Multistage Manufacturing Processes Research Articles

Weighted Self-Regulation Complex Network-Based Variation Modeling and Error Source Diagnosis of Hybrid Multistage Machining Processes

Manufacturing of complex parts is often finished in a hybrid multistage machining process, in which various error transfers and accumulates generally happen in multiple processes. To identify the key processing features and error sources in the processing of complex parts, a complex network-based method is proposed in this paper. First, a weighted self-regulation variation propagation network (WSRVPN) is developed to describe the relationship between the actual error and the machining process in serial–parallel multistage manufacturing processes (SP-MMP). Then, the weighted LeaderRank algorithm is employed to recognize the key processing features in the constructed WSRVPN. Finally, the traveling algorithm is proposed to search the error propagation paths for potential problem nodes. A quantifying index called ContributionIndex is proposed to identify the error source from the propagation paths to determine the out-of-control machining nodes. An industrial case (i.e., machining process of the main bearing cap) based on SP-MMP is used to verify the effectiveness of the proposed method. The experimental results indicate that the proposed method can effectively identify key nodes and variation sources in SP-MMP.

Multistage Quality Control Using Machine Learning in the Automotive Industry

Product dimensional variability is a crucial factor in the quality control of complex multistage manufacturing processes, where undetected defects can easily be propagated downstream. The recent advances in information technologies and consequently the increased volume of data that has become readily available provide an excellent opportunity for the development of automated defect detection approaches that are capable of extracting the implicit complex relationships in these multivariate data-rich environments. In this paper, several machine learning classifiers were trained and evaluated on varied metrics to predict dimensional defects in a real automotive multistage assembly line. The line encompasses two automated inspection stages with several human-operated assembly and pre-alignment stages in between. The results show that non-linear models like XGBoost and Random Forests are capable of modelling the complexity of such an environment, achieving a high true positive rate and showing promise for the improvement of existing quality control approaches, enabling defects and deviations to be addressed earlier and thus assist in reducing scrap and repair costs.

Robust model-based control of multistage manufacturing processes

This paper presents a novel method that utilizes in-process measurements of product quality and models that relate those measurements with the underlying manufacturing process parameters to drive down the product quality errors via strategic adjustments of the controllable process parameters. Uniqueness of the new method is its robustness to inevitable inaccuracies in the underlying models, as well as the absence of traditional, but restrictive assumptions of Gaussianity and independence of measurement and process noise terms. The new approach was demonstrated using models and data from an automotive cylinder head machining process and an industrial-scale semiconductor lithography overlay process.

Sustainable Lot Size in a Multistage Lean-Green Manufacturing Process under Uncertainty

Optimal lot sizing is the primary tool applied by lean practitioners to reduce inconsistency in the manufacturing system to cut down inventories, which are often considered as a type of waste in the lean culture. Managers attempt to consider environmental impacts of the manufacturing system and find ways to reduce these effects while making efforts to achieve environmental protection. From a sustainability standpoint, carbon emissions are the major source of environmental contamination and degradation. In this context, this research provides an economic production quantity model with uncertain demand and process information in a multistage manufacturing process. This imperfect manufacturing process produces defective products at an uncertain rate, and is reworked to convert them into perfect quality products and reduce wastages. To control this uncertainty in the manufacturing process, the decomposition principle and the signed distance method of fuzzy theory are applied. The manufacturing process is analyzed with regard to environmental concerns, and a sustainable lot size is obtained through an interactive Weighted Fuzzy Goal Programming (WFGP) approach for the simultaneous achievement of economic and environmental sustainability. An experimental study is performed to verify the practical implication of the model, and results are evaluated through a sensitivity analysis. Important managerial insights and graphical illustrations are provided to elaborate the model.

