Serial Production Lines Research Articles

Energy cost optimisation in two-machine Bernoulli serial lines under time-of-use pricing

Energy cost optimisation in manufacturing systems has gained more and more attention. Although there are many papers about energy consumption optimisation in serial production lines, energy cost optimisation in serial production lines has rarely been focused. In this paper, we formulate an energy cost optimisation problem in two-machine Bernoulli serial line under time-of-use pricing. We analyse the structural characteristics of the problem and transform the problem into optimally allocating the production rate among the time periods of different electricity rates. A definition of the extreme allocation is proposed and completed, and the optimal allocation is proved to be one of the extreme allocations. Using the property, an efficient method to solve the optimal allocation is proposed. With the help of the method, the multi-electricity-rate problem is transformed into several single-electricity-rate problems, which has been solved in the literature.

A finite state method in improvement and design of lean Bernoulli serial production lines

Production systems have a significant impact on national and global economies. It is, therefore, of particular interest to develop, master, and apply new methodologies improving the performance of the production systems. A systematic approach to the improvability and design for lean production in the case of the serial Bernoulli production lines is presented in this research using the finite state method and application of the Markovian modeling approach. A new concept of the bottleneck identification procedure is presented based on the analytical relationship between line properties and performance measures. Also, a new design methodology combining features of the finite state method and genetic evolution algorithm is developed to enable the efficient lean design of the serial Bernoulli production lines. The developed procedures are applied in the case of an automotive paint shop system including bottleneck identification concerning the production rate, the work-in-process, and the probabilities of blockage and starvation. Finally, the lean automotive paint shop system is obtained in an efficient and designer-friendly way.

Energy-efficient control in serial production lines: Modeling, analysis and improvement⋆

Control of production operations is one of the most cost-effective methods to improve energy efficiency in production lines. Although many research efforts have been devoted to utilizing production control to boost production energy efficiency, it is not clear how production dynamics change in reaction to the energy saving operations, which is important to capture energy saving opportunities. In this paper, an optimal energy saving control method based on N-policy is introduced to reduce the system energy consumption while only slightly decreasing productivity. A Markov chain model is established to interpret the dynamics of serial production lines with N-policy. The analytical formulas of production rate and energy consumption are derived. Then, an energy-efficient control model is formulated to balance the productivity and energy consumption. For two-machine production lines, the monotonicity property of the model is proved, and the analytical N-policy is derived. For longer production lines, an adaptive genetic algorithm is utilized to solve the energy-efficient control model. Case studies validate the effectiveness of the proposed optimal energy saving control method.

Energy-efficient buffer allocation problem in unreliable production lines

The buffer allocation problem (BAP) is one of the major optimization problems considered by production systems designers. The BAP is widely studied in the literature, since buffers have a great impact on improving the efficiency of production systems especially for mass production. Nevertheless, with all upcoming changes in world’s economy, industrial requirements and ecological awareness, the necessity of considering energy efficiency in manufacturing systems design becomes crucial. This paper introduces the first variant of energy-efficient buffer allocation problem (EE-BAP). It addresses a novel integrated analytical method to find the best buffer allocation that optimizes both throughput and energy consumption for serial production lines with unreliable machines. A large experimental study, based on different instances well established in the literature, is carried out to analyze the relevance of the proposed method. The obtained results show that considerable energy savings with low or almost no productivity loss are allowed. The energy-efficient BAP is a novel problem that considers the complex trade-off between throughput and energy efficiency performance and introduces new perspectives of research in the field of sustainable manufacturing systems design.

Similarity Transformation of Discrete Part Production Lines

In order to improve the computational efficiency of models of production lines with unreliable machines, we propose the similarity transformation method by adjusting the design parameters, such as times between machine failures, times to repair and buffer capacities, in an equal proportion. It is found that the system throughput remains almost unchanged under the similarity transformation. The typical manufacturing system, serial production line, is taken as an example to verify the proposed method. Theoretical analysis and numerical experiments have been performed to confirm the feasibility and effectiveness of the method. The results show that the throughput error is no more than 5% and the calculation time is much less than that of the original model. Therefore, it is helpful to apply the similarity transformation method to large production line models to reduce computational complexity.

