Data Mining in Manufacturing Industry

Advancements in manufacturing techniques and computers have made it easy to collect and store data in the field of manufacturing. But the problem faced by engineers and managers is to understand and interpret this large volume of data which is collected. New tools are used for extracting the large amount of data is been developed in data mining. These methods have a huge impact on the manufacturing industry.

Various factors affect the productivity and efficiency of a manufacturing industry. Some of the factors are product design, manufacturing process, decision making, and management and so on. Powerful data acquisition systems are actively used in manufacturing industry. Huge data related to bill of materials, product design, manufacturing process planning and scheduling, production process and systems, monitoring and diagnosis, market forecasting are collected and stored easily and quickly in the database at various stages / phases of manufacturing. This data acquisition technique enable manufacturing engineers and managers to capture, transfer, extract, understand and use the information / knowledge form large amount of datasets. Figures, charts and tables can be displayed within the office and production areas, reports and manuals can be generated from the large datasets which are acquired and mined.

For example these outputs can be used to inter-relate various entities / information. For example detecting failures in the manufacturing process and systems or finding defects of the quality of the products. In most of the cases just the human intervention might not be a powerful tool to inspect the large datasets which are collected. Whereas computer based data analysis and mining approaches would bring radical changes in the effectiveness of running a manufacturing industry.

Convectional statistical data analysis methods are no longer considered as the best methods that can be used in the manufacturing industry. Data mining based on fuzzy logic systems, machine learning and advanced statistics are some of the powerful new tools which are used for extraction of the information in manufacturing industries.

Data mining principles can also be used in various streams like finance, economics, business problems, bio-informatics, public administration, health care & science and manufacturing industry as discussed earlier and so on.

In my next blog I will continue on the same topic and discuss 2 case studies on manufacturing maintenance and assembly quality form data mining point of view. This will be followed by the benefits and limitation of data mining methods when used in manufacturing sector.

References - 

1. Applying data mining to manufacturing: The nature and implications - Kesheng Wang

2. Data mining for improving the quality of manufacturing: a feature set decomposition approach. Journal of Intellectual Manufacturing Rokach & Maimon, 2006