Multi-stage Production Research Articles

Hybrid feedback and reinforcement learning-based control of machine cycle time for a multi-stage production system

With the increasing need of flexible manufacturing systems, machine flexibility has become one of the major impact factors. An important aspect of machine flexibility is the ability to change an individual machine’s capacity (or cycle time) to improve the overall system efficiency. In this paper, a novel control method is proposed for multi-stage production systems to dynamically change the individual machines’ cycle time to improve overall system efficiency. The proposed control method integrates distributed feedback control scheme and a Reinforcement Learning (RL) control scheme based on an extended actor-critic algorithm. The feedback control will determine whether a machine is turned on or off using real-time system status, while the RL control scheme will decide how to increase or decrease a machine’s cycle time when a machine is on. An improved actor-critic RL algorithm is developed to add an auxiliary model-based path to the standard model-free RL to enhance the learning performance. To demonstrate the effectiveness of the proposed method, numerical case studies have been performed that clearly show improvements in the overall profits and energy savings compared to other methods.

Energy-Saving Control in Multistage Production Systems Using a State-Based Method

Manufacturers are pursuing energy-efficient production in response to the fluctuating energy price, growing global competition, rigorous international laws, and severe environmental crisis. This article proposes to boost the energy efficiency of production systems by controlling the production. It extends the existing energy-saving control research by presenting integrated modeling, analyzing, and controlling approaches. The work starts from the modeling of the production systems and establishes an analytical model to systematically quantify the production loss resulted from energy-saving control and the various disruptions. A dynamic control algorithm is proposed to reduce energy consumption and maintain desirable productivity. Simulation studies are utilized to demonstrate the application of the proposed method and verify its effectiveness. Note to Practitioners—The previous research indicates that it is possible to strategically turn off stations during production for energy saving. However, production systems are complex dynamic systems consisting of interconnected stations and supporting subsystems. Similar to station random failures, turning off stations for energy saving can severely jeopardize the production, and deviate the production from the desired target. Therefore, a quantitative method is established to calculate the production loss resulted from the energy-saving control and disruptions. The method is important to understand the real control cost. Based on the analysis, a dynamic control algorithm is formulated to balance the achieved control benefit and the production loss. It provides plant managers a useful tool to make energy-saving control decisions with a thorough understanding of production system dynamics. The presented research is established for serial batch production systems. The examples of batch stations include the refrigerator foaming equipment in refrigerator assembly lines and the vacuum oven in battery assembly lines. The model cannot be directly applied in serial-parallel production systems.

A Lot-Sizing Model for a Multi-State System with Deteriorating Items, Variable Production Rate, and Imperfect Quality

Conventional production systems assume that during the manufacturing processes, machines operate without breakdown over an infinite planning horizon and manufacture only products of good quality. Imperfect production processes as a result of machine degradation are common in manufacturing. This paper deals with a problem that concerns the modelling and evaluation of the performance of a multi-state production system that is subject to degradation and its effect on lot sizing. Here, we consider that the cycle starts with a particular production rate until a point when the inventory reaches a certain level after which the failure mode is activated due to the deterioration of certain components, leading to a reduction in the production rate in order to ensure the continuity of supply until the maximum inventory level is reached. Production then stops to restore the machine and the cycle starts again. We have assumed that the rate at which inventory deteriorates is exponential and that demand is constant. A numerical example is used to illustrate the model application, followed by sensitivity analysis. This paper contributes to lot sizing in the area of machine reliability by considering a production system in a degraded state with a non-increasing production rate for deteriorating items with imperfect quality and partial backlogging.

