|
|
Minitab Case Study - Analyzing Data for a Delicate Finish: Six Sigma Ranch,
Vineyards, and Winery
Some people take it easy when they retire. But Kaj Ahlmann, retired chairman,
president and CEO of General Electric’s Employers Reinsurance division, is just
getting started. In 1999, he united two of his passions—wine and statistics—in a
new venture: the Six Sigma Ranch, Vineyards and Winery. This pairing may seem
unlikely. “Winemaking” conjures images of pastoral fields of lush grapes, grown
by experts who divine the ideal time to pick and ferment them. Six Sigma, on the
other hand, is frequently associated with mass production, and massaging reams
of data to wring maximum efficiency from the factory. When asked how the
romantic art of winemaking can coexist with data-driven decision-making, Ahlmann
suggests that his passions complement each other perfectly. Six Sigma theory
posits that great products are informed by the voice of the customer, while
great winemakers seek to ferment a wine that delights the drinker. Why not use
Six Sigma to create great wines that answer the voice of the customer? “We
aren’t trying to take the romance out of winemaking— we’re trying to improve the
wine,” he says. “Six Sigma helps us focus on factors that contribute to the
wine’s quality and what customers want.” And when it comes to analyzing data to
optimize those factors, Minitab Statistical Software plays a critical role.
The Challenge
Making great wine isn’t easy. Vintners must carefully consider soils, grape
varieties and ripeness, the vessel the wine is aged in, and many more factors.
They must try to make the wine consistent within a batch or vintage year to
satisfy the customer’s desire to get the same flavor with every bottle. Even
slight variations can result in an unacceptable product(s)—and when something isn’t
right, a connoisseur’s taste buds can detect it immediately. Some vintners
appear to be able to create great vintages by relying on intuition. As romantic
as that approach might be, Ahlmann sees clear advantages in applying Six Sigma
to the many variables involved in the process of turning grapes into something
more: he and his team harness science, statistics, and their senses to create
fine wine. While the vineyard and winery offer a different environment than many
Six Sigma practicioners would encounter on a daily basis, Ahlmann and his team
gather data in ways any quality improvement professional would recognize. For
example, they monitor critical-to-quality attributes within the wine
fermentation vats, including temperature, alcohol level, acidity, and malolactic
fermentation, among others. They also apply the DMAIC strategy to eliminate
variability in the vinification process. “We decide where we want to be, then we
make sure that our process ensures that our wines meet that level of quality,”
Ahlmann says. “We emphasize consistency, and eliminate variation where it would
affect quality.” Grape sorting is just one of many places where Six Sigma
tools—and Minitab—have made a difference in how the wine gets made. The winery’s
vineyards are typically harvested over a two-month period. Bad grapes must be
removed before crushing and fermentation.
It takes 5 to 15 labor-intensive hours for employees to sort the grapes from
each day’s harvest. First, the grapes are placed in small bins, which makes it
easier for sorters to spot and remove entire clusters that might be bad.
Clusters that pass this test move to the sorting table, where immature grapes
are shaken out through a grid and discarded, leaving only mature grapes, which
then face a final round of hand-sorting. The Six Sigma Ranch wanted to be sure
their sorters, who return for each year’s harvest season, all knew the
difference between good and bad grapes. Ahlmann and his winemakers began by
defining specific standards for high-quality grapes. By randomly tasting grapes
before harvest, they could identify characteristics that indicate flavor in the
skin of the grape, whether the seeds have become nutty or are still bitter, and
other critical factors. “An expert winemaker’s palette is his measuring
tool—even if we don’t usually think of it that way,” Ahlmann notes. “Factors
like these may not seem like hard data points on the surface, but you can get
data around them.”
How Minitab Helped
Once they clearly established criteria for sorting grapes, the winery had a
solid foundation from which to make sure all sorters followed the same
procedures to make sound pass/fail decisions at the sorting table. Tanuj
Pasricha and Akashdeep Khera, two Six Sigma Black Belts working as consultants
with the winery, used Minitab to conduct a measurement systems analysis of the
sorting process. To measure the sorters’ consistency in recognizing grape
quality, they used Attribute Agreement Analysis to assess the consistency and
correctness of sorters’ ratings against the standards. By conducting their
analysis and evaluating the results, they were able to effectively train all
operators to ensure an acceptable level of agreement. But sorting grapes is just
one component of the winemaking process. Wherever they collect data, the winery
turns to Minitab Statistical Software for tools like descriptive statistics,
graphical analysis, control charts, and capability studies. For example,
Minitab’s Individuals and Moving Range (I-MR) control charts have proven to be a
particularly powerful tool in assessing whether the winery’s processes are in
control, and are used to monitor alcohol and acidity levels.
All processes
exhibit some natural variation, and control charts help you distinguish this
natural variation from the non-random, “special cause” variation that can be
detrimental. Minitab’s I-MR charts help the winery identify problematic
variation so they can provide customers with consistent wine taste and quality.
The charts also help reveal patterns that may indicate quality improvement
opportunities, such as shifts at the beginning or end of the winemaking process.
To ensure their processes are producing quality wine within acceptable limits,
the winery turns to Minitab’s capability analysis tools. These studies help the
team ensure that fermentation and other processes under way at the winery are
capable of producing results that meet the specifications they’ve set. “These
charts easily show us the picture of how well we are doing,” Pasricha says. “We
also can use this data to make improvements in the next wine-making cycle.”
Another aspect of Minitab Statistical Software that has been tremendously useful
is the Help system, which features detailed explanations of statistical
techniques, explains how to perform them, and even helps users interpret
results. “I usually refer to Minitab’s Help menu as the ‘Google of Statistical
Tools’ because we use it so often,” says Tanuj Results Ahlmann and his
colleagues now are looking at additional aspects of production, trying to define
standards for all of the the critical-to-quality parameters that must be
controlled at each step, then making sure their processes meet those standards.
In keeping with the commitment to quality implied in its name, all of the Six
Sigma Ranch, Vineyard, and Winery’s processes are in control and capable of
delivering within specification limits. But that doesn’t mean Ahlmann’s quest to
produce great wine is complete—far from it.
Additional projects are under way, and the winery team continues to use Minitab
to reveal the meaning of the data they collect. They’re even using Minitab’s DOE
(Design of Experiments) functionality to efficiently assess promotional
campaigns and to redesign the company Web site. Through it all, Ahlmann’s focus
remains squarely on the most important person in the winemaking process: the
person who drinks it. Six Sigma Ranch, Vineyards and Winery continues to learn
about customers wants by soliciting feedback and conducting voice of the
customer projects. “That’s what Six Sigma is really about,” Ahlmann says. “We
want to understand what attracts a potential customer and adds the most value,
and then deliver it.” |
|