Stacked denoising autoencoder‐based feature learning for out‐of‐control source recognition in multivariate manufacturing process

AbstractIn multivariate statistical process control (MSPC), regular multivariate control charts (eg, T2) are shown to be effective in detecting out‐of‐control signals based upon an overall statistic. But these charts do not relieve the need for multivariate process pattern recognition (MPPR). MPPR would be very useful for quality operators to locate the assignable causes that give rise to out‐of‐control situation in multivariate manufacturing process. Deep learning has been widely applied and obtained many successes in image and visual analysis. This paper presents an effective and reliable deep learning method known as stacked denoising autoencoder (SDAE) for MPPR in manufacturing processes. This study will concentrate on developing a SDAE model to learn effective discriminative features from the process signals through deep network architectures. Feature visualization is performed to explicitly present feature representations of the proposed SDAE model. The experimental results illustrate that the proposed SDAE model is capable of implementing detection and recognition of various process patterns in complicated multivariate processes. Analysis from this study provides the guideline in developing deep learning‐based MSPC systems.

Ensuring Product Quality, Consistency and Patient Supply over Time for a Large-Volume Biologic: Experience with Remicade®.

Biologics are produced from living organisms in complex, multi-stage manufacturing processes and contain inherent variability, which must be understood and controlled during manufacturing to avoid unexpected changes in key quality attributes that may contribute to clinically meaningful differences. The process must also meet large commercial demand, while simultaneously being able to accommodate change without sacrificing product consistency. The four key components of successful biologics manufacturing are (1) a stable, well-defined proprietary cell line; (2) a good manufacturing practice (GMP)-compliant supply chain with a process control strategy defining acceptable levels of variability for target product/process attributes and capable of managing complex material flows; (3) a tightly controlled procedure for implementation of proposed process changes that ensures product consistency; and (4) built-in redundancy and flexibility providing the ability to adapt rapidly to unexpected developments. This report describes the requirements for the manufacturing and distribution of biologics, using Remicade® (infliximab, Janssen Biotech, Horsham, PA, USA) as an example of best practices. Since Remicade’s first marketing approval in 1998, Janssen has manufactured > 150 million vials used to treat > 2.6 million patients around the world for a variety of inflammatory diseases. Remicade displays a highly consistent quality attribute profile and meets all product/process specifications across multiple manufacturing sites and process scales. Janssen’s experience with Remicade demonstrates that deep product knowledge, extensive manufacturing experience, diligent product/process monitoring and a sustained commitment to compliance and research are required to ensure quality, consistency and uninterrupted patient supply for large-volume biologics over the long term.

A Minimal-Sensing Framework for Monitoring Multistage Manufacturing Processes Using Product Quality Measurements

For implementing data analytic tools in real-world applications, researchers face major challenges such as the complexity of machines or processes, their dynamic operating regimes and the limitations on the availability, sufficiency and quality of the data measured by sensors. The limits on using sensors are often related to the costs associated with them and the inaccessibility of critical locations within machines or processes. Manufacturing processes, as a large group of applications in which data analytics can bring significant value to, are the focus of this study. As the cost of instrumenting the machines in a manufacturing process is significant, an alternative solution which relies solely on product quality measurements is greatly desirable in the manufacturing industry. In this paper, a minimal-sensing framework for machine anomaly detection in multistage manufacturing processes based on product quality measurements is introduced. This framework, which relies on product quality data along with products’ manufacturing routes, allows the detection of variations in the quality of the products and is able to pinpoint the machine which is the cause of anomaly. A moving window is applied to the data, and a statistical metric is extracted by comparing the performance of a machine to its peers. This approach is expanded to work for multistage processes. The proposed method is validated using a dataset from a real-world manufacturing process and additional simulated datasets. Moreover, an alternative approach based on Bayesian Networks is provided and the performance of the two proposed methods is evaluated from an industrial implementation perspective. The results showed that the proposed similarity-based approach was able to successfully identify the root cause of the quality variations and pinpoint the machine that adversely impacted the product quality.

Multi-scale computation of multistage manufacturing process signatures of glassy polymers multi-functionalisation

To functionalize a glassy polymer surface like the Polycarbonate (PC) several process are needed: milling, turning, polishing and a chemical attack process. These processes sign the surface on a large band frequency from mm-wavelength for the milling, to the nm-wavelength for the polishing and the chemical attack process, passing through the µm-wavelength for turning. However, the band frequency available with a measurement is limited by the size of the measure and the step scale of the sampling. In this paper, a method to characterize the evolution of the surface has been developed with Wavelet Transform approach. Surfaces of each step were measured with an interferometer under different magnifications (x5, x50, x115). Their signatures calculated with Mexican hat wavelet transform were lengthened by continuity between the different magnification. Some conditions of the continuity are needed as having the same size and the same step scale on the measurement by adding interpolation points for the lower magnification. The results show clear correlation between the multistage process signatures and the function printability evolution like optical, design and mechanical.

