Data Mining-driven Manufacturing Process Optimization

Advanced manufacturing such as aerospace, semi-conductor, and flat display device often involves complex production processes, and generates large volume of production data. In general, the production data comes from products with different levels of quality, assembly line with complex flows and equipments, and processing craft with massive controlling parameters. The scale and complexity of data is beyond the analytic power of traditional IT infrastructures. To achieve better manufacturing performance, it is imperative to explore the underlying dependencies of the production data and exploit analytic insights to improve the production process. However, few research and industrial efforts have been reported on providing manufacturers with integrated data analytical solutions to reveal potentials and optimize the production process from data-driven perspectives. In this paper, we design, implement and deploy an integrated solution, named PDP-Miner, which is a data analytics platform customized for process optimization in Plasma Display Panel (PDP) manufacturing. The system utilizes the latest advances in data mining technologies and Big Data infrastructures to create a complete analytical solution. Besides, our proposed system is capable of supporting automatically configuring and scheduling analysis tasks, and balancing heterogeneous computing resources. The system and the analytic strategies can be applied to other advanced manufacturing fields to enable complex data analysis tasks. Since 2013, PDP-Miner has been deployed as the data analysis platform of ChangHong COC 1 . By taking the advantages of our system, the overall PDP yield rate has increased from 91% to 94%.

Journal of Manufacturing Science and Engineering-transactions of The Asme, 2006

The paper reviews applications of data mining in manufacturing engineering, in particular production processes, operations, fault detection, maintenance, decision support, and product quality improvement. Customer relationship management, information integration aspects, and standardization are also briefly discussed. This review is focused on demonstrating the relevancy of data mining to manufacturing industry, rather than discussing the data mining domain in general. The volume of general data mining literature makes it difficult to gain a precise view of a target area such as manufacturing engineering, which has its own particular needs and requirements for mining applications. This review reveals progressive applications in addition to existing gaps and less considered areas such as manufacturing planning and shop floor control.

Journal of Intelligent Manufacturing, 2009

In modern manufacturing environments, vast amounts of data are collected in database management systems and data warehouses from all involved areas, such as product and process design, assembly, materials planning, quality control, scheduling, maintenance, fault detection and so on. Data mining has emerged as an important tool for knowledge acquisition in manufacturing databases. This paper reviews the literature dealing with knowledge discovery and data mining applications in the broad domain of manufacturing with an special emphasis on the type of functions to be performed on data. The major data mining functions to be performed include characterization and description, association, classification, prediction, clustering and evolution analysis. The papers reviewed have therefore been categorized in these five categories. It has been shown that there is a rapid growth in the application of data mining in the context of manufacturing processes and enterprises in the last 3 years. This review reveals the progressive applications and existing gaps identified in the context of data mining in manufacturing. A novel text mining approach has also been applied to the abstracts and keywords of 150 identified literatures to identify the research gaps and find the linkages between knowledge area, knowledge type and data mining tools and techniques applied.

IFAC-PapersOnLine, 2017

The Industry 4.0 concept assumes that modern manufacturing systems generate huge amounts of data that must be collected, stored, managed and analysed. The case study is focused on predicting the manufacturing process behaviour according to production data. The paper presents the way of gaining knowledge about the future behaviour of manufacturing system by data mining predictive tasks. The proposed simulation model of the real manufacturing process was designed to obtain the data necessary for the control process. The predictions of the manufacturing process behaviour were implemented varying the input parameters using selected methods and techniques of data mining. The predicted process behaviour was verified using the simulation model. The authors analysed different methods. The neural network method was selected for deploying new data by PMML files in the final phases. The objectives of the research are to design and verify the data mining tools in order to support the manufacturing system control by aiming at improving the decisionmaking process. Based on the prediction of the goal production outcomes, the actual control strategies can be precisely modified. Then they can be used in real manufacturing system without risks.

2012

In recent years data mining has become a very popular technique for extracting information from the database in different areas due to its flexibility of working on any kind of databases and also due to the surprising results. This paper is an attempt to introduce application of data mining techniques in the manufacturing industry to which least importance has been given. A taste of implement-able areas in manufacturing enterprises is discussed with a proposed architecture, which can be applied to an individual enterprise as well as to an extended enterprise to get benefit of data mining technique and to share the discovered knowledge among enterprises. The paper proposes conceptual methods for better use of different data mining techniques in product manufacturing life cycle. These techniques include statistical techniques, neural networks, decision trees and genetic algorithms. An integrated and unified data mining platform is anticipated then to improve overall manufacturing process.

