Manufacturing Scheduling Research Articles

Cloud–edge collaboration task scheduling in cloud manufacturing: An attention-based deep reinforcement learning approach

Cloud Manufacturing (CMfg), as a service-oriented manufacturing mode, aims to provide consumers on-demand manufacturing services. The CMfg platform requires task scheduling technology to schedule manufacturing tasks efficiently, and improve resource utilization and customer satisfaction. Existing scheduling models for manufacturing tasks mainly consider maximizing the quality of service for customers but ignore the actual production execution, which will lead to low-quality execution or delayed delivery. To maximize customer satisfaction and balance production, this article studies a cloud–edge collaboration manufacturing task scheduling in CMfg (CETS). CETS refines manufacturing services deployed in the cloud to the factory process level, and schedules tasks according to the real-time production information on the edge side and manufacturing service information on the cloud side. Considering the dynamics of CETS and the complexity of state information in CETS, an attention-based deep reinforcement learning (DRL) algorithm is proposed to solve CETS. First, the CETS is mathematically represented and built as a partially observable Markov decision process. Second, on-policy maximum a posteriori policy optimization (V-MPO) with gated transformer-XL (GTrXL) named AV-MPO is developed. The effectiveness, training stability, generalizability, scalability, and robustness of AV-MPO are investigated. Rule-based algorithms and some state of art DRL algorithms, such as proximal policy optimization (PPO), soft actor-critic (SAC), and dueling deep q network (Dueling DQN), are compared with AV-MPO. The experimental results validate that AV-MPO can deal with the CETS problem more effectively.

Robustness Evaluation Process for Scheduling under Uncertainties

Scheduling production is an important decision issue in the manufacturing domain. With the advent of the era of Industry 4.0, the basic generation of schedules becomes no longer sufficient to face the new constraints of flexibility and agility that characterize the new architecture of production systems. In this context, schedules must take into account an increasingly disrupted environment while maintaining a good performance level. This paper contributes to the identified field of smart manufacturing scheduling by proposing a complete process for assessing the robustness of schedule solutions: i.e., its ability to resist to uncertainties. This process focuses on helping the decision maker in choosing the best scheduling strategy to be implemented. It aims at considering the impact of uncertainties on the robustness performance of predictive schedules. Moreover, it is assumed that data upcoming from connected workshops are available, such that uncertainties can be identified and modelled by stochastic variables This process is supported by stochastic timed automata for modelling these uncertainties. The proposed approach is thus based on Stochastic Discrete Event Systems models and model checking techniques defining a highly reusable and modular process. The solution process is illustrated on an academic example and its performance (generecity and scalability) are deeply evaluated using statistical analysis. The proposed application of the evaluation process is based on the technological opportunities offered by the Industry 4.0.

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.

An improved memetic algorithm to solve the energy-efficient distributed flexible job shop scheduling problem with transportation and start-stop constraints.

In the current global cooperative production environment, modern industries are confronted with intricate production plans, demanding the adoption of contemporary production scheduling strategies. Within this context, distributed manufacturing has emerged as a prominent trend. Manufacturing enterprises, especially those engaged in activities like automotive mold production and welding, are facing a significant challenge in managing a significant amount of small-scale tasks characterized by short processing times. In this situation, it becomes imperative to consider the transportation time of jobs between machines. This paper simultaneously considers the transportation time of jobs between machines and the start-stop operation of the machines, which is the first time to our knowledge. An improved memetic algorithm (IMA) is proposed to solve the multi-objective distributed flexible job shop scheduling problem (MODFJSP) with the goal of minimizing maximum completion time and energy consumption. Then, a new multi-start simulated annealing algorithm is proposed and integrated into the IMA to improve the exploration ability and diversity of the algorithm. Furthermore, a new multiple-initialization rule is designed to enhance the quality of the initial population. Additionally, four improved variable neighborhood search strategies and two energy-saving strategies are designed to enhance the search ability and reduce energy consumption. To verify the effectiveness of the IMA, we conducted extensive testing and comprehensive evaluation on 20 instances. The results indicate that, when faced with the MODFJSP, the IMA can achieve better solutions in almost all instances, which is of great significance for the improvement of production scheduling in intelligent manufacturing.

