ZIFFPAGE TITLEFinding An Edge inBy John Moore | Posted 2005-05-23 Email Print
WEBINAR: On-demand webcast
Next-Generation Applications Require the Power and Performance of Next-Generation Workstations REGISTER >
As rival textile producers move their operations overseas to save money, W.L. Gore is staying put in the U.S. Its competitive advantage? Democracy.Process">
Finding An Edge in Process
The trick for Gore is to orchestrate and support the activities of myriad ad hoc groups that have parlayed the company's polymer technology into products such as guitar strings and dental floss. Each of these groups must boost quality and improve production processes in a tough manufacturing market.
About 4 1/2 years ago, Gore set out to address the difficulty of coordinating production across its four divisions: fabric, medical, industrial and electronics.
"A group of us wanted to make the life of a process engineer easier," says Jose Ramirez, an industrial statistician at Gore.
Until a couple of years ago, process engineers used an array of Excel spreadsheets and process control products to keep Gore humming. Engineers, Ramirez says, "were trained in all the tools of statistical process control and capability analysis, but we wanted to create a common denominator for all four divisions and provide a tool that just concentrates on the decision-making process rather than number crunching."
The result: Gore's Process Health Assessment Tool. It uses a custom-built Web service that lets engineers access reports depicting the state of the company's production processes. The tool pulls in production data that factory managers cull from a variety of sources: spreadsheets, SQL Server databases and enterprise resource planning software, among others. The collected data covers the minutiae of manufacturing, down to such details as the temperature at which a given process takes place.
Once the data is extracted, SAS Institute's business analytics software suite performs the "number crunching" Ramirez mentioned. Individual software tools perform functions such as data mining, business process modeling, statistical analysis and quality improvement. The software generates reports that help Gore engineers get a handle on manufacturing operations.
The reports prepared by the Process Health Assessment Tool provide a snapshot of the state of a given processa factory floor pulse-taking. In Gore's textile business, the reports focus on the different process steps in the production of fabric laminates such as Gore-Tex. Lamination, a key process step, involves the gluing of a polymer membrane onto fabric. A sufficient amount of adhesive must be used to ensure adequate bonding, but overuse can decrease the fabric's breathability.
In statistical process control, engineers attempt to manage variations of this kind. Donna Fulenwider, a strategist in SAS' worldwide marketing department, says otherwrinkles can crop up in areas such as manufacturing equipment and even human resources. "There are a lot of nuggets of variables you have to deal with," she says.
Chuck Seipel, a Gore associate in process engineering, says the tool helps point the way toward reducing or eliminating pockets of variability. It diagnoses the effectiveness of processes in two ways: the ability of a process to meet customer specifications, and the predictability of that process.
An inquiry into the first aspect, for example, might determine that excessive use of adhesive causes a given production line to miss the quality mark. Or it could be that the flaw lies in the chemistry of the adhesive. Using the Process Health Assessment Tool, Gore engineers can pinpoint the problem.
"Based on that analysis, we can determine the health of the process and take appropriate actions," Ramirez explains.
Actions that ratchet up quality provide an important edge for textile makers. While Gore enjoyed a first-mover advantage in performance-outerwear fabrics, the company now faces competition from several rivals. "Gore invented the category and set the standards, but the category has been attacked from every direction," notes John Cooley, a spokesman for Marmot, one of Gore-Tex' first customers in the mid- 1970s.
The second aspectpredictabilityinvolves determining whether a process will consistently yield the same result. Predictability has ramifications for the entire supply chain. Planning and logistics become easier to manage when a company has a predictable process, according to Ramirez.
Predictability looms particularly large in the outerwear market. Ski jackets and other seasonal items require supply chain timing that leaves little margin for error.
In Marmot's case, the company tries to secure retailers' orders for the next fall's product line around Nov. 1about the time the current season's goods begin to sell. About 60 to 90 days prior to that date, Marmot must give Gore and other fabric suppliers an idea of what it will need in raw material, Cooley explains. The suppliers agree to set aside a certain amount of greigerolls of unfinished fabric. Marmot then develops a better sense of which colors and fabric finishes will sell. The company places its fabric order with the supplier, generally before Jan. 1. The material ships to a contract manufacturer.
Sewing normally begins by March, and finished goods are shipped in July to Marmot, which in turn ships to retailers in August or September. Cooley emphasizes, however, that the timing is subject to change and varies with suppliers.
Apparel retailers at the end of the supply chain have their own particular time-driven issues. At the retail level, time-to-shelf is the key objective. According to BearingPoint's Kauffeld, retailers that produce a shorter span of time from factory to shelf can keep a broader variety of products available and thus enjoy a stronger competitive posture. "You can be more responsive to shifts in what is hot with consumers," he says.
Seipel notes that on-time delivery is among Gore's overriding metrics. Another is first-pass yieldsthe ability to create fabric that meets customer specifications the first time. Remaking a batch of fabric affects delivery schedules, he adds.
In typical Gore fashion, process engineers from two divisions collaborated on the Process Health Assessment Tool project. Joining them were Gore's enterprise information-technology group, which manages the company's network infrastructure, and technology associates affiliated with Gore's manufacturing centers. At Gore, each plant has its own cadre of information systems specialists.
The system for analyzing the health of processes took about six months to complete, according to Ramirez. Still, it remains a work in progress. "This tool has a life of its own," he says. "Now that it's been in existence for four years and everyone has access to it, people are taking it in different directions."
For example, engineers with access to supplier data now tap the tool to compare suppliers and make decisions on which one to use, he adds.
In addition, Seipel said Gore uses the tool to get greater insight into the raw materials purchased from suppliers and how those materials impact the company's processes. "It helps us articulate our needs to suppliers," he says.
But whether it is used for tweaking internal processes or evaluating suppliers, the Process Health Assessment Tool gives Gore an engineering lingua franca. "The common language is the key," Ramirez says. Each Gore division developed its own lingo over the years, he explains.
"But now, you allow the engineers and scientists to communicate at the same level," Ramirez says. The data they analyze may be vastly different, but the health assessment remains the same, he adds. The engineers measure with the same yardsticks: customer expectation and predictability.
"We have engineers from four different divisions communicating with each other," Ramirez points out. Such dialogue lies at the heart of Gore's democratic ethos.