A new approach for quality prediction and control of multistage production and manufacturing process based on Big Data analysis and Neural Networks

As consumers care more and more about product quality, it is important to mine the deep correlations between production and manufacturing parameters and the evaluation of product quality through the analysis of industrial big data. The existing research of product quality prediction faces several major problems: the lack of diverse quality features, the poor tractability of abnormal parameters, the strong nonlinearity of parameters, the obvious sequential property of data, and the severe time lag of data. To solve these problems, this paper explores the quality prediction and control of multistage MP process (MPMP) based on big data analysis. Firstly, the prediction strategy and flow were specified for MPMP product quality prediction, and the features were extracted from MPMP product quality. After that, the MPMP product quality features were described in multiple dimensions, the attention mechanism was introduced to the prediction process. In addition, the recurrent neural network was improved, and an MPMP product quality prediction model was established on bidirectional long short-term memory (BiLSTM) network. Our model was compared with AdaBoost and XGBoost through experiments. The effectiveness of our model was demonstrated by the results of the appearance quality PQ1, and the area under the curve (AUC) for each process parameter. In general, our model is superior to other algorithms in the accuracy, mean accuracy, and precision of product quality prediction.

РОЛЬ АРТ-МЕНЕДЖМЕНТА В МУЗЫКАЛЬНОМ ШОУ-БИЗНЕСЕ МЕЙДЖОР-ЛЕЙБЛОВ

Current article discloses principal functions of Arts Administration in music industry on the example of World biggest labels’ practice. Practical evaluation of the proposed information lays in an opportunity to apply foreign experience in order to develop local music industry. The novelty of our investigation lies in consideration of production and post-production processes of musical product through the prism of marketing and social psychology – after all exactly an comprehension of potential customers’ needs and wants differentiate a pure music art from musical show-business. Major labels – are the brands and trade marks of the biggest companies, engaged in all multistage production and post-production processes of musical and video products. Precisely on the example of Major label activities it is possible to descry all the subtleties in musical show-business

Analyzing the Performance of a Multi Product Producer Company in Ramanathapuram District of Tamil Nadu

Farmer producer companies were started with the objective of increasing the income of farmers through aggregation, adoption of improved technologies and access to the service and markets. Sample Farmer Producer Company was started during 2015 and it is a multi product producer company. Financial performance is one of the important criteria for sustainability of the business by the FPCs. Hence this study was proposed with the objective of analyzing the performance of multi product Producer Company. Descriptive study was helped to full the objective of this research. Primary and secondary data were collected for analyzing the performance of sample FPC. Primary data analysis showed that sample FPC established strong forward and backward linkage. Based on t test the members of the sample FPC benefited in inputs purchase, access to farm machinery, credit, and crop insurance than the non members. Even though the company’s operational performance was better, financial performance was poor. Based on the altman Z score model results, the company will bankrupt in the near future. The external borrowing of the company limits the profitability. Hence support from the Government and better financial management will make the company financially sound in the near future.

Nonlinear fusion method for multistage product design decision-making using plant growth simulation algorithm

Design decision-making is a vital activity for selecting an optimal scheme for product development. Owing to the uncertainty and ambiguity of design requirements and constraints, several product design phases are often deployed for concept refinement, which makes multistage product design decision-making (MPDDM) and the effective fusion of MPDDM data indispensable. However, few existing methods have considered the nonlinear relationships among the MPDDM information. Therefore, a nonlinear fusion method for MPDDM was proposed in this study. This method applies a three-parameter interval grey number to depict decision-makers’ judgement about product design schemes. Based on converting linguistic judgements into interval scales, an interval analytic hierarchy process (AHP) method was employed to calculate the weights of the design criteria, decision-makers, and decision-making stages. Considering the advantage of integrating multiple matrices without requiring external control parameters, a multistage decision-making fusion process using a plant growth simulation algorithm (PGSA) was proposed to aggregate multistage decision-making data for product design. A case study was conducted to collect multistage decision-making data, and the PGSA was developed. Through comparison with the extant method, the effectiveness and feasibility of the fusion of MPDDM was verified. The results indicate that (1) uncertainty perceived by decision-makers at three stages accounted for 96.7%, 95%, and 97.2%, respectively. The “center of gravity” of a three-parameter interval grey number, which reflects the largest possibility of decision-makers’ preferences, is not always equally distant from its maximum or minimum value (73.9%). (2) The optimization model using interval AHP to calculate the weights of decision-making indicators and stages is conducive to reducing the decision-maker’s uncertainty. (3) The global search mechanism of the PGSA can effectively realize the nonlinear fusion of MPDDM.