Monitoring and control of beta-distributed multistage production processes

Multistage statistical process control (SPC) is an effective procedure for ensuring quality of products in multistage manufacturing processes. Effective SPC approaches for monitoring and controlling quality in multistage processes are limited. Beta distributed data explained by process input variables are generally prevalent in multistage process industries. In this paper, we propose a generalized linear model for beta-distributed data in multistage processes in order to estimate the distribution parameters and develop two approaches for process monitoring and control: they are model-based SPC charts and a beta regression model. The proposed control charts outperform the existing control charts in terms of the out-of-control average run length for the detection of process mean shifts.

Influence of pre-polishing process on site flatness values of polished wafers

The preparation of semiconductor silicon polished wafer is a multi-stage manufacturing process. This paper analyzed the influence of polished wafers’ SFQR (Site flatness front least square range) values with different pre-polishing process. In this study, the pre-polishing processes included dual-side lapping & etching, dual-side grinding and back-side polishing. Among them, the etching process was divided into caustic etching and acid etching with different removal amount and different rotation speed. The experimental results show that different pre-polishing processes have significant effects on SFQR values of polished wafers. Caustic etching, dual-side grinding and back-side polishing are more suitable for polished wafer's SFQR, while the center area of acid etching wafers show worse polished wafer's SFQR due to the etching mechanism.

Multiple Testing in Regression Models With Applications to Fault Diagnosis in the Big Data Era

ABSTRACTMotivated by applications to root-cause identification of faults in multistage manufacturing processes that involve a large number of tools or equipment at each stage, we consider multiple testing in regression models whose outputs represent the quality characteristics of a multistage manufacturing process. Because of the large number of input variables that correspond to the tools or equipments used, this falls in the framework of regression modeling in the modern era of big data. On the other hand, with quick fault detection and diagnosis followed by tool rectification, sparsity can be assumed in the regression model. We introduce a new approach to address the multiple testing problem and demonstrate its advantages over existing methods. We also illustrate its performance in an application to semiconductor wafer fabrication that motivated this development. Supplementary materials for this article are available online.

Residual Life Prediction of Multistage Manufacturing Processes With Interaction Between Tool Wear and Product Quality Degradation

Multistage manufacturing processes (MMPs) usually exhibit an interactive relationship between tool wear and product quality degradation. On one hand, the tool wear in a stage may result in the quality degradation of the products fabricated on that stage. On the other hand, the quality degradation at a preceding stage may lead to the change of the operational condition and thus affect the tool wear in subsequent stages. This interaction needs to be considered to accurately predict the residual life distribution (RLD) of MMPs, which will benefit condition-based maintenance and tool inventory management. In this paper, we propose an interaction model that utilizes a linear model to represent the impact of tool wear on quality degradation and a stochastic differential equation model to capture the impact of quality degradation on the instantaneous rate of tool wear. We then propose a Bayesian framework that incorporates real-time quality measurements to online update the RLD of MMPs. Our methodology is a generalization of an existing “QR-chain model,” which is dedicated into a similar research and application area. We conduct numerical studies to test the performance of our methodology and compare with the QR-chain model. The results show that our methodology outperforms the QR-chain model through capturing the impact of quality degradation on the process of tool wear and incorporating real-time quality measurements.