In this paper basic concepts of machine learning and data mining are introduced. Machine learning algorithms extract knowledge from diverse data bases that can be used to build decision-making systems. For example, based on the operational engineering data, equipment faults can be detected, the number of items to be ordered can be predicted, optimal control parameters can be determined. A framework for organizing and applying knowledge for decision-making in manufacturing and service applications is presented. The framework uses decision-making constructs such decision tables, decision maps, and atlases. It offers a new data-driven paradigm of importance to modern manufacturing and service organisations. Examples of data mining applications in industrial, medical, and pharmaceutical domains are presented. It is envisioned that the data-driven framework presented in the paper will enhance these applications.

Indonesian Journal of Electrical Engineering and Computer Science, 2022

Recently due to the explosion in the data field, there is a great interest in the data science areas such as big data, artificial intelligence, data mining, and machine learning. Knowledge gives control and power in numerous manufacturing areas. Companies, factories, and all organizations owners aim to benefit from their huge; recorded data that increases and expands very quickly to improve their business and improve the quality of their products. In this research paper, the knowledge discovery in databases (KDD) technique has been followed, "association rules" algorithms "Apriori algorithm", and "chi-square automatic interaction detection (CHAID) analysis tree" have been applied on real datasets belonging to (Emisal factory). This factory annually loses tons of production due to the breakdowns that occur daily inside the factory, which leads to a loss of profit. After analyzing and understanding the factory product processes, we found some breakdowns occur a lot of days during the product lifecycle, these breakdowns affect badly on the production lifecycle which led to a decrease in sales. So, we have mined the data and used the mentioned methods above to build a predictive model that will predict the breakdown types and help the factory owner to manage the breakdowns risks by taking accurate actions before the breakdowns happen.

2004

Accurate planning of produced quantities is a challenging task in semiconductor industry where the percentage of good parts (measured by yield) is affected by multiple factors. However, conventional data mining methods that are designed and tuned on "well-behaved" data tend to produce a large number of complex and hardly useful patterns when applied to manufacturing databases. This paper presents a novel, perception-based method, called Automated Perceptions Network (APN), for automated construction of compact and interpretable models from highly noisy data sets. We evaluate the method on yield data of two semiconductor products and describe possible directions for the future use of automated perceptions in data mining and knowledge discovery.

International Journal of Computer Integrated Manufacturing , 2022

Nowadays, manufacturing companies worldwide are focusing on innovative ways to design the most efficient and effective internal processes. The datasets generated from both material and information flows within a production environment are precious resources that can be used to understand the current situation and to support managerial decisions. This develops a process mining based approach that aims to support the automated mapping and controlling of production support processes that are directly connected to the manufacturing process performance. This kind of processes can be made of several process variant thus traditional approaches (eg. interviews) can be biased by subjective opinion end not very effective. The goal of this work is to automatically map and analyse production support processes starting from execution data recorded by IT systems. In order to overcome the limitations of existing methodologies, the proposed approach exploits the joint application of two process mining algorithms: heuristic and inductive miner. As application problem, the main phases of the procure-to-pay process (P2P) of a manufacturing company are automatically identified and analysed. The managerial implications of using this approach are described within the application to a real case study. Tangible benefits have been provided to show the proposed approach effectiveness.

Proceedings of The Institution of Mechanical Engineers Part B-journal of Engineering Manufacture, 2006

Modern manufacturing systems equipped with computerized data logging systems collect large volumes of data in real time. The data may contain valuable information for operation and control strategies as well as providing knowledge of normal and abnormal operational patterns. Knowledge discovery in databases can be applied to these data to unearth hidden, unknown, representable, and ultimately useful knowledge. Data mining offers tools for discovery of patterns, associations, changes, anomalies, rules, and statistically significant structures and events in data. Extraction of previously unknown, meaningful information from manufacturing databases provides knowledge that may benefit many application areas within the enterprise, for example improving design or fine tuning production processes. This paper examines the application of association rules to manufacturing databases to extract useful information about a manufacturing system's capabilities and its constraints. The quality of each identified rule is tested and, from numerous rules, only those that are statistically very strong and contain substantial design information are selected. The final set of extracted rules contains very interesting information relating to the geometry of the product and also indicates where limitations exist for improvement of the manufacturing processes involved in the production of complex geometric shapes.