Welcome to the Adept Scientific Press Room

Statistical Process/Quality Control exploits the Internet

Lexmark International, a laser-printer toner manufacturer, faced challenges common to global manufacturers: consistent product quality. Said Mark Watkins, control engineer for the company’s Boulder, Colo., facility: “With our continuous processes, real-time detection of problems is important to prevent production of large amounts of scrap material.”

The company wanted to achieve the same level of quality and yields in all facilities. Watkins indicated Lexmark needed a way to compare and contrast process and quality data among plants.

Keeping yields as high as possible is another issue. The company needed to meet tight price and product goals. “With manufacturing today, you keep a lot less material on hand, so there’s less room for mistakes,” he remarked.

Information overload!

The Boulder plant starts collecting data for SPC/statistical quality control (SQC) analysis with its distributed control system (DCS). The DCS controls the process and collects data such as temperature, speed, cycle time, and weight. It tracks data continuously, in real time, for every process and product.

In addition to process data, large amounts of manually entered analytical test data, printer test results performed on the finished toner, yield and rates by shift, and tracking data for every bag of toner produced are collected. A Sybase SQL server stores three years’ worth of process and test data, with older data archived on another machine.

The company discovered that with so much data being collected and stored, its greatest challenge was to make it easy for people throughout the organisation to analyse and use the data for quality monitoring, research and development, and process improvement.

The Solution

To serve these data sharing needs, Lexmark built an intranet-based solution. The Windows NT system runs an Apache general Web server and a NWA Quality Analyst Web Server for SPC/SQC charting. Programs for accessing the database and SPC charts were written in Perl.

The system generates the data file from the Sybase database based on user requests. It then builds the necessary Analyst Web server header and data files on the fly and asks that server to generate the SPC chart. The Perl programs connect to the database using open database connectivity (ODBC) Perl libraries and NT Sybase ODBC drivers.

Lexmark chose the Analyst Web server because of its connectivity options, based on an open architecture and ability to be connected to any ODBC database. But the company had to work through a few minor problems with the Analyst Web server. One was system efficiency. Originally, the company configured the server to access the database for every chart displayed, but the arrangement didn’t give it the desired result.

Lexmark wanted to access data once for each user connection, building all the charts in one step. To do this, the company used the same automation program (the Quality Analyst Run File Interpreter) that the Analyst Web server uses—but that caused conflicts. However, Lexmark, with NWA’s assistance, distributed the different automation tasks and avoided the conflicts.

Having successfully implemented Web-based SPC, the company’s challenge in the immediate future is to use those capabilities globally to help facilities share problems encountered, lessons learned, and innovative solutions.

NWA Quality Analyst Web Server is supplied and supported in the UK by Adept Scientific plc, Amor Way, Letchworth, Herts. SG6 1ZA; telephone (01462) 480055, fax (01462) 480213, email quality@adeptscience.co.uk; or see Adept’s World Wide Web site http://www.adeptscience.co.uk/. Adept Scientific is one of the world’s leading suppliers of software and hardware products for research, scientific, engineering and technical applications on desktop computers.

With offices in the UK, USA, Germany and throughout the Nordic region, Adept Scientific is one of the world’s leading suppliers of software and hardware products for research, scientific, engineering and technical applications on desktop computers.