Robustness Indices in Multistage Problems

The study of robustness in single‐stage manufacturing has been surveyed by many researchers, while modeling of robustness in multistage manufacturing using multivariate data is seldom explored. The aim of this paper is to develop a methodology to model process robustness in a multistage manufacturing system. This paper proposes a methodology to study the robustness of a multistage process that creates a component and process capability that acts as a surrogate measure of robustness. A powerful technique to facilitate the analysis of the mentioned problem, response surfaces express the behavior of responses considering the above controllable factors (quality characteristics). A common approach to estimating response surfaces using experimental data is the ordinary least squares method. Because this method is very sensitive to outliers and correlated responses, in this paper, a robust coefficient estimation method is proposed for multiresponse surfaces in multistage processes based on M‐estimators. The results show that considering the proposed robustness criteria based on a multivariate process capability index, the robust multistage process would perform better than the individual stage process. This paper adds to the body of knowledge on the robustness of multistage manufacturing processes, and also the relationship between robustness and capability is to be explored as surrogate robustness indices in the multistage process. The efficiency of the proposed approach is assessed by a small simulation study besides a real manufacturing case study. Copyright © 2017 John Wiley & Sons, Ltd.

A Goal-Oriented, Sequential, Inverse Design Method for the Horizontal Integration of a Multistage Hot Rod Rolling System

The steel manufacturing process is characterized by the requirement of expeditious development of high quality products at low cost through the effective use of available resources. Identifying solutions that meet the conflicting commercially imperative goals of such process chains is hard using traditional search techniques. The complexity in such a problem increases due to the presence of a large number of design variables, constraints and bounds, conflicting goals and the complex sequential relationships of the different stages of manufacturing. A classic example of such a manufacturing problem is the design of a rolling system for manufacturing a steel rod. This is a sequential process in which information flows from first rolling stage/pass to the last rolling pass and the decisions made at first pass influence the decisions that are made at the later passes. In this context, we define horizontal integration as the facilitation of information flow from one stage to another thereby establishing the integration of manufacturing stages to realize the end product. In this paper, we present an inverse design method based on well-established empirical models and response surface models developed through simulation experiments (finite-element based) along with the compromise decision support problem (cDSP) construct to support integrated information flow across different stages of a multistage hot rod rolling system. The method is goal-oriented because the design decisions are first made based on the end requirements identified for the process at the last rolling pass and these decisions are then passed to the preceding rolling passes following the sequential order in an inverse manner to design the entire rolling process chain to achieve the horizontal integration of stages. We illustrate the efficacy of the method by carrying out the design of a multistage rolling system. We formulate the cDSP for the second and fourth pass of a four pass rolling chain. The stages are designed by sequentially passing the design information obtained after exercising the cDSP for the last pass for different scenarios and identifying the best combination of design variables that satisfies the conflicting goals. The cDSP for the second pass helps in integrated information flow from fourth to first pass and in meeting specified goals imposed by the fourth and third passes. The end goals identified for this problem for the fourth pass are minimization of ovality (quality) of rod, maximization of throughput (productivity), and minimization of rolling load (performance and cost). The method can be instantiated for other multistage manufacturing processes such as the steel making process chain having several unit operations.

Multistage manufacturing process control robust to inaccurate knowledge about process noise

A method for robust control of quality errors in a multistage manufacturing process (MMP) is formulated, guaranteeing the product quality even under the worst-case scenario of uncertainties in the knowledge of noise characteristics. Simulations based on the error flow models in automotive cylinder head machining and lithography overlay processes show that the new robust control method increasingly outperforms the traditional stochastic control approach as uncertainties in the knowledge of process noise characteristics increase. It is also shown that controlling MMP quality, while ignoring inaccuracies in the noise model could make quality even worse than if no control is used.

The Numerical Simulation of Multi-Stage Sheet Metal Forming Based on Modified Yoshida-Uemori Hardening Model

Bauschinger effect is significant for metal forming, particularly for aluminum. A material constitutive model especially for multi-stage sheet metal forming is presented in this paper, which is improved based on Yoshida-Uemori(Y-U) model assumes that there exists different coefficient on equivalent back stress and boundary surface between stages. The prediction of this model is validated through real tension and compression test. Compared to other hardening rules, it can be shown that a more accurate result can be predicted by this model. This model is also successfully applied to be used in the numerical simulation of a multi-stage manufacturing process of an A-pillar, the experimental result demonstrates the advantage of this model in springback analysis in multi-stage simulation over other constitutive model.

