3.
Production
Process Characterization
3.4.
Data Analysis for PPC
3.4.5.

Assessing Process Stability


A process is stable if it has a constant mean and a
constant variance over time

A manufacturing process cannot be released to production until it
has been proven to be stable. Also, we cannot begin to talk about
process capability until we have demonstrated
stability in our process. A process is said to be stable when all of
the response parameters that we use to measure the process have both
constant means and constant variances over time, and also have a
constant distribution. This is equivalent to our earlier definition
of controlled variation.

The graphical tool we use to assess stability
is the scatter plot or the control chart

The graphical tool we use to assess process stability is the
scatter plot. We
collect a sufficient number of independent samples (greater than 100)
from our process over a sufficiently long period of time (this can
be specified in days, hours of processing time or number of parts
processed) and plot them on a scatter plot with sample order on the
xaxis and the sample value on the yaxis. The plot should look like
constant random variation about a constant mean. Sometimes it is
helpful to calculate control
limits and plot them on the scatter plot along with the data. The
two plots in the controlled
variation example are good illustrations of stable and unstable
processes.

Numerically, we assess its stationarity using
the autocorrelation function

Numerically, we evaluate process stability through
a times series analysis concept know as
stationarity.
This is just another way of saying that the process has a constant mean
and a constant variance. The numerical technique used to assess
stationarity is the
autocovariance
function.

Graphical methods usually good enough

Typically, graphical methods are good enough for
evaluating process stability. The numerical methods are generally only
used for modeling purposes.