Transfer lines with multi-machine stages and non-deterministic processing time: transient performance analysis and control

Serial transfer lines with parallel machines in each stage are now a fairly common manufacturing system. However, its analysis and predictive-reactive control during transients remains mostly unexplored. This manufacturing system consists of a series of stages where in each stage a finite number of machines work concurrently. A two multi-state stage transfer line is studied in this paper. To describe the probability of one part being completed from the several machines in a stage, an equivalent multi-state, perfectly reliable machine is proposed. This multi-state model enables variance in the processing time between job completions to be captured stochastically. Based on historic data or engineering experience, an equivalent multi-state machine is simulated from the cumulative distribution function of the aggregated behavior of the parallel machines. In this paper, a serial production line with two equivalent, multi-state machines is used to analyze transient performance. Analytical closed-form expressions are derived to evaluate system transient behaviors. This system is also used with a controller to implement practical production control. Using model predictive control, system control parameter values provide decision support for managers to control short-term production. The proposed methodology can be applied to improve resource efficiency in other operation management problems.

Development of an Efficient Prediction Model for Optimal Design of Serial Production Lines

One of the problems encountered in the design and implementation of a serial production line (SPL) is the buffer size between the machine tools. The buffer size of the SPL has an important impact on the productivity of the whole production system. The machine tools’ characteristics including their uptimes and downtimes and the process parameters are the main factors that affect the decision regarding the buffer size, and thus the productivity of the SPL. Due to the dynamic nature of this problem, it is complex to find the optimal buffer size in SPL. Thus, in this paper, an Efficient Prediction Model (EPM) is developed using Artificial Neural Network (ANN). The purpose of the developed EPM is to find the buffer size between each succeeding pair of machine tools in SPL at any given uptimes and downtimes of machine tools. An optimization model based on genetic algorithms (GA) is used to generate the learning data for the prediction model to find the optimal or near optimal buffer size of the bay of each machine tool in SPL. The proposed approach integrates the optimization and prediction methodologies to evaluate, and predict the optimal buffer sizes for maximum productivity. Including uptime and downtime parameters enable the proposed method to be used to improve the design of running SPL as well as to design a new SPL. Numerical examples for five and fifteen machine tools were conducted independently in this research and the results show the ability of the proposed method to determine the optimal buffer sizes in a reasonable amount of time. In particular, the results of case studies show that the developed model accurately predict the optimal buffer size, especially for the case of five machines and even for a higher number of machine tools yet with acceptable but less accuracy. Finally, the performance of the proposed approach was compared with some results of the state of the art methods reported in the literature. The comparison shows the superiority of the present approach to identify buffer sizes for higher throughput under the same uptimes and downtimes.

Data-Enabled Real-Time Modeling for Production Systems with Variable Cycle Time

Cycle time is one of the essential parameters that could lead to a significant improvement on the productivity of a manufacturing system. In the traditional manufacturing systems, cycle time for each station is recommended by the vendor and determined based on production line design and through repeated experiments and simulations. Once determined, the cycle time is typically unchangeable during the production operation. Nowadays, the development of smart manufacturing enables a flexible production operation, and the cycle time of each machine can be made changeable to quickly respond to market demand fluctuation and to further improve the system efficiency. To evaluate the performance of a serial production line with variable cycle time, a novel data-enabled manufacturing system model is developed, in which sensor data inputs are allowed and cycle times are modeled as input variables. In addition, an opportunity window concept, which has been studied in the previous study and has been treated as a real-time system status measurement, is extended to the variable cycle time scenario. Numerical case studies are presented to validate the fidelity of the data-enabled model and demonstrate the influence of variable cycle time of individual machines on the production line throughput and the opportunity windows.

