectives and find non-dominated solutions through various solution selection strategies. Three key problem-specific issues for solving

Experimental results and discussions

Comparative experiments were conducted on our integrated scheduling model and the typical scheduling model. In a typical scheduling model, different optimization phases are determined separately. The upper phases make optimization decisions for the overall planning based on the average situation of the underlying production procedure, rather than accurate scheduling. By comparing the scheduling of the integrated and separated models, we would like to verify the practicality of the integrated

Conclusion

Considering the cooperative barriers between enterprises and customer requirements in the CMfg environment, this paper studied a multi-phase integrated scheduling problem that combines order priority assignment, supplier and production process selection, and production line scheduling. Multiple production indicators of the manufacturing process were evaluated and optimized to measure the scheduling scheme more comprehensively. To better address the complex hybrid scheduling problems, six of

Declaration of Competing Interest

None.

Acknowledgments

This paper was supported by National Key R&D Program of China (No. 2018YFB1003703) and partially supported by the Natural Science Foundation of China (Grant no. 61703015) .

References (54)

  • N.K. Shah et al.

    Integrated production planning and scheduling optimization of multisite, multiproduct process industry

    Comput. Chem. Eng.

    (2012)
  • X. Shao et al.

    Integration of process planning and scheduling—a modified genetic algorithm-based approach

    Comput. Oper. Res.

    (2009)
  • J. Zhou et al.

    An adaptive multi-population differential artificial bee colony algorithm for many-objective service composition in cloud manufacturing

    Inf. Sci.

    (2018)
  • F. Chen et al.

    A flexible QoS-aware web service composition method by multi-objective optimization in cloud manufacturing

    Comput. Ind. Eng.

    (2016)
  • S. Türk et al.

    Multi-objective optimisation in inventory planning with supplier selection

    Expert Syst. Appl.

    (2017)
  • K. Deb

    An efficient constraint handling method for genetic algorithms

    Comput. Methods Appl. Mech. Eng.

    (2000)
  • R. Kolisch et al.

    PSPLIB – a project scheduling problem library

    Eur. J. Oper. Res.

    (1997)
  • B. Li

    Cloud manufacturing: a new service-oriented networked manufacturing model

    Comput. Integr. Manuf. Syst.

    (2010)
  • F. Tao et al.

    Cloud manufacturing: a computing and service-oriented manufacturing model

    Proc. Instit. Mech. Eng. Part B

    (2011)
  • B. Huang et al.

    Cloud manufacturing service platform for small- and medium-sized enterprises

    Int. J. Adv. Manuf. Technol.

    (2012)
  • F. Tao et al.

    IoT-Based intelligent perception and access of manufacturing resource toward cloud manufacturing

    IEEE Trans. Ind. Inform.

    (2014)
  • H. Akbaripour et al.

    Cloud manufacturing service selection optimization and scheduling with transportation considerations: mixed-integer programming models

    Int. J. Adv. Manuf. Technol.

    (2017)
  • C. Hu et al.

    Study of classification and modeling of virtual resources in cloud manufacturing

    Appl. Mech. Mater.

    (2011)
  • D.Y. Kim et al.

    Supplier selection problem revisited from the perspective of product configuration

    Int. J. Prod. Res.

    (2012)
  • H. Mezni et al.

    A negotiation-based service selection approach using swarm intelligence and kernel density estimation

    Software

    (2018)
  • D. Tai et al.

    Cloud manufacturing based on cooperative concept of SDN

    Adv. Mater. Res.

    (2012)
  • S. Wang et al.

    Resource allocation model in cloud manufacturing

    Proc. Inst. Mech. Eng. Part C

    (2015)
  • Cited by (55)

    • Multi-search-routes-based methods for minimizing makespan of homogeneous and heterogeneous resources in Cloud computing

      2023, Future Generation Computer Systems
      Citation Excerpt :

      Xu et al. [41] applied NSGA-III to optimize the execution time and energy consumption of IoT-enabled Cloud-edge computing. MOGA proposed by Jiang et al. [42] and MOEAs proposed by Cong et al. [20] improved the search route strategies based on NSGA-II and were utilized to settle resource scheduling in Cloud. FireFly including FA [43] and FIMPSOA [44], PSO including MOPSO [45], APDPSO [46], and TSPSO [47], are other meta-heuristic algorithms applied in scheduling of Cloud.

    View all citing articles on Scopus
    View full text