Industry 4.1. Группа авторов

Чтение книги онлайн.

СКАЧАТЬ for scholars, practitioners, and C‐suite executives alike.

This book “Industry 4.1 – Intelligent Manufacturing with Zero Defects” focuses on improving the quality of products and processes, and is the culmination of the brilliant but down‐to‐the‐earth efforts of the team led by Professor Fan‐Tien Cheng over the past many years. The efforts started with Virtual Metrology. In view of the incompatible paces of fast production and slow metrology, 100% inspection is impossible, and sample inspection has been the practice. With the advancements in sensing, metrology, analytics and Industry 4.0 technologies, the team innovatively integrated physical metrology with its cyber counterparts, Virtual Metrology (VM). The resulting Automatic Virtual Metrology (AVM) system presented in this book is capable of predicting the quality of a product based on machine parameters, sensor data in the production equipment, and off‐line sampling measurements. It also provides on‐line and real‐time total inspection of all work pieces to timely detect abnormalities during production. As a result, the sampling rate of real measurements can be cut down, the production costs can be reduced, and the goal of nearly zero defects of deliverables can be achieved.

      Effective implementation of Automatic Virtual Metrology, however, is not easy, especially if we want it to be scalable to large factories and transferrable to other companies and other industries. Major infrastructure needs to be established efficiently and flexibly. Based on the team’s successful research, development, implementation, and redeployment at many factories and across multiple industries, this book methodically presents the essential infrastructure components. The content includes data collection and management and feature extraction; communication standards; computation infrastructure of cloud, edge, Internet of Things and big data; container‐related software development, deployment, and management technologies of Docker and Kubernetes; the overall architecture of the advanced manufacturing “Cloud of Things” framework, and the specific design and implementation of key components such as cyber‐physical agents, big data analytics application platform, the automated construction scheme for manufacturing services, and AVM and other servers.

      Extending the ideas, methods, and infrastructure presented above, the book then focuses on Intelligent Predictive Maintenance (IPM). Predictive maintenance, sometimes known as “condition‐based maintenance,” is to monitor the performance and conditions of equipment during operations to predict when equipment performance is deteriorating and when equipment is going to fail, followed by scheduled or corrective maintenance. Intelligent Predictive Maintenance presented in this book detects the abnormality of key components of manufacturing tools based on advanced fault detection and classification techniques and predicts their Remaining Useful Lives (RUL) using time series prediction algorithms. Factory‐wide implementation is then discussed to improve tool availability and prevent unscheduled down of manufacturing tools.

      Since modern manufacturing facilities are generally capital intensive, it is critical to have consistently high yields to justify the investment and to have a positive bottom line. Intelligent Yield Management (IYM) presented in this book is a closely related cousin of Intelligent Predictive Maintenance, with the purpose to effectively detect root causes that affect the yield. It consists of data collection and management; statistical, big data, and machine learning tools for defect and yield analysis; and timely resolution of issues discovered while maintaining the requisite quality and reliability standards. The kernel of the above is the “Key‐variable Search Algorithm” (KSA), which includes new root‐cause search methods for solving the high‐dimensional variable selection problem, and modules for checking the quality of input data and for evaluating the reliability of search results.

      The current Industry 4.0‐related technologies emphasize productivity improvement but not on quality enhancement. They can have the faith of achieving nearly Zero‐Defect Manufacturing but without effective methods to achieve it. By developing and implementing the novel methods, technologies, and infrastructure presented above, zero defects of products can be effectively achieved. This is what is defined as Industry 4.1 in the book. The actual deployment cases in seven industries, including flat panel display, semiconductor, solar cell, automobile, aerospace, carbon fiber, and blow molding, are presented in the final Chapter 11. The ingenuity is outstanding, the effort is tremendous, and the impact is far‐reaching and long‐lasting.

Since many acronyms are used throughout the book, readers are advised to have Abbreviation Lists handy when reading the book. Beyond this point, I sincerely hope that you enjoy reading the book, and delightfully discover the wonderful world of Industry 4.1.

      

      Peter B. Luh

      Board of Trustees Distinguished Professor

      SNET Professor of Communications & Information Technologies

      Dept. of Electrical & Computer Engineering

      University of Connecticut

      Since Germany brought up Industry 4.0 in 2012, the trend of Intelligent Manufacturing has boomed globally. By integrating the innovative information‐and‐communication technologies such as IoT, Cloud, Big Data, AI, etc., various Cyber‐Physical Systems are developed to promote factory process optimization, yield improvement, efficiency enhancement, and cost reduction. Besides, in response to changes in consumers' habits, Zero Defects, High Variety Low Volume, and Rapid Change have become mandatory indicators for Intelligent Manufacturing.

      Advanced Semiconductor Engineering Inc. (ASE), is the leading provider of independent semiconductor manufacturing services in assembly and test. ASE develops and offers complete turnkey solutions in IC packaging, design and production of interconnect materials, front‐end engineering test, wafer probing, and final test. In 2011, ASE started to vigorously promote Intelligent Manufacturing and established over 15 lights‐out factories in response to changes in the global industrial environment. Moreover, ASE also collaborated with various top universities in Taiwan, ROC for R&D of IoT, Cloud, Big Data, and AI technologies, which have cultivated more than 400 professionals in the automation field via co‐hosting educational trainings and industry programs to improve the automation capability within ASE.

      ASE began the industry‐university collaboration with Prof. Fan‐Tien Cheng in 2014. Initially, we implemented Automatic Virtual Metrology (AVM) to achieve total inspection in an efficient and economic way so as to reduce the measurement cost. The project was a great success, and ever since then Prof. Cheng has become one of our major collaborators. The subsequent cooperation includes Intelligent Yield Management (IYM), Intelligent Predictive Maintenance (IPM), Advanced Manufacturing Cloud of Things (AMCoT), and Scheduling, which can be said to be the practical applications of all the research essence of Prof. Cheng on the production line.

      The Industry 4.1 proposed by Prof. Cheng aims at Zero Defects, it applies AVM to accomplish total inspection and utilizes IYM to find the root causes of a yield loss. In addition to enhancing production efficiency, it also improves product yield and makes products close to Zero Defects, which is a great step forward in the realm of Industry 4.0.

      Although Intelligent Manufacturing is a hot subject nowadays, it is challenging for the enterprises to actually carry it out; many enterprises still struggle to realize the vision of Intelligent Manufacturing. The implementation of novel technologies isn’t the only core for Intelligent Manufacturing, the shaping of the ecological chain of the automation industry and the cultivation of talents are also important factors.

As the development of hardware like sensor, microcontroller, Automatic Material СКАЧАТЬ