Hyper-heuristics applications to manufacturing scheduling: overview and opportunities

modern industrial challenges require solutions that are both efficient and reactive. Heuristic-based approaches allow production systems to react quickly to unexpected events and disturbances. Thus, the term Hyper-Heuristic (HH) covers a wide variety of techniques that allow the selection or generation of heuristics. To conciliate the speed of heuristics with the necessary global performance, a number of mechanisms have been proposed in the literature. This paper first presents the HHs. Afterward, inspired from previous works, a classification is given to categorize them. In addition, an enhancement is proposed. Indeed, a third type of HHs is added (i.e. to classic selection HHs and generation HHs). This new category is named “mixed” HHs. Then, different contributions from the literature to dynamic scheduling are highlighted. Before concluding, a number of trends and future directions such as the use of Machine Learning and simulation are explored.

Perspectives on future research directions in green manufacturing for discrete products

With the increasing concern due to climate change caused by a higher atmospheric concentration of CO2 and other greenhouse gases, reducing environmental impact is becoming more important for every part of society. Manufacturing is responsible for a significant amount of energy/material consumption and environmental burden and, therefore, has a great opportunity to reduce its impact through green manufacturing. Green manufacturing presents opportunities across the manufacturing enterprise to increase the efficient usage of energy and material resources. These opportunities include designing products to consume fewer materials and energy during manufacturing and use, incorporating more efficient manufacturing processes, streamlining and optimizing manufacturing schedules and plans, and circularizing products. The goal of this paper will be to provide a perspective from the authors on the opportunities that exist within green manufacturing for discrete products through a review of pertinent topics and future directions. The paper will focus on processes, manufacturing equipment, manufacturing systems, recovering value at a product’s end-of-life, and additional thoughts that include metrics and indicators, techno-economic assessment, and a discussion of efficiency and effectiveness. Key findings from this review include a need for social indicators and renewable energy considerations in scheduling and process planning, integrating Industry 4.0 into circular economy along with social and institutional dimensions, consistency in the ability to measure and conceptualize metrics and indicators, a detailed evaluation of the life cycle impacts and cost of Addit Manuf, and more human and environment-oriented considerations for smart manufacturing.

Approximation algorithms for bi-objective parallel-machine scheduling in green manufacturing

We consider bi-objective parallel-machine scheduling in green manufacturing to minimize the makespan and total processing cost. Each machine has a different constant processing cost per unit time. For the objective of minimizing the makespan, given a total cost budget, we provide an approximation algorithm with a worst-case ratio of 33+14≈1.686, which improves the previous bound of 2. For the objective of minimizing the total processing cost, subject to all the jobs must be completed before a given common deadline, we provide an approximation algorithm with a worst-case ratio of 2+r3, where r is the ratio of the maximum to the minimum processing cost per unit time on a machine.

A branch-and-price algorithm to perform single-machine scheduling for additive manufacturing

Additive manufacturing (AM) has attracted significant attention in recent years based on its wide range of applications and growing demand. AM offers the advantages of production flexibility and design freedom. In this study, we considered a practical variant of the batch-processing-machine (BPM) scheduling problem that arises in AM industries, where an AM machine can process multiple parts simultaneously, as long as the two-dimensional rectangular packing constraint is not violated. Based on the set-partitioning formulation of our mixed-integer programming (MIP) model, a branch-and-price (B&P) algorithm was developed by embedding a column-generation technique into a branch-and-bound framework. Additionally, a novel labelling algorithm was developed to accelerate the column-generation process. Ours is the first study to provide a B&P algorithm to solve the BPM scheduling problem in the AM industry. We tested the performance of our algorithm using a modern MIP solver (Gurobi) and real data from a 3D printing factory. The results demonstrate that for most instances tested, our algorithm produces results similar or identical to those of Gurobi with reasonable computation time and outperforms Gurobi in terms of solution quality and running time on some large instances.

The Effectiveness of a Vehicle Manufacturing Line Using Probabilistic Process Timing

Abstract: Automotive manufacturing involves several steps, starting with the casting of components and moving on to the machining and assembly processes. The assembly system, particularly for the Single Cylinder S.I. Engine, will be the only emphasis of this project. A considerable number of components must be integrated throughout the engine assembly process. The creation of a decision assistance system based on a discrete event simulation model is discussed in the work provided in this paper. The four closed-loop network arrangement of vehicle assembly and preassembly lines connected by conveyors is targeted at a particular type of manufacturing lines. The design and execution of such industrial processes frequently include simulation. This simulation research aims to select the process throughput for the Single Cylinder S.I. Engine Assembly line. Any possible machine bottlenecks will be identified by the simulation study to avoid blockages and starvation and increase process throughput. Machine cycle times for various processes should be included in the input data for the research. The Arena simulation program will be utilized in this investigation. The goal is to use existing equipment to boost output on the present Head Assembly Line and Sub-Assembly Lines to an anticipated capacity. The outcomes of the simulation will be used to pinpoint the assembly-line procedures that may have an impact on the S.I. Engine's manufacturing schedule. By contrast the study's findings with the actual process throughput, the findings will be illustrated. Additionally, statistical analysis will be done to boost performance.

Analysis of a Collaborative Scheduling Model Applied in a Job Shop Manufacturing Environment

Collaborative Manufacturing Scheduling (CMS) is not yet a properly explored decision making practice, although its potential for being currently explored, in the digital era, by combining efforts among a set of entities, either persons or machines, to jointly cooperate for solving some more or less complex scheduling problem, namely occurring in job shop manufacturing environments. In this paper, an interoperable scheduling system integrating a proposed scheduling model, along with varying kinds of solving algorithms, are put forward and analyzed through an industrial case study. The case study was decomposed in three application scenarios, for enabling the evaluation of the proposed scheduling model when envisioning the prioritization of internal–makespan-or external–number of tardy jobs-performance measures, along with a third scenario assigning a same importance or weight to both kinds of performance measures. The results obtained enabled us to realize that the weighted application scenario permitted reaching more balanced, thus a potentially more attractive global solution for the scheduling problem considered through the combination of different kinds of scheduling algorithms for the resolution of each underlying sub problem according to the proposed scheduling model. Besides, the decomposition of a global more complex scheduling problem into simpler sub-problems turns them easier to be solved through the different solving algorithms available, while further enabling to obtain a wider range of alternative schedules to be explored and evaluated. Thus, contributing to enriching the scheduling problem-solving process. A future exploration of the application in other types of manufacturing environments, namely occurring in the context of extended, networked, distributed or virtual production systems, integrating an increased and variable set of collaborating entities or factories, is also suggested.

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.

Analysis of proliferated low Earth orbit constellation resilience against solar weather radiation effects

Proliferated low Earth orbit constellations are becoming more common both in reality and in planning. The resilience of these constellations to the radiation effects of adverse solar weather is of interest given that some of them pass over polar regions. This paper stochastically simulates solar weather events over the course of solar cycle 25 in order to quantify radiation effects on a model of Planet Lab’s Flock constellation under various design and programmatic decisions. The Flock constellation is a highly proliferated, commercial, and polar low Earth orbit earth observation architecture. Lifecycle and radiation failures are simulated along with a reconstitution system in order to quantify architectural performance as a function of constellation population. This time-based measure of performance is then used to quantify the constellation resilience and allow for programmatic and design variables to be altered in order to observe the effects on the resilience of proliferated constellations against adverse solar weather radiation effects. A sensitivity analysis is performed in order to assess which decisions most improve or degrade resilience. Broadly, the results show that commercial off the shelf components and programmatic decisions such as launch and manufacturing schedules may improve constellation resilience against adverse solar weather effects. More specifically, the results also show that commercial off the shelf components with even basic shielding levels provide high resilience to cumulative total ionizing dose effects and single event upset radiation effects. However, any further reductions to shielding would have significant impacts on single event upset rates whereas total ionizing dose would remain negligible. Furthermore, manufacturing time for Flock satellites is sufficient to maintain a resilient reconstitution strategy. Finally, results show that in order to improve the resilience afforded by a reconstitution strategy, improvements to manufacturing and launch times should be made in parallel, rather than one or the other in isolation.

A hybrid manufacturing scheduling optimization strategy in collaborative edge computing

A hybrid manufacturing scheduling optimization strategy in collaborative edge computing

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.

Time-based resilience metric for smart manufacturing systems and optimization method with dual-strategy recovery

Smart manufacturing (SM) enables production scheduling to automatically adjust the original plan to meet customer demands. The deep integration of advanced information technologies also makes SM systems prone to a wide range of possible attacks. However, rapid recovery in response to damage has not aroused enough recognition. This paper addresses the issue of resilience metrics and recovery optimization for SM systems. First, the characteristics of smart manufacturing scheduling (SMS), coupled with its execution, damage, and dual-strategy recovery behaviors, are analyzed. Afterward, a novel time-based resilience metric oriented toward the quantification of rapid recovery is proposed. Furthermore, a decision-making framework composed of a joint optimization model and a modified pigeon-inspired optimization algorithm with an enhanced learning strategy and crossover operator (PIOLC) is established. Last, a case study of a flexible job shop scheduling problem with 8 jobs and 8 machines is conducted to verify the effectiveness of the work. Experimental results show that the proposed approach can achieve dual-strategy recovery optimization by increasing the system resilience to 86.2%.

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.

Quantum Optimization with Lagrangian Decomposition for Multiple-process Scheduling in Steel Manufacturing

Quantum Optimization with Lagrangian Decomposition for Multiple-process Scheduling in Steel Manufacturing

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.

Symbolic Scheduling of Robotic Cellular Manufacturing Systems With Timed Petri Nets

To reduce the computational burden in the scheduling of robotic cellular manufacturing (RCM) systems based on Petri nets’ (PNs) reachability graphs, existing methods mainly focus on the relaxation of a graph search algorithm at the cost of schedule optimality. Different from that, this article presents a method that accelerates the search process by using symbolic representations and manipulations. The proposed method uses binary decision diagrams (BDDs) to functionally represent and evolve discrete and place-timed PNs of RCM systems and then schedule them with a symbolic A* search to find an optimal solution with minimal makespan. It uses compact BDD structures to represent sets of states, instead of individual states, and then performs efficient Boolean manipulations to evolve the net and search for a system schedule. A heuristic function and its Boolean implementations for the symbolic A* search are developed. The admissibility of the proposed heuristic function is proven, which guarantees the optimality of the obtained schedule. The computational time is thus reduced without sacrificing the schedule optimality. Finally, experiments are carried out to show the effectiveness and efficiency of the presented method.

Optimizing job release and scheduling jointly in a reentrant hybrid flow shop

In a reentrant manufacturing system, job release and scheduling are of great importance. However, the two decisions have been made separately in the past. This study mainly aims at optimizing job release and scheduling jointly in a reentrant hybrid flow shop (RHFS-OJRSJ) with unrelated parallel machines. To promote energy conservation and emission reduction, a mathematical model for minimizing the makespan and total energy consumption is proposed. Due to the frequent reentrances of jobs and the complexity of the RHFS-OJRSJ problem, it is a strongly NP-hard problem. To effectively solve the RHFS-OJRSJ problem, an improved multi-objective evolutionary algorithm based on inverse optimization and decomposition (IMOEA/IOD) is proposed. First, the improved multi-objective evolutionary algorithm based on decomposition is employed to search the solution space. Second, a genetic algorithm is integrated to determine an initially best job release plan, based on which the scheduling problem can be encoded. Third, in the decoding, inverse optimization is introduced to further optimize the job release plan. Finally, to avoid falling into a local optimum and ensure population diversity, an adaptive neighborhood updating strategy is proposed. The experimental results demonstrate that the IMOEA/IOD algorithm bridges the gap of the job release and scheduling problem and the proposed algorithm can effectively solve the RHFS-OJRSJ problem.