Search

We analyse the joint scheduling of spare parts production and service engineers in a multi-site maintenance system. In the system, for each failure, a service engineer with the required spare part needs to be allocated for on-site maintenance. If one of the required resources is not readily available, the maintenance task can be satisfied via an external channel with quick responsiveness but high involved costs. Service providers need to make an efficient plan, including production scheduling of spare parts, allocation of service engineers and outsourcing strategy. First, a mathematical model with the objectives of minimum total outsourcing costs and minimum weighted tardiness is formulated. Then, a novel and knowledge-based heuristic algorithm is developed, namely progressive Pareto algorithm based on step-size (PPAS). To verify the performance of PPAS, it is compared against other well-known algorithms, including NSGA-II, MOEA/D, NNIA, RIPG, MMOIG and MNNIA. Finally, the effectiveness of the joint scheduling mode is demonstrated by comparing it with the separate optimisation mode of spare parts production and workforce allocation. Highlights Analyse the joint scheduling of spare parts production and service engineers Construct a bi-objective optimisation model Develop a progressive Pareto algorithm based on step-size Verify the superiority of the proposed algorithm and joint scheduling mode [ABSTRACT FROM AUTHOR]

The previous studies on service resource scheduling for maintenance, repair and operation (MRO) ignore the diverse service processes from different complex equipment and the coordination of task scheduling and resource allocation, which is less realistic when multiple tasks exist. In this work, we propose an integrated scheduling model of distributed service resources for on-site MRO, which simultaneously considers different service processes, service teams and material resources. The MRO service process is decomposed into multiple subtasks with hierarchical relationships (dependency and independency). Within the duration of the subtask, all multi-skill technicians in service team stay together and material resources are delivered from service providers. Resource allocation and task scheduling are integrated to minimise the makespan, the excessive human resources and the cost performance index of material resources. Further, we introduce six classic multi-objective evolutionary algorithms (MOEAs) to solve the scheduling problem, in which a novel and practical encoding method mixed with integers and decimals is developed. The design-of-experiment (DOE) method is used to determine the optimal combination of crucial parameters. Finally, 20 benchmark instances of multiple MRO tasks are tested. Experiment results show MOEA/D-DU could obtain higher quality scheduling solutions with relatively low computational costs in the integrated scheduling model. [ABSTRACT FROM AUTHOR]

The previous studies on scheduling problem with tool changes take processing time as the only reason for the tool wear, which is not accurate in the real manufacturing system. This paper takes processing speed and processing time into consideration simultaneously and proposes a new unrelated parallel machine scheduling problem (UPMSP) with tool changes caused by the tool wear, in which the energy consumption rate of the parallel machines is influenced by two factors: tool changes and corresponding processing speed. A new effective heuristic evolutionary algorithm (NHEA) is presented to solve the proposed UPMSP with objectives of optimising total energy consumption and makespan. For the NHEA, some effective operators such as target-searching operators are designed to accelerate the search efficiency and further exploit the solution space. A first fit decreasing algorithm is presented and incorporated into the NHEA to reduce the number of tool changes. The Taguchi method of Design of Experiments is used to obtain the best combination of key parameters of the NHEA. Extensive computational experiments are carried out to compare the NHEA with some well-known algorithms. The results validate that the proposed NHEA is able to obtain better Pareto solutions for UPMSP with tool changes. [ABSTRACT FROM AUTHOR]

The traditional flexible job shop scheduling problem (FJSP) considers machine flexibility but not worker flexibility. Given the influence and potential of human factors in improving production efficiency and decreasing the cost in practical production systems, we propose a mathematical model of an extended FJSP with worker flexibility (FJSPW). A hybrid artificial bee colony algorithm (HABCA) is presented to solve the proposed FJSPW. For the HABCA, effective encoding, decoding, crossover and mutation operators are designed, and a new effective local search method is developed to improve the speed and exploitation ability of the algorithm. The Taguchi method of Design of Experiments is used to obtain the best combination of key parameters of the HABCA. Extensive computational experiments carried out to compare the HABCA with some well-performing algorithms from the literature confirm that the proposed HABCA is more effective than these algorithms, especially on large-scale FJSPW instances. [ABSTRACT FROM AUTHOR]

Remanufacturing has been widely studied for its potential to achieve sustainable production in recent years. In the literature of remanufacturing research, process planning and scheduling are typically treated as two independent parts. However, these two parts are in fact interrelated and often interact with each other. Doing process planning without considering scheduling related factors can easily introduce contradictions or even infeasible solutions. In this work, we propose a mathematical model of integrated process planning and scheduling for remanufacturing (IPPSR), which simultaneously considers the process planning and scheduling problems. An effective hybrid multi-objective evolutionary algorithm (HMEA) is presented to solve the proposed IPPSR. For the HMEA, a multidimensional encoding operator is designed to get a high-quality initial population. A multidimensional crossover operator and a multidimensional mutation operator are also proposed to improve the convergence speed of the algorithm and fully exploit the solution space. Finally, a specific legalising method is used to 'legalise' possible infeasible solutions generated by the initialisation method and mutation operator. Extensive computational experiments carried out to compare the HMEA with some well-known algorithms confirm that the proposed HMEA is able to obtain more and better Pareto solutions for IPPSR. [ABSTRACT FROM AUTHOR]