Phase I monitoring of simple linear profiles in multistage processes with cascade property

When a multistage manufacturing process is monitored statistically, the cascade property results in a more complicated condition compared to the case when a single-stage process is controlled. The cascade property usually exists in different stages of a multistage process, where the quality of a stage influences the performance of the next stage. Moreover, sometimes the quality of a product/process is best characterized by a functional relationship. This relationship is referred to as a profile. In this paper, phase I monitoring of simple linear profile is addressed for a multistage process involving the cascade property. To aim this, the capabilities of the methods that may be used to monitor a profile in a multistage process are first assessed. Then, a statistic, named the U statistic, is introduced to provide the opportunity of removing the cascade property. This statistics provides quality engineers a way to reduce the complicated condition of monitoring a multistage process. The new approach also helps quality engineers to diagnose effectively the stage responsible for the out-of-control condition. To evaluate the effectiveness of the proposed approach, different simulated cases are analyzed numerically. In addition, a case study is provided to illustrate the applicability of the proposed method in real-world manufacturing environments.

Monitoring multistage processes with autocorrelated observations

In multistage manufacturing processes, autocorrelations within stages over time are prevalent and the classical control charts are often ineffective in monitoring such processes. In this paper, we derive a linear state space model of an autocorrelated multistage process as a vector autoregressive process, and construct novel multivariate control charts, CBAM and Conditional-based MEWMA, for detecting the mean changes in a multistage process based on a projection scheme by incorporating in-control stage information. When in-control stages are unknown, finding in-control stages is a challenging issue due to the autocorrelations over time and the sequential correlations between stages. To overcome this difficulty, we propose a conditional-based selection that chooses stages with strong evidences of in-control stage using the cascading property of multistage processes. The information of selected stages is effectively utilised in obtaining powerful test statistics for detecting a mean change. The performance of the proposed charts is compared with other existing procedures under different scenarios. Both simulation studies and a real example show the effectiveness of the conditional-based charts in detecting a wide range of small mean shifts compared with the other existing control charts.

New multivariate process capability indices for measuring the performance of multivariate processes subject to non-normal distributions

Purpose – In the past few years, several capability indices have been developed for evaluating the performance of multivariate manufacturing processes under the normality assumption. However, this assumption may not be true in most practical situations. Thus, the purpose of this paper is to develop new capability indices for evaluating the performance of multivariate processes subject to non-normal distributions. Design/methodology/approach – In this paper, the authors propose three non-normal multivariate process capability indices (MPCIs) RNMC p , RNMC pm and RNMC pu by relieving the normality assumption. Using the two normal MPCIs proposed by Pan and Lee, a weighted standard deviation method (WSD) is used to modify the NMC p and NMC pm indices for the-nominal-the-best case. Then the WSD method is applied to modify the multivariate ND index established by Niverthi and Dey for the-smaller-the-better case. Findings – A simulation study compares the performance of the various multivariate indices. Simulation results show that the actual non-conforming rates can be correctly reflected by the proposed capability indices. The numerical example further demonstrates that the actual quality performance of a non-normal multivariate process can properly reflected by the proposed capability indices. Practical implications – Process capability index is an important SPC tool for measuring the process performance. If the non-normal process data are mistreated as a normal one, it will result in an improper decision and thereby lead to an unnecessary quality loss. The new indices can provide practicing managers and engineers with a better decision-making tool for correctly measuring the performance for any multivariate process or environmental system. Originality/value – Once the existing multivariate quality/environmental problems and their Key Performance Indicators are identified, one may apply the new capability indices to evaluate the performance of various multivariate processes subject to non-normal distributions.

Research on Machining Allowance Distribution Optimization based on Processing Defect Risk

It is necessary to control every key process to make sure the output quality of final product is stability. Because there existing many fluctuating factors during multistage manufacturing processes, it is inevitable lead to numerous manufacturing defects, which would make the quality of final product at risk. In this paper, a model about processing defect risk was proposed to measure the quality of products, which was associated with multistage machining allowance. Then, it was used to optimize machining allowance distribution program of multistage manufacturing process. A machining example of slide valve & sleeve matching parts was presented to verify the rationality of this model.