Manufacturing Scheduling Problems Research Articles

Combinatorial Benders decomposition for single machine scheduling in additive manufacturing with two-dimensional packing constraints

This paper addresses a single-machine scheduling problem in additive manufacturing (AM). We focus on the direct metal laser sintering (DMLS) technology of an AM process, where parts can be produced simultaneously in a batch. In this problem, we consider simultaneously the assignment of parts to batches and the packing of parts onto two-dimensional surfaces of the build chamber to minimize the makespan objective. To solve this problem, a mixed integer linear programming model is formulated, and an approximation algorithm is proposed, followed by an exact algorithm based on combinatorial Benders decomposition method with various Benders cuts and acceleration strategies (Algorithm CBD). Finally, we conduct a comprehensive computational study to verify the efficiency of our proposed Algorithm CBD. The computational results show that our proposed Algorithm CBD can effectively solve the test instances with different sizes.

An ensemble of brain storm optimization and Q-learning methods for distributed flexible job shop scheduling problems with distribution operations

Distributed manufacturing scheduling problems have attracted much concern from both industrial and academic areas. Nevertheless, distributed scheduling problems with distribution operations are seldom studied. This work proposes a distributed flexible job shop scheduling problem with distribution operations. A set of jobs is handled at distributed flexible job shops, and then the finished jobs are transported to their corresponding customers following given due dates. First, a mixed integer programming model is established to minimize total tardiness. Second, an ensemble of brain storm optimization and Q-learning methods is developed to solve the formulated model. Six heuristics are hybridized to generate a high-quality initial population. A Q-learning method is devised by fully employing found search information to guide subsequent search processes instead of using fixed parameters as basic brain storm optimization. A variable neighborhood search method combining problem-specific knowledge is designed to further refine the found best individual. At last, the formulated model and method are compared with three state-of-the-art metaheuristics and a mathematical programming solver CPLEX via using a group of problem instances. The results and analysis demonstrate that the developed model and algorithm have more powerful competitiveness in addressing the studied problem.

Long-sighted dispatching rules for manufacturing scheduling problem in Industry 4.0 hybrid approach

The introduction of Cyber–Physical Systems, the implementation of machine learning, and the proposal of other technologies are some of the main features of Industry 4.0, which aims to change the way of production by introducing more flexibility and more autonomous decision-making by machines. The presence of centralized and decentralized systems and the continuous change in the market scenario has led to a transition of production systems from highly centralized or decentralized processes to a hybrid architecture that incorporated advantages of the production control approaches. The analysis of the production performances of recent semi-heterarchical architectures encourage to assess these architectures by introducing new scheduling approach that overcome the risk to take a myopic perspective: the long-sighted approach. To this aim, two new dispatching rules are proposed, both evaluated with different variability. This is significant because it allows the measurement of the performance of the production system, as well as the definition of the reactive and proactive capacity of semi-heterarchical architectures in an increasingly competitive market.

Logistics-involved task scheduling in cloud manufacturing with offline deep reinforcement learning

As an application of industrial information integration engineering (IIIE) in manufacturing, cloud manufacturing (CMfg) integrates enterprises’ manufacturing information and provides an open and sharing platform for processing manufacturing tasks with distributed manufacturing services. Assigning tasks to manufacturing enterprises in the CMfg platform calls for effective scheduling algorithms. In recent years, deep reinforcement learning (DRL) has been widely applied to tackle cloud manufacturing scheduling problems (CMfg-SPs) because of its high generalization and fast-responding capability. However, the current DRL algorithms need to be trial-and-error through online interaction with the environment, which is costly and not allowed in the real CMfg platform. This paper proposes a novel offline DRL scheduling algorithm that alleviates the online trial-and-error issue while retaining DRL’s original advantages. First, we describe the system model of CMfg-SPs and propose the sequential Markov decision process modeling strategy, where all tasks are regarded as one agent. Then, we introduce the framework of the decision transformer (DT), which converts the online scheduling decision-making problem into an offline classification problem. Finally, we construct an attention-based model as the agent’s policy and train it offline under the DT architecture. Experimental results indicate that the proposed method consistently matches or exceeds online DRL algorithms, including deep double q-network (DDQN), deep recurrent q-network (DRQN), proximal policy optimization (PPO), and the offline learning algorithm behavior cloning (BC) in terms of scheduling performance and model generalization.

