Page 169 - Six Sigma Demystified

150 Six SigMa DemystifieD

item 3, regarding the prevalence of human error, suggests benefits in
mistake- proofing the process. Controlling human behavior by other means, such
as written procedures, often produces less than desirable results (as is evident in
this example). in the case of software data entry, error prevention is often a matter
of automating data entry to remove or reduce human influence or adding
prompts, warnings, and/or required fields into the data- entry forms. Unfortu-
nately, the current process uses a mass- market customer relationship manage-
ment (CRM) system for storing these data that does not provide data- entry
customization.
item 4 had particular resonance among team members: if the Web order data
are available to order- entry personnel, the additional data entry is limited to the
license codes. The license codes are generated from internal software, which
could be easily modified to write to the Web- order database. if all orders could be
entered directly into the Web- order database, then order entry into the CRM sys-
tem could cease immediately.
The team developed a revised flow assuming use of the Web- order database
and received the sponsor’s approval for information technology (iT) support to
implement these changes to the Web- order database.

Simulations

When a process model is known, either from experimentation or by design,
then process simulations can be used to find optimal solutions. While a regres-
sion model is deterministic in nature, in that the factors in the model are fixed
values, factors in processes are random variables. Simulations provide a means
of using probability estimates for the random variables to discover the impact
of their joint probabilities.

Simulations have several advantages over experimentation. They certainly
are cheaper in almost all cases, allowing the test of many more conditions than
is practically possible with experiments. This makes them very suitable for
“what- if” scenarios to test the response under worst- case circumstances. Simu-
lations also can be used for planning, such as in a design for Six Sigma (DFSS)
approach to estimate the effects of increased business activity on resources or
for new- product processing.
Key uses of simulations include (Pyzdek and Keller, 2009)

• Verifying analytical solutions
• Studying dynamic situations
• Determining significant components and variables in a complex system