Economic assessment of a serial production system with random imperfection and shortages: A step towards sustainability

Recent paradigm shifts in production modeling have evolved away from only economic considerations and toward the three pillars of sustainability: economic, environmental, and social, in order to accomplish sustainable development goals. While production modeling research has undergone a radical shift, quite a few studies in the production management literature have addressed sustainability challenges. These classic sustainable production models are only applicable to single-stage production systems and overlook a variety of real-world situations associated with random defective output and resulting shortages during demand fulfillment. To address this concern about developing sustainable strategies for multi-stage production, this paper analyzes a cleaner multi-stage production management system for carbon emissions reduction and active participation in corporate social responsibility activities, while also advancing the system economically. Three scenarios of random imperfect proportion are constructed, and a reworking opportunity is implied to clean the multi-stage production system. A planned backorder policy is advised to address anticipated shortages and improve the system’s service level. Analytical optimization techniques are used in the developed situations to determine the appropriate batch size and backorder quantity, hence minimizing the system’s overall cost and achieving global optimum solutions. Numerical evaluation and sensitivity analysis are used to conduct a comparative examination of the model scenarios. Significant managerial insights are developed to demonstrate the suggested cleaner multi-stage production model’s actual relevance. According to the model results, a lower batch size is the optimal strategy in response to increasing holding and CSR activity costs, but otherwise for the setup cost in a serial production system.

Advanced Springback Compensation Strategy through elimination of avoidable elastic strain energy

The stamping of dimensionally accurate sheet metal body components still represents a huge challenge in the automotive industry. This is partly caused by the multi-stage production process which complicates the design of an appropriate compensation strategy dramatically. A key prerequisite of an appropriate compensation strategy is to eliminate any undesired influence of elastic energy being potentially induced when closing the blankholders in the several operations throughout the production process. Here, a universally applicable compensation strategy is presented which fulfils this requirement thoroughly; by applying this process, the involved strain energy is reduced to a minimum due to proper part position in all operations. Technically this compensation strategy is achieved by, first, simulating all operations of the stamping process individually, second, by individually calculating the springback after each operation and, third, by accumulating the calculated deviation vector fields for each operation appropriately, which are then used for springback compensation. The process is time and cost effective and the required efforts are moderate even for complicated multistage stamping operations.

Machine learning derived dynamic operating reserve requirements in high-renewable power systems

Accurately forecasting wind and solar power output poses challenges for deeply decarbonized electricity systems. Grid operators must commit resources to provide reserves to ensure reliable operations in the face of forecast errors, a process which can increase fuel consumption and emissions. We apply neural network-based machine learning to expand the usefulness of median point forecast data by creating probabilistic distributions of short-term uncertainty in demand, wind, and solar forecasts that adapt to prevailing grid conditions. Machine learning derived estimates of forecast errors compare favorably to estimates based on incumbent methods. Reserves derived from machine learning are usually smaller than values derived using incumbent methods, which enables fuel savings during most hours. Machine learning reserves are generally larger than incumbent reserves during times of higher forecast error, potentially improving system reliability. Performance is tested using multistage production simulation modeling of the California Independent System Operator system. Machine learning reserves provide production cost and greenhouse gas (GHG) emission reductions of approximately 0.3% relative to historical 2019 requirements. Savings in the 2030 timeframe are highly dependent on battery storage capacity. At lower levels of battery capacity, savings of 0.4% from machine learning reserves are shown. Significant quantities of battery storage are expected to be added to meet California's resource adequacy needs and GHG reduction targets. The addition of these batteries saturates reserve needs and results in minimal within-hour balancing costs in 2030.

A Novel Analytical Modeling Approach for Quality Propagation of Transient Analysis of Serial Production Systems.