Deep Reinforcement Learning-Based Multi-Task Scheduling in Cloud Manufacturing Under Different Task Arrival Modes

Abstract Cloud manufacturing is a service-oriented networked manufacturing model that aims to provide manufacturing resources as services in an on-demand manner. Scheduling is one of the key techniques for cloud manufacturing to achieve the aim. Multi-task scheduling with dynamical task arrivals is a critical problem in cloud manufacturing. Many traditional algorithms such as the genetic algorithm (GA) and ant colony optimization algorithm (ACO) have been used to address the issue, which, however, either are incapable of or perform poorly in tackling the problem. Deep reinforcement learning (DRL) as the combination of deep learning (DL) and reinforcement learning (RL) provides an effective technique in this regard. In view of this, we employ a typical DRL algorithm—Deep Q-network (DQN)—and propose a DQN-based approach for multitask scheduling in cloud manufacturing. Three different task arrival modes—arriving at the same time, arriving in random batches, and arriving one by one sequentially—are considered. Four baseline methods including random scheduling, round-robin scheduling, earliest scheduling, and minimum execution time (min-time) scheduling are investigated. A comparison of results indicates that the DQN-based scheduling approach is effective and performs best among all approaches in addressing the multitask scheduling problem in cloud manufacturing.

A review of intelligent optimization for group scheduling problems in cellular manufacturing

A review of intelligent optimization for group scheduling problems in cellular manufacturing

Sequence generation for multi-task scheduling in cloud manufacturing with deep reinforcement learning

Cloud manufacturing is a manufacturing model that aims to deliver on-demand manufacturing services to consumers. Scheduling is an important problem that needs to be addressed carefully and effectively for cloud manufacturing to achieve that aim. Cloud manufacturing allows consumers to submit their requirements to the cloud platform simultaneously and therefore requires cloud manufacturing scheduling systems to be able to handle multiple tasks effectively. It is further complicated when multiple composite tasks are submitted to the system and to be addressed. A vast majority of existing studies have proposed various algorithms, including meta-heuristics, heuristics, and reinforcement learning algorithms, to address cloud manufacturing scheduling (CMfg-Sch) problems, but only a very small fraction of them deal with scheduling of multiple composition tasks with deep reinforcement learning. In this work, we leverage DRL coupled with sequence generation for addressing CMfg-Sch problems. Different from all existing works, we first propose two sequence generation algorithms for generating scheduling sequences of multiple composite tasks prior to scheduling. Coupled with this a Deep Q-Networks (DQN) and a Double DQN-based scheduling algorithms are proposed, respectively. Performance of the proposed algorithms is compared against seven baseline algorithms using makespan, cost, and reliability as evaluation metrics. Comparison indicates that sequence generation algorithm II (SGA-II) overall has a greater advantage over algorithm I (SGA-I), especially in terms of the makespan, and the Double DQN-based scheduling algorithm outperforms the DQN-based algorithm, which in turn performs better than other baseline algorithms.

Flexible job shop scheduling problem under Industry 5.0: A survey on human reintegration, environmental consideration and resilience improvement

The Job Shop Scheduling Problem (JSSP) has been widely studied in recent decades. Various approaches have been proposed to support scheduling decisions according to the evolving production environment. The emergence of technological advancements in the context of Industry 4.0 has brought many changes and made production scheduling more and more efficient. Today’s Industry 5.0 environment pays much attention to human considerations, sustainability, and resilience. These modern production environments can be accurately represented by the flexible shop floor scheduling problem in which various coordinating machines (with many alternative routing possibilities) and different operators are challenging. Recent literature on JSSP, which considers the human in the loop, has shown that the well-being and skills of workers significantly affect scheduling performance. In addition, knowing that industries are responsible for a significant part of the world’s energy consumption and greenhouse gas (GHG) emissions, new studies in scheduling focus on environmental factors. This paper introduces the Sustainable Flexible Scheduling Problem (SFJSSP) as a human and energy-efficiency-centered scheduling problem. First, we review the last decade’s literature on Flexible Job Shop Scheduling Problems (FJSSP) with human and/or environmental considerations. Next, we analyze the development trends in manufacturing scheduling problems. Finally, we discuss future research challenges to move towards scheduling 5.0 and suggest a mathematical model that considers human and environmental factors (in addition to the factors considered by the Classical Flexible Job Shop Scheduling Problem (CFJSSP)).

