February 2008 Edition
Just how good is your process?
Check your O.E.E. and find out
By Richard Clark
There is no better statistical tool to use when
evaluating the all-around efficiency of a production
process than Overall Equipment Effectiveness. O.E.E.
is a simple series of formulas that be can used with
any calculator and some basic data that are most
likely currently at your disposal. There are 3
factors within the production process that are
evaluated using O.E.E.
• Availability: The actual uptime of the
process divided by the scheduled available runtime.
Example: The scheduled runtime was 7.2 hours (432
minutes). A broken parts feeder stopped production
for 45 minutes. The uptime for the shift was 387
minutes. 387/432 = 0.89583 or 89.6 percent.
Bottom line– Availability represents machine
breakdowns.
• Performance efficiency (P.E.): During
the actual uptime, how efficient was the process
when compared the designed optimum cycle time.
Example: The optimum cycle time for the process is
15 seconds per part or 4 parts per minute. For 387
minutes of uptime, the process would produce 1548
parts if it never stopped once. During our shift, we
produced 1357 parts. 1357/1548 = 0.87661 or 87.7
percent.
Bottom line- Performance efficiency represents
short stoppages of the process.
• Rate of quality product (R.O.Q.P): Of
the total number of parts produced during the
uptime, what percentage was conforming. Example:
During our 387 minutes in which we produced 1357
parts, 14 were defective and 12 need re-worked. 1331
conforming parts were made. 1331/1357 = 0.9808 or
98.1 percent.
Bottom line- Lower your in-process scrap and
rework.
The Overall Equipment Effectiveness is computed
by the formula – Availability x P.E. x R.O.Q.P.
0.896 x 0.877 x 0.981 equals 77.1 percent, which
is our O.E.E. for this process.
A "World Class" process would produce a
consistent 85 percent O.E.E. (or an average of 95
percent in each category).
There are some simple items to look at for an
O.E.E. improvement. Availability deals with
scheduled uptime. Be sure routine items such as
machine changeovers, preventive maintenance, and
department meetings are scheduled through the
production controller (however titled) at your
facility.
Improving P.E. means correcