CATEGORY: Data Mining and Visualization

REVIEWER: Barbara Manville, director of corporate software development for Photronics, Inc.

BACKGROUND: Photronics is a leading worldwide manufacturer of photomasks, high-precision quartz plates that contain microscopic images of electronic circuits. A key element in the manufacture of semiconductors, photomasks are used to transfer circuit patterns onto semiconductor wafers during the fabrication of integrated circuits. Photomasks are produced in accordance with circuit designs provided by customers at strategically located manufacturing facilities in Asia, Europe and North America.

PLATFORMS: We are using an Oracle database which is on an Intel/Solaris platform for the back-end data storage. For the user front end, we are using Windows 98 clients with a networked install of the software to minimize future administrative overhead.

PROBLEM SOLVED: Because of the high-precision nature of the semiconductor industry, it is important that we control our products and processes to ensure that we ship quality photomasks to our customers. By standardizing our statistical analysis tools, we are able to share gains in the quality improvement process between our manufacturing sites. With today's software-driven systems, quality and productivity benefits can be quickly realized with a well conceived statistical process control (SPC) implementation. Additionally, SPC implementation is a prerequisite for QS9000 certification.

PRODUCT FUNCTIONALITY: At Photronics, the STATISTICA Enterprise-Wide SPC System (SEWSS) implementation has been coupled with a rigorous SPC training program for the quality and engineering organizations. This training focuses on using the comprehensive analytical tools included in the STATISTICA Analysis Pack, which includes hundreds of statistical analyses and graphs. In the future, we will standardize our approach to Design of Experiment (DOE) and Failure Mode and Effect Analysis (FMEA) by providing additional training for the Analysis Pack (APACK) modules. We are also looking at the Automated Data Collection (ADC) module of SEWSS.

STRENGTHS: The scalability and flexibility of SEWSS make it suitable for almost any manufacturing environment. We are running this product on an Oracle database at six sites internationally and have plans to have six more sites online by March of 2001. This product could easily run on any other relational database platform. During our selection process, it was important to our engineering community that they be able to set up charts with data from our existing MES system; SEWSS provides the ability to use either an internal or external data source.

WEAKNESSES: When using an external data source, we noticed that some functionality in the product is not quite mature at this time. In particular, we are struggling with recording out-of-control action plan (OCAP) results using the write back to the database feature. We are working with StatSoft to make this feature more elegant. Also, we implemented a SEWSS database at each of our manufacturing sites to optimize operator performance. At this time, there is no utility to easily share a control chart defined at one site with another site.

SELECTION CRITERIA: We selected SEWSS because of its ease of integration with our existing manufacturing execution system and flexibility beyond our existing systems. A cross-functional team developed a matrix of selection criteria, conducted a thorough investigation of several SPC packages and used a weighted voting process to choose SEWSS.

DELIVERABLES: Our team decided to create a limited number of charts for operations to use initially and provided training for the engineers so that they could conduct their own analysis to determine what "critical-to-quality" characteristics should be monitored.

VENDOR SUPPORT: We have developed a strong working relationship with our sales and technical resources at StatSoft. During our implementation, StatSoft representatives worked closely with us to resolve issues specific to our organization. The best part of working with StatSoft has been the expertise of the technical support staff who are well versed in both statistics and the inner workings of the package.

DOCUMENTATION: The introductory classroom materials get the user off to a good start with the basics. The users can easily find answers to their more advanced questions using the online help which is extremely thorough and intuitive.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access