Parameter Identification for Bernoulli Serial Production Line Model

Model-based analysis of production systems is one of the main areas in manufacturing research. The foundation of the successful application of these theoretical studies is the availability of valid and high-fidelity mathematical models that are capable of capturing the behavior of job flow in production systems. The modeling process of a production system, however, may require a significant amount of nonstandardized work that can only be done properly by someone with solid training in the area and extensive experience through real case studies. This poses a critical challenge in the effective implementation of these valuable theoretical results in the Industry 4.0 era. To overcome this, we propose a new production systems modeling paradigm inspired by system identification: calculate production system model parameters that best match the standard system performance metrics measured on the factory floor. Specifically, in this article, we consider production lines characterized by the Bernoulli serial line model and develop algorithms that identify model parameters to fit the system throughput and work-in-process. Analytical algorithms are derived to solve this problem in a two-machine line case and then extended to multi-machine lines. The accuracy and computational efficiency of the algorithms are demonstrated through extensive numerical experiments. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —A high-fidelity mathematical model is of critical importance to the implementation of any model-based production system analysis method. Currently, the construction of such models is carried out in an ad hoc manner. The quality of the resulting models may heavily depend on the training, experience, intuition, and personal preference of the modeler. The proposed model parameter identification method focuses on standard key performance indices commonly measured on the factory floor. The advantage is twofold. First, these standard performance metrics are consistently defined regardless of industry, thus avoiding any data-ambiguity issue that may occur when using complex machine/equipment status data. Second, measuring these performance metrics in real time is typically convenient and cost effective, even for manufacturing plants without high-end IT infrastructure, thus making the technology accessible to not only large but also small- and mid-sized manufacturers. Using the algorithms developed in this article, a practitioner can quickly construct a serial production line model and then utilize it to access the rich library of production analysis, design, and control methods available in the literature.

Modeling and Dynamic Assignment of the Adaptive Buffer Spaces in Serial Production Lines

Abstract In production systems, the buffer capacities have usually been assumed to be fixed during normal operations. Inspired by the observations from the real industrial operations, a novel concept of Adaptive Buffer Space (ABS) is proposed in this paper. The ABS is a type of equipment, such as movable racks or mobile robots with racks, which can be used to provide extra storage space for a production line to temporarily increase certain buffers’ capacities in a real-time fashion. A good strategy to assign and reassign the ABS can significantly improve real-time production throughput. In order to model the production systems with changing buffer capacities, a data-driven model is developed to incorporate the impact of buffer capacity variation in system dynamics. Based on the model, a real-time ABS assignment strategy is developed by analyzing real-time buffer levels and machine status. The strategy is demonstrated to be effective in improving the system throughput. An approximate dynamic programming algorithm, referred to as ABS-ADP, is developed to obtain the optimal ABS assignment policy based on the strategy. Traditional ADP algorithms often initialize the state values with zeros or random numbers. In this paper, a knowledge-guided value function initialization method is proposed in ABS-ADP algorithm to expedite the convergence, which saves up to 80% computation time in the case study.

Product Completion Time Prediction Using A Hybrid Approach Combining Deep Learning and System Model

The prediction of product completion time is critical in real time production scheduling and control to achieve customer demand satisfaction. However, it is a challenging task due to the increasing complexity of production systems and greater diversity of products. The recent advancement in data-driven approach and machine learning algorithms have provided unprecedent opportunities to tackle such problems that otherwise very difficult to solve using conventional methods in the manufacturing industry. However, most existing studies on product completion time prediction adopt a purely data-driven approach while ignoring the prospect of integrating domain knowledge in their machine learning models. In this paper, we propose a hybrid approach to predicting product completion time by combining the strengths of both machine learning techniques and analytical system model. A mathematical model for multi-product serial production line is proposed to describe the real-time dynamics of the system. With this model, the strict lower bound of product completion time can be efficiently computed for given system status, where the lower bound represents the least possible product completion time when assuming no random downtime in the system. Instead of directly predicting the product completion time, a deep learning model is developed to only predict the distance between the lower bound and actual product completion time. Guided by properties of production system, we discover a recurrent sequence in the prediction problem by modeling each machine and product as recurrent units. The Long Short-Term Memory (LSTM) method, a prominent variant of recurrent neural network (RNN), is used to combine with the system model to predict the product completion time in a real-time fashion.

Multifidelity Modeling for Analysis and Optimization of Serial Production Lines

Recent advances in sensing, data analytics, and manufacturing technologies (e.g., 3-D printing, soft robotics, nanotechnologies, etc.) provide the potential to produce highly customized products by allowing flexible system design, endless device configurations, and unprecedented information flows. These opportunities also increase the complexity of controlling such systems optimally, which typically requires fast exploration of an increasingly large number of alternative operation strategies. Simulation and stochastic models have been particularly successful to support control and optimization of production systems, and methods have been developed to exploit them separately. Herein, we argue that the simultaneous use of these models can allow for better control and optimization by balancing the simulation accuracy, and related high computational costs, with the computational efficiency and lower accuracy of stochastic models. In this article, we assume that high fidelity models have higher accuracy and computational costs, and we present a novel multifidelity approach, which utilizes several models at different levels of fidelity to efficiently and effectively estimate and optimize the performance of asynchronous serial production lines with machines suffering from multiple failure types. Experimental results show that the multifidelity approach leads to better estimations, requiring less computational effort for optimization compared with the use of only high fidelity simulations.

Real-Time Bottleneck in Serial Production Lines With Bernoulli Machines: Theory and Case Study

In production systems, the bottleneck is typically understood as the operation that affects the system performance in the strongest manner. Identification and mitigation of bottleneck are among the most important problems in production systems research and practice. The current literature offers numerous methods for the identification of production bottlenecks. Most of these methods focus on steady-state system behavior under constant parameters, and the corresponding mitigation approach is typically to prioritize the improvement of that particular operation. This applies when the improvement cannot be dynamically reallocated to other operations in the system, and the improvement is to be fixed to an operation for a long period of time. In this article, we attempt to study the bottleneck problem in a dynamic environment. Especially, we consider serial production lines with finite capacity buffers and the Bernoulli reliability machines and use a control-theoretic approach to formulate the bottleneck identification and mitigation problem as a state-based feedback control problem with the objective being the maximization of steady-state throughput. Computation formulas and procedures are developed to calculate the performance metrics of such systems with dynamic bottleneck control. Properties of the optimal control policies are investigated, and a computationally efficient algorithm is developed to obtain an effective real-time bottleneck indicator. Numerical experiments are used to demonstrate the efficacy of the proposed method, and a case study is presented to illustrate its application. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —It is very common in manufacturing practice that operators, engineers, and managers need to make decisions on-the-fly to ensure efficient production operations. One of the key tasks involved is to identify real-time bottleneck (RTBN). This is typically carried out based on the practitioners’ intuition and experience. With the advances in sensing, computing, and communication technologies emerging in the Industry 4.0 era, manufacturers now have the capability to collect, process, and visualize an enormous amount of data in real time. This provides a foundation for more rigorous real-time production management and control. The research described in this article is intended to provide such a tool for RTBN identification in production lines and, thus, guide timely decision-making on the factory floor for smarter production activities.

A Finite State Method in the Performance Evaluation of the Bernoulli Serial Production Lines

Research on the performance measure evaluation of Bernoulli serial production lines is presented in this paper. Important aspects of the modeling and analysis using transition systems within the Markovian framework are addressed, including analytical and approximation methods. The “dimensionality curse” problems of the large scale and dense transition systems in the production system engineering field are pointed out as one of the main research and development obstacles. In that respect, a new analytically-based finite state method is presented based on the proportionality property of the stationary probability distribution across the systems’ state space. Simple and differentiable expressions for the performance measures including the production rate, the work-in-process, and the probabilities of machine blockage and starvation are formulated. A finite state method’s accuracy and applicability are successfully validated by comparing the obtained results against the rigorous analytical solution.

Data-Enabled Modeling and Analysis of Multistage Manufacturing Systems with Quality Rework Loops

In a multistage serial production line, multiple inspection stations and repair processes are typically involved to ensure high product quality. Quality rework is the activity to repair or repeat the work on the defect parts during manufacturing processes. The rework process after each inspection can add cost and cycle time to the normal process and impose negative impact on the throughput. This paper studies real-time performance of multistage serial manufacturing systems with quality rework loops and machine random failures. A production line with multiple quality rework loops is first unify by segmenting it into a set of serially connected quality rework loops. An event-based data-enabled mathematical model is developed to evaluate real-time production rate of each machine for such a system. In addition, the system properties are analyzed and permanent production loss due to quality rework loops and random machine failures are identified respectively. The permanent production loss attribution to each disruption event and machine can be used as real-time performance indicators to diagnose production system inefficiency. The mathematical model and system performance identification methodology are studied analytically and validated through numerical case studies.

