Samsung Electro-Mechanics is a global electronics parts manufacturer with production facilities in eight locations around the globe. The company measures productivity and competitiveness to increase its yield. In order to carry out this analysis, Samsung Electro-Mechanics deployed Microsoft® SQL Server® 2005 and Microsoft® Office Excel® 2007 within its Business Integration (BI) framework, mapped to its business models. Following its deployment, Samsung Electro-Mechanics has reduced time spent on product yield analysis from hours to minutes, and can perform a multi-dimensional analysis, which was not previously possible. These enhancements to the manufacturing process and reduction in yield analysis times mean Samsung Electro-Mechanics can rapidly and accurately run its business.
Samsung Electro-Mechanics, located in Suwon, Republic of Korea, is a global electronics parts manufacturer exporting more than 80 percent of its production worldwide. Established in 1973, the company operates three business locations in Korea and seven overseas production subsidiaries in Europe, China and Southeast Asia. Samsung Electro-Mechanics’ goal is to rank in the top three companies globally in the parts industry in 2010. As part of its efforts to realize this goal, Samsung Electro-Mechanics was integrating Business Intelligence (BI) applications into its production processes to increase the quality of its products.
||With SQL Server 2005 we have reduced the product yield trend analysis time from four hours to less than one minute.
||Jo Kye Woong
Senior Manager and Professional Yield Analyst
The company has complex manufacturing processes needs to eliminate mistakes that can occur during production. Samsung Electro-Mechanics must manage yield as well as increase its degree of accuracy. When errors occurred, the company needed to rapidly identify the problem areas. As a result, Samsung Electro-Mechanics focused on BI as a solution to resolve its issues of speed and accuracy.
The BI application project began with a yield analysis center being built and field engineers gathering data accumulated in manufacturing to systemize the analysis process.
In 2007, Samsung Electro-Mechanics deployed online analytical processing (OLAP) tools and tried to implement a yield analysis system that could be applied to the production of one product. However, the company realized that this solution did not meet its expectations. Park Chan Hyung, Senior Engineer, Manufacturing Engineering R&D Institute, Samsung Electro-Mechanics, explains, “We realized through experience that there are limits to having tools that are difficult to use by field managers, particularly when we need data analysis to actually enhance yield.” He adds, “For BI to work effectively, common consent among users is required on why it is necessary and how to use it. When performing tasks using existing BI solutions, we spent more time providing training on using the tools instead of building this understanding.”
This recognition led to a decision to improve the existing BI environment. At the end of 2007, Samsung Electro-Mechanics set new goals. The company wanted to increase scalability and flexibility, while enabling an open approach and increasing employee usability.
The company reviewed several technology solutions and selected Microsoft® SQL Server® 2005. In 2008, Samsung Electro-Mechanics began the project for ‘Y-Cuby Development’, its integrated yield analysis system. The company chose to work with Inbrein, a Microsoft Korea BI partner.
Microsoft® SQL Server® 2005 Integration Services and Microsoft® SQL Server® 2005 Analysis Services were used for data extraction, changing and manipulation of data from backbone systems located in eight business sites of Samsung Electro-Mechanics, as well as for analysis and reporting. The system also used Microsoft Excel® 2007in Microsoft® Office SharePoint® Server 2007 to provide users with easy access to various statistics and analysis critical to the business.
In the next stage, the project team added modularized professional analysis including search and statistics for manufacturing engineers and mining professionals.
Park comments “In the case of Samsung Electro-Mechanics, all products have their own specific manufacturing process, so rather than having an enterprise wide integrated BI environment, we required a BI environment focused on the product units. In 2007, we reconstructed the BI framework based on Microsoft technology for the 32 product families that had already been targeted for yield enhancement using BI. In 2008, we applied the BI framework to the production of two products.”
Building Y-Cuby for each product was a complex process, as interfaces were required with several other systems. With the help of Microsoft Korea, Samsung Electro-Mechanics was able to complete its BI project. Park says, “Microsoft provided a BI diagnostic service for a three month period, which enabled us to check standards and determine the direction of the system implementation.”
The Y-Cuby system, based on SQL Server 2005, was implemented by the end of 2008 and has been used for the production of three products, producing key benefits for the company.
Reduced Yield Analysis Time
The company reports significant reductions in time taken to perform various complex analyses. Jo Kye Woong and Jong Yoon Jun, Senior Managers, Samsung Electro-Mechanics, working as professional yield analysts, explain “In the past, when deterioration of yield occurs we had to search legacy system data from all production sites linked with the manufacturing of the relevant product in order to gather data, and then manually perform the analysis. As a result, all analysis processes took a long time.
“With SQL Server 2005 we have reduced the product yield trend analysis time from four hours to less than one minute. In the case of single manufacturing process source analysis, the time was reduced from 12 hours to 30 minutes while the deterioration yield decline source analysis was reduced from 37 hours to eight hours.” Samsung Electro-Mechanics can rapidly and accurately determine the issues in its manufacturing processes using its SQL Server 2005 based solution.
Improved Quality of Analysis
The company notes significant improvements in the quality of their analysis since implementing the Microsoft BI solution. “Samsung Electro-Mechanics was able to easily integrate information held in legacy systems of several production sites according to the flow of logistics, as well as extract and fabricate data related to long term and mass production. This means Samsung Electro-Mechanics can perform multi-dimensional relationship analysis,” comments Park. “As a result, Samsung Electro-Mechanics can easily perform an analysis of complicated problems, including chronic defect reasoning, which was difficult to analyze in the past,” he adds.
Improved Manufacturing Processes
Samsung Electro-Mechanics expects to see further benefits realized from increased BI measurement, specifically in yield management. A greater range of users are now able to perform analyses, which means more people can participate in the decision-making process for improving manufacturing.
Jo says, “It was difficult to train users in the professional OLAP tools it, but with Microsoft BI framework everyone can easily run statistics and analysis. Field staff with responsibilities in manufacturing, production process and quality can always view and analyze data related to their products.”
“If a yield issue occurs, field managers can immediately determine and resolve the issue. Moreover, the manufacturing staff can perform multi-dimensional analysis to achieve results that were not possible in the past,” Jo adds. With field engineers performing manufacturing history searches and basic statistical analyses, the yield analysis team can concentrate on their primary roles as professional analysts and leading the manufacturing field.
Established model for expansion
Samsung Electro-Mechanics plans to expand the application of Microsoft BI framework to other products in the future, to increase its yield enhancement levels. Park says, “We are building a model framework based on best practices and optimizing it according to products and manufacturing fields.”