Production system modeling (PSM) for quality propagation involves mapping the principles between components and systems. While most existing studies focus on the steady-state analysis, the transient quality analysis remains largely unexplored. It is of significance to fully understand quality propagation, especially during transients, to shorten product changeover time, decrease quality loss, and improve quality. In this paper, a novel analytical PSM approach is established based on the Markov model, to explore product quality propagation for transient analysis of serial multi-stage production systems. The cascade property for quality propagation among correlated sequential stages was investigated, taking into account both the status of the current stage and the quality of the outputs from upstream stages. Closed-form formulae to evaluate transient quality performances of multi-stage systems were formulated, including the dynamics of system quality, settling time, and quality loss. An iterative procedure utilizing the aggregation technique is presented to approximate transient quality performance with computational efficiency and high accuracy. Moreover, system theoretic properties of quality measures were analyzed and the quality bottleneck identification method was investigated. In the case study, the modeling error was 0.36% and the calculation could clearly track system dynamics; quality bottleneck was identified to decrease the quality loss and facilitate continuous improvement. The experimental results illustrate the applicability of the proposed PSM approach.

Optimization of Multi-Stage Production Scheduling of Automated Production

Optimization of Multi-Stage Production Scheduling of Automated Production

Validation of satellite-based sea surface temperature products against in situ observations off the western coast of Sumatra

This study validated the sea surface temperature (SST) datasets from the Group for High-Resolution SST Multi Product Ensemble (GMPE), National Oceanic and Atmospheric Administration (NOAA) Optimal Interpolation (OI) SST version 2 and 2.1 (OIv2 and OIv2.1), and Estimating the Circulation and Climate of the Ocean, Phase II (ECCO2) in the area off the western coast of Sumatra against in situ observations. Furthermore, the root mean square differences (RMSDs) of OIv2, OIv2.1, and ECCO2 were investigated with respect to GMPE, whose small RMSD < 0.2 K against in situ observations confirmed its suitability as a reference. Although OIv2 showed a large RMSD (1–1.5 K) with a significant negative bias, OIv2.1 (RMSD < 0.4 K) improved remarkably. In the average SST distributions for December 2017, the differences among the 4 datasets were significant in the areas off the western coast of Sumatra, along the southern coast of Java, and in the Indonesian inland sea. These results were consistent with the ensemble spread distribution obtained with GMPE. The large RMSDs of OIv2 corresponded to high clouds, and it was suggested that the change in the satellites used for SST estimation contributed to the improvement in OIv2.1.

Ownership and Productivity in Vertically Integrated Firms: Evidence from the Chinese Steel Industry

AbstractWe study productivity differences in vertically integrated steel facilities using equipment-level information on inputs and output for each of the main stages in the value chain. We obtain stage-level productivity estimates by estimating a multistage production system and then integrate them into estimates for integrated facilities. At this level, we do not find statistically significant differences in productivity by ownership. This conceals important differences upstream and downstream in the value chain: private firms outperform in pig iron and steelmaking but lag in sintering. Inferior access to higher-quality raw materials and use of less automated technology are likely sources of these differences.

Application of a Machine Learning Algorithm in a Multi Stage Production System

Application of a Machine Learning Algorithm in a Multi Stage Production System

ProdSim: An Open-source Python Package for Generating High-resolution Synthetic Manufacturing Data on Product, Machine and Shop-Floor Levels

Due to steadily increasing product and manufacturing complexity, the mutual dependencies between production system states and product properties increase as well. The flow of information in a multi-stage production setup also represents a material flow, as the metadata about the product and the production process grows. Changes of the production system state or product properties can influence downstream process steps and thus create hidden dependencies across a production system. However, if the individual process steps are examined in isolation, which is common for currently available simulation software, the generated data do not reflect these dependencies. Therefore, this work introduces the open-source simulation environment ProdSim, which brings together process simulations on product, machine, and shop-floor levels. The shop-floor simulation layer models the material flow between individual process steps, ensuring that the generated data also map interactions across multiple production steps. In contrast, the machine level maps specific production processes using user-definable Python functions. ProdSim is specifically designed to synthesize high-resolution data with complex interaction patterns that can be used to develop and apply data-driven quality prediction techniques. Furthermore, the shop-floor level offers flexible modeling options, such as assembly hierarchies, buffer storage, different production types, multi-stage, and multi-product production. ProdSim is distributed as an open-source Python package, using the discrete-event simulation framework SimPy as a simulation core.