Deep reinforcement learning based preventive maintenance policy for serial production lines

In the manufacturing industry, the preventive maintenance (PM) is a common practice to reduce random machine failures by replacing/repairing the aged machines or parts. The decision on when and where the preventive maintenance needs to be carried out is nontrivial due to the complex and stochastic nature of a serial production line with intermediate buffers. In order to improve the cost efficiency of the serial production lines, a deep reinforcement learning based approach is proposed to obtain PM policy. A novel modeling method for the serial production line is adopted during the learning process. A reward function is proposed based on the system production loss evaluation. The algorithm based on the Double Deep Q-Network is applied to learn the PM policy. Using the simulation study, the learning algorithm is proved effective in delivering PM policy that leads to an increased throughput and reduced cost. Interestingly, the learned policy is found to frequently conduct “group maintenance” and “opportunistic maintenance”, although their concepts and rules are not provided during the learning process. This finding further demonstrates that the problem formulation, the proposed algorithm and the reward function setting in this paper are effective.

Bottleneck identification and alleviation in a blocked serial production line with discrete event simulation: A case study

Bottleneck identification and alleviation in a blocked serial production line with discrete event simulation: A case study

Hybrid multi-objective opposite-learning evolutionary algorithm for integrated production and maintenance scheduling with energy consideration

While conventional scheduling researches take production efficiency, cost and quality as objectives, increasingly serious ecological problems and energy shortage have turned scholars’ attention to energy-efficient scheduling. Meanwhile, maintenance activities are of great importance to equipment availability and production continuity. This paper addresses an energy-efficient multi-objective scheduling problem of a serial production line (SPSP) integrating production and maintenance, with the criteria of minimizing the makespan and the total energy consumption simultaneously. The impact of integrating preventive maintenance (PM) is analyzed, and the better interval mode is picked out. An energy-saving strategy combining shutdown windows of PM and energy-saving windows on idle machines is designed to cut down the total energy consumption. In order to tackle this NP-hard problem, a hybrid multi-objective opposite-based learning evolutionary algorithm (HMOLEA) is developed, in which the opposite-based learning and IGD indicator-based evolutionary mechanism are combined together and a special rank assignment is designed to improve the computational efficiency. Furthermore, a self-adaptive weighted mutation operator is fused into the framework of HMOLEA to enhance the exploration of the algorithm. Extensive computational experiments are carried out to verify the effectiveness and superiority of HMOLEA in solving the SPSP.

Order-Reduced Dynamic Decoupling Approach for Performance Evaluation of Multitype and Small-Batch-Based Serial Lines With Adjustments and Resets

In the existing literature, extensive investigations focus on evaluating the performance of production systems that typically assumed to be operating under rigid production mode. Mass production as well as single type of products is the representative characteristics of this production mode. However, by applying prompt equipment adjustments and resets, modern production lines are capable of making different products in small batches with varying processing requirements. This is referred to as flexible manufacturing mode. With the flexible manufacturing mode, which has been widely used in recent years, real-time performance evaluation and prediction, small-batch-based real-time scheduling, and management of such systems, etc., are having important research significance. In this article, by considering adjustments and resets on serial production lines with multitype and small-batch-based production, the transient performance evaluation problem is studied. Specifically, for the system with machines having the Bernoulli reliability model and buffers having finite capacities, the mathematical model is formulated first. Then, we develop a Markovian model and a computational procedure for evaluating various system real-time performance indicators, such as production rates and completion times of different batch-based production tasks, etc. In addition, an order-reduced dynamic decoupling approach is proposed to approximate system real-time performance with high accuracy and computational efficiency. Finally, theoretic properties of the system under the flexible manufacturing mode are discussed.

WITHDRAWN: Deep Reinforcement Learning based Preventive Maintenance Policy for Serial Production Lines

The Publisher regrets that this article is an accidental duplication of an article that has already been published, http://dx.doi.org/10.1016/j.eswa.2020.113701 . The duplicate article has therefore been withdrawn. The full Elsevier Policy on Article Withdrawal can be found at https://www.elsevier.com/about/our-business/policies/article-withdrawal