The use of Quality Analyst software from NorthWest Analytics has helped glass container maker Saint-Gobain chart variables and make comparisons in order to implement an effective and efficient quality programme
Saint-Gobain Containers, formerly Ball-Foster Glass Container, makes between 1000-1500 different glass container products across its 18 US plants. When the company wanted to implement a company-wide quality data management programme at the plants, it chose to standardise on NWA Quality Analyst software from Northwest Analytics. The software is available in the UK from Adept Scientific.
“Our quality has always been very good, but to be the best in this marketplace, we wanted even higher levels of efficiency, productivity and quality,” said Mike Puhl, corporate senior process engineer. “Standardising on NWA QA has created efficiencies and easy comparisons.”
Inside the plants sand, soda ash and limestone are mixed with recycled glass and fed into a furnace, which melts the glass at 2,700°F over about 24 hours. After going through a temperature conditioning unit, the glass is cut into pieces that are the exact weight of the finished jar or bottle. The pieces are then dropped into a forming machine and transferred onto a conveyor for hot-end coating. Then a cold-end coating is applied and the finished bottles are inspected for multiple criteria such as thickness, cracks and bubbles.
During the process, NWA Quality Analyst is used to chart variables such as temperature in the hot-end coating hood, temperature of glass at cold-end coating, flow rates and furnace temperatures. If the control charts show data points trending outside the upper and lower control limits, mid-course corrections can be made before defects are produced.
“We focus on key characteristics of large-volume products,” said Puhl. “That’s where the quality programme can make the most difference.”
Because NWA QA is off-line software, St. Gobain also uses it to see how processes perform over time. The software makes it easy to compare data collected over days, weeks, months and even years, allowing them to spot large trends and unusual patterns that might otherwise be missed in the day-to-day views. They also use the software for Pareto analysis of customer complaints, sorting them by defect types to help prioritise quality efforts.
One of the plants, in Pevely, MO, manufactures over three million beer bottles per day. One way they use NWA QA is to monitor and extend the life of bottle moulds. Moulds break down over time and need retooling or replacing at a cost of between $50,000 to $60,000. Replacing a mould too early – before it is necessary – reduces capacity and increases costs. Replacing the mould too late causes product scrap and rejects.
By control charting container capacity measurements, the plant can determine more exactly when to make process adjustments to keep the product on target and extend mould life. “If a product is budgeted for four moulds in a year and we only use three, that $50,000 goes straight to the bottom line,” said Paul Delaney, plant QC manager. “And if you consider that each machine produces 20 to 40 products, each with their own set of moulds, you can see there is great potential for savings.”
The plant also uses NWA QA to track yields (how many bottles make it into final inspection from production). They monitor a production shop’s 30-day daily performance against its per cent pack (yield) target and review reasons for deviations. Using the software in their daily production meetings allows the whole team to review shop performance data, calculating and projecting control charts, histograms and statistics on a screen for everyone to see.
Delaney says pursuing SQC with the NWA software has many benefits:
- It allows for comparisons between areas, shops, tools and products.
- It has removed the subjectivity from describing shop performance; there is more focus on numbers and trends.
- Trends are visible; they are able to see problems developing earlier while there’s time to prevent losses.
- Having clear data stops people from saying “we can’t do that” when the data shows they can, and “we can do that” when the data shows they can’t.
- The charts and graphs allow management to set goals based on data and tighten up performance. Goals are based on control and capability rather than on just meeting specifications.
For Delaney, the flexibility and ease-of-use of the software are what matter most. From the corporate point of view, Puhl agrees, but he also highlights the importance of the quality and consistency of the software company’s customer support. “Whenever I call I receive advice from a knowledgeable human being that already understands the issues I’m experiencing,” he said. “That’s invaluable.”
Adept Scientific
