Statistical Software Supports Six Sigma at Dow Chemical
JMP for Six Sigma
Few companies are comfortable
with product defects; companies striving to maintain Six
Sigma production levels abhor them. Dow Chemical Co. belongs
to the latter group of businesses--a devout Six Sigma company
dedicated to streamlined processes and high-quality products.
It's understandable that Dow subscribes to Six Sigma initiatives
with such ardor; it's much easier to eliminate defects before
they begin than to scrap thousands of gallons of chemicals
due to poor production procedures. Remaining at Six Sigma
production levels is no easy task, even for a powerhouse
like Dow. For help, the company turned to the statistical
quality control software of JMP, a business unit of SAS.
"We were looking for statistical software that would
meet the needs of 95 percent of the people 95 percent of
the time," says Dan Obermiller, Master Black Belt of
corporate R&D Six Sigma at Dow. "The product needed
to work under Windows and be affordable." Dow conducted
an extensive evaluation of numerous statistical software
packages, focusing on typical statistical techniques as
well as advanced procedures. Dow also rated the software
on ease of use, accuracy, documentation, expandability,
customizability and price. At the end of the analysis, JMP
emerged as the best product to fulfill Dow's requirements.
Since its adoption at Dow, JMP statistical software has
become the stand-ardized statistical software throughout
the company. "Every department at Dow uses JMP,"
says Obermiller. "It's our global standard for statistical
software."
JMP statistical software links statistics with graphics,
allowing interactive exploration, visualization and ultimately
comprehension of data. In addition to traditional statistical
tools, JMP software includes design of experiments and statistical
quality control, elements that make for a comfortable fit
with Dow. These attributes coupled with a gently sloping
learning curve have made JMP software ubiquitous at Dow.
"A large population of Dow employees are familiar with
JMP and how it works," says Obermiller. "This
has raised the level of statistical awareness throughout
the company and provided a common language related to data
analysis." As evidence of Dow's extensive use of JMP
software, Obermiller points to a set of data miners who
use JMP's 5.0 partition models, PLS and neural networks
relentlessly. To some at Dow, the benefits of JMP statistical
software were manifested long ago. "Dow's R&D effort
has been utilizing JMP's design of experiments capabilities
for about 12 years," notes Obermiller.
Although JMP has provided Dow employees with a common
spoken language for data analysis, the software has also
garnered praise for its scripting language, which allows
users to record, repeat, program, automate and customize
tasks. The structure of the scripting language enables skilled
users to bundle multiple analytics into a template before
distributing scripts to be run by less experienced operators.
This multitiered utilization of the software makes it attractive
to large companies where there may be a wide spectrum of
employee proficiency levels.
By providing applications that are both functional and
user-friendly, JMP has helped Dow assert itself well within
the Six Sigma specification for product quality. "It
has raised our expectations of how one should look at data,"
asserts Obermiller. "JMP is our primary analysis tool
for our Six Sigma programs."
JMP for Six Sigma
- Statistical results linked to graphic representations
- Analytics support all stages of the DMAIC design process
- DOE custom designer
- Available for Windows or Macintosh operating systems
www.jmp.com |