Agent-Based Distributed Data Analysis in Industrial Cyber-Physical Systems

Cyber-Physical System (CPS) is a key concept in the fourth industrial revolution, acting as the backbone infrastructure to develop Industry 4.0 compliant solutions based on decentralized networks of cyber-physical entities. In such systems, the distribution of intelligence and data analytics capabilities by different computational layers assumes a crucial relevance to support monitoring, diagnosis, prediction, and optimization tasks. Multiagent Systems (MAS) is a suitable approach to support this distribution through vertical and horizontal dimensions. In this context, this article introduces a modular agent-based architecture to balance the distribution of data analysis capabilities in industrial CPS. The proposed MAS solution was implemented in an industrial automotive factory plant, which applicability, novelty, and benefits were recognized, e.g., through its inclusion in the Innovation Radar of the Great EU-funded Innovations due to its contribution for the implementation of Zero Defects Manufacturing strategies in multistage production systems.

Predictive Maintenance Decision Making Based on Reinforcement Learning in Multistage Production Systems

Predictive maintenance has become increasingly prevalent in modern production systems that are challenged by high-mix low-volume production and short production life cycle. It is very helpful to prevent costly equipment failures, and reduce significant production loss caused by unscheduled machine breakdown. Although important, decision models for joint predictive maintenance and production in manufacturing systems have not been fully explored. Therefore, we propose a reinforcement learning based decision model, that brings together production system modeling and approximate dynamic programming. We start from the development of a state-based model by analyzing the dynamics of a multistage production system with predictive maintenance. It provides an approach to quantitatively evaluate the various disruptions as well as the maintenance decision’s impact on production. Then a reinforcement learning method is proposed to explore optimal maintenance policies, that optimize the production and maintenance cost. To further improve the performance of the production system, machine stoppage bottlenecks are defined. An event-based indicator is proved to identify bottlenecks with production data. We test the proposed models in simulation case studies. The proposed predictive maintenance decision model is compared with three policies, which are state-based policy (SBP), time-based policy (TBP) and greedy policy (GP). The numerical studies show that the proposed decision model outperforms the policies, and it has the lowest system cost that is 9.68%, 39.07%, and 39.56% lower than SBP, TBP, and GP, respectively. In addition, the research shows that bottleneck identification and mitigation could help manufacturing systems to achieve more than 9.00% throughput improvement.

Part Variation Modeling to Avoid Scrap Parts in Multi-stage Production Systems

Manufacturing systems for today's products are complex systems requiring a variety of different processes in order to be able to manufacture all necessary part features. This also applies to the production of rotating components, which have experienced increasing demand at the latest due to the growth in mobility. As in almost every manufacturing process, quality-reducing defects can occur due to deviations for example tool wear, which cannot always be avoided. Those, that have accumulated from previous process steps can cause the occurrence of superimposed defects. This leads to complex relationships between quality defects in the end product and the numerous parameters of the manufacturing processes. To remain competitive, production must be optimized in order to identify defects as early as possible, as well as their dependencies and variation patterns. The paper presents an approach to identify and model part variations within multi-stage production systems. Subsequently, based on a detected deviation, a downstream compensation strategy can be proposed at an early stage of the manufacturing process, which uses the capability of the overall system to fundamentally eliminate rejects.

Dynamic Hoist Scheduling for Multi-Recipe and Multi-Stage Production Lines: A Logical Framework

Dynamic Hoist Scheduling for Multi-Recipe and Multi-Stage Production Lines: A Logical Framework