The Potential of Low-Power, Cost-Effective Single Board Computers for Manufacturing Scheduling

The ongoing fourth industrial revolution is pushing manufacturing towards smart, cognitive, and more adaptable systems. These new paradigms will require a huge number of computational resources. To meet these needs, the evolution of Industrial Internet added a layer of low-power heterogeneous devices with processing capabilities on the edge of the network, closer to the area where it is needed. Single board computers are the most common hardware platform found in the edge nodes. Due to its computation performance and power efficiency, the resources of this layer on the edge should be explored to solve processing demanding tasks. Maybe systems like the enterprise resource planning systems (ERP) and the manufacturing execution systems (MES) that are typically hosted in the industrial cloud can take advantage of sharing services between the cloud and the edge. This work aims to explore the edge layer collaboration with the cloud, sharing computational power to process the manufacturing scheduling problems linked to the ERP and MES. Therefore, in this preliminary study, we investigate the potential of single board computers to solve manufacturing scheduling problems. We present the experiments with four versions of Raspberry Pi, a well-known single board computer, evaluating them from the energy consumption and computation time points of view. Experiments show that it is possible to solve these scheduling problems in a reasonable time, with a lower energy consumption than other high-end computing devices.

Population-Based Meta-Heuristic Algorithms for Integrated Batch Manufacturing and Delivery Scheduling Problem

This paper addresses an integrated scheduling problem of batch manufacturing and delivery processes with a single batch machine and direct-shipping trucks. In the manufacturing process, some jobs in the same family are simultaneously processed as a production batch in a single machine. The batch production time depends only on the family type assigned to the production batch and it is dynamically adjusted by batch deterioration and rate-modifying activities. Each job after the batch manufacturing is reassigned to delivery batches. In the delivery process, each delivery batch is directly shipped to the corresponding customer. The delivery time of delivery batches is determined by the distance between the manufacturing site and customer location. The total volume of jobs in each production or delivery batch must not exceed the machine or truck capacity. The objective function is to minimize the total tardiness of jobs delivered to customers with different due dates. To solve the problem, a mixed-integer linear programming model to find the optimal solution for small problem instances is formulated and meta-heuristic algorithms to find effective solutions for large problem instances are presented. Sensitivity analyses are conducted to find the effect of problem parameters on the manufacturing and delivery time.

Solving task scheduling problems in cloud manufacturing via attention mechanism and deep reinforcement learning

Cloud manufacturing (CMfg) offers a cloud platform for both consumers and providers, and allocating consumer tasks to service providers requires many-to-many scheduling. Deep reinforcement learning (DRL) has been gradually employed to solve CMfg scheduling problems and has achieved satisfactory performance in dynamic and uncertain cloud environments. Nonetheless, we need better scheduling algorithms and more accessible modeling methods to enable practical implementation. In this study, an end-to-end scheduling algorithm is proposed to address task scheduling problems in CMfg. The proposal extracts intercorrelations in enterprise–enterprise and enterprise–task with the multi-head attention mechanism and is trained by DRL. Our proposal has extremely low time-response compared to heuristic algorithms and can generate a scheduling solution in seconds. In contrast to other DRL algorithms, our proposal can achieve better scheduling performances and uses a more accessible modeling method: the objective function alone is sufficient to train the model stably, without the need for a step-based reward function. Applying multi-head attention and DRL to scheduling problems is an exploratory attempt to achieve positive results. Experimental results on a case of automobile structure part processing in CMfg indicated that our proposal showed competitive scheduling performances and running time compared to eight DRL algorithms, two heuristic algorithms, and two priority dispatching rules. Besides, the results proved that our proposal’s generalizability and scalability were better than the other eight DRL algorithms.

Learning-Based Grey Wolf Optimizer for Stochastic Flexible Job Shop Scheduling

This work considers a stochastic flexible job shop scheduling with limited extra resources and machine-dependent setup time in a semiconductor manufacturing environment, which is an NP-hard problem. In order to obtain its reliable and high-performance schedule in a reasonable time, a learning-based grey wolf optimizer is proposed. In it, an optimal computing budget allocation-based approach, which is designed for two scenarios from real manufacturing environments, is proposed to intelligently allocate computing budget and improve search efficiency. It extends the application area of optimal computing budget allocation by laying a theoretic foundation. Besides, to obtain proper control parameters iteratively, a reinforcement learning algorithm with a newly designed delay update strategy is used to build a parameter tuning scheme of a grey wolf optimizer. The scheme acts as a guide for balancing global and local search, thereby enhancing effectiveness of the proposed algorithm. The theoretic interpretation of the developed optimal computing budget allocation-based approach and the convergence analysis results of the proposed algorithm are presented. Various experiments with benchmarks and randomly generated cases are performed to compare it with several updated algorithms. The results shows its superiority over them. Note to Practitioners—Meta-heuristic are often deployed to solve semiconductor manufacturing scheduling problems. However, they face to two thorny issues when they face stochastic manufacturing environments. 1) their computational efficiency is quite low, thus requiring substantial improvement, since a stochastic optimization problem requires Monte Carlo sampling to estimate the actual objective function values in a precise manner; and 2) most of them are parameter-sensitive, and choosing their proper parameters is highly challenging in such environments. To address the first issue, we develop an optimal computing budget allocation-based method for deciding the optimal numbers of sampling times based on both prior knowledge and simulation results. To address the second one, we propose a reinforcement learning algorithm to self-adjust the parameters of our proposed method called Learning-based Grey Wolf Optimizer. In addition, we design a delay update strategy to enhance its robustness, and thus, a feasible and high-quality schedule can be founded in a short time for real-time scheduling problems. Theoretic proofs and experimental results show that the proposed method is effective and efficient. Consequently, it can be readily applicable to practical semiconductor manufacturing systems.

Deep Reinforcement Learning Approach for Material Scheduling Considering High-Dimensional Environment of Hybrid Flow-Shop Problem

Manufacturing sites encounter various scheduling problems, which must be dealt with to efficiently manufacture products and reduce costs. With the development of smart factory technology, many elements at manufacturing sites have become unmanned and more complex. Moreover, owing to the mixing of several processes in one production line, the need for efficient scheduling of materials has emerged. The aim of this study is to solve the material scheduling problem of many machines in a hybrid flow-shop environment using deep reinforcement learning. Most previous work has ignored some conditions, which were critical for solving practical problems. Such critical conditions make the scheduling more complex and difficult to solve. They expand the size of the state and large action space and make learning in an environment with many machines problematic. In this study, a reinforcement learning approach was developed considering practical factors such as the processing time and material transfer to solve realistic manufacturing scheduling problems. Additionally, a method to simplify the high-dimensional environmental space at manufacturing sites for efficient learning was established to solve the problem of learning in a high-dimensional space. Through experiments, we showed that our approach could optimally schedule material scheduling in multi-process lines, which contributes to realistic manufacturing intelligence.

Dynamic scheduling of tasks in cloud manufacturing with multi-agent reinforcement learning

Cloud manufacturing provides a cloud platform to offer on-demand services to complete consumers’ tasks, but assigning tasks to enterprises with different services requires many-to-many scheduling. The dynamic cloud environment puts forward higher requirements on scheduling algorithms’ real-time response and generalizability. Additionally, complex manufacturing tasks with flexible processing sequences also increase the decision-making difficulty. The existing approaches either have difficulty meeting the requirements of dynamics and fast-respond or struggle to effectively capture features of tasks with flexible processing sequences. To address these limitations, we develop a novel scheduling algorithm to solve a dynamic scheduling problem in the group service cloud manufacturing environment. Our proposal is formulated and trained by multi-agent reinforcement learning. The graph convolution network encodes tasks’ graph-structure features, and the recurrent neural network records each task’s processing trajectories. We independently design the action space and the reward function and train the algorithm with a mixing network under the centralized training decentralized execution architecture. Multi-agent reinforcement learning and graph convolution networks are rarely used to cloud manufacturing scheduling problems. Contrast experiments on a case study indicate that our proposal outperforms the other six multi-agent reinforcement learning-based scheduling algorithms in terms of scheduling performance and generalizability.

Adjusted Iterated Greedy for the optimization of additive manufacturing scheduling problems

As a disruptive technology, additive manufacturing (AM) is revolutionizing manufacturing supply chains. AM consists of producing 3-dimensional objects through layer-by-layer addition of compound material based on digital models. The scheduling of AM operations differs from traditional (i.e., subtractive and injection molding) manufacturing with a single production run involving several parts/geometries; this makes the jobs heterogeneous. Limited studies have investigated the Additive Manufacturing Scheduling Problems (AMSP). This study extends the Iterated Greedy algorithm to solve the AMSPs considering a single-machine production setting. For this purpose, several computational mechanisms are customized to account for AM-specific characteristics of production scheduling. Numerical analysis shows that the vast majority of the best-found solutions are yielded by the Adjusted Iterated Greedy (AIG) algorithm considering both solution quality and stability; the outperformance becomes more significant with an increase in problem size. Statistical analysis confirms that AIG’s performance is notably better than that of the existing solution algorithm in terms of solution quality and stability. This study is concluded by providing directions for future development of AM and AMSPs to extend the industrial reach of 3D printing technology.

Mathematical Models for Minimizing Total Tardiness on Parallel Additive Manufacturing Machines

In this research we tackle the scheduling problem in additive manufacturing for unrelated parallel machines. Both the nesting and scheduling aspects are considered. Parts have several alternative build orientations. The goal is to minimize the total tardiness of parts. We propose a mixed-integer linear programming model which considers the nesting subproblem as a 2D bin-packing problem, as well as a model which simplifies the nesting subproblem to a 1D bin-packing problem. The computational efficiency and properties of the proposed models are investigated by numerical experiments. Results show that the total tardiness optimization significantly increases the complexity of the problem, only the simple instances are solved optimally, whereas the makespan variant is able to solve all testing instances. Using the 1D bin-packing simplification allows for solving more instances to optimality, but with a risk of obtaining nesting-infeasibility. We also observed the compromise between the total tardiness and makespan objectives, which originates from the dilemma of “packing more parts to benefit from the common machine setup/recoating time” or “packing less parts to maintain the flexibility for handling distributed duedates”.

State-of-the-art in optimisation and heuristics to solve manufacturing scheduling problem

State-of-the-art in optimisation and heuristics to solve manufacturing scheduling problem

A Review of Reinforcement Learning Based Intelligent Optimization for Manufacturing Scheduling

As the critical component of manufacturing systems, production scheduling aims to optimize objectives in terms of profit, efficiency, and energy consumption by reasonably determining the main factors including processing path, machine assignment, execute time and so on. Due to the large scale and strongly coupled constraints nature, as well as the real-time solving requirement in certain scenarios, it faces great challenges in solving the manufacturing scheduling problems. With the development of machine learning, Reinforcement Learning (RL) has made breakthroughs in a variety of decision-making problems. For manufacturing scheduling problems, in this paper we summarize the designs of state and action, tease out RL-based algorithm for scheduling, review the applications of RL for different types of scheduling problems, and then discuss the fusion modes of reinforcement learning and meta-heuristics. Finally, we analyze the existing problems in current research, and point out the future research direction and significant contents to promote the research and applications of RL-based scheduling optimization.

Additive manufacturing scheduling problem considering assembly operations of parts

Additive manufacturing scheduling problem considering assembly operations of parts

A hybrid multi-objective algorithm to solve a cellular manufacturing scheduling problem with human resource allocation

A cellular manufacturing system (CMS) is a suitable system for the economic manufacture of part families. Scheduling the manufacturing cells plays an effective role in successful implementation of the manufacturing system. Due to the fact that in the CMS, bottleneck machine and human resources are two important factors, which so far have not been studied simultaneously in a mathematical model, there should be a model to consider them. Therefore, this research develops a bi-objective model for CMS in a three-dimensional space of machine-part and human resources. The main objective is to minimize the maximum completion time of all tasks in the system and reduce the number of intercellular translocation based on bottleneck machines’ motion and human resources. Due to the NP-hardness of the studied problem, applying the conventional solution methods is very time-consuming, and is impossible in large dimensions. Therefore, the use of metaheuristic methods will be useful. The accuracy of the proposed model is investigated using LINGO by solving a small example. Then, to solve the problem in larger dimensions, a hybrid Multi-Objective Tabu Search-Genetic Algorithm (MO-TS-GA) is designed and numerical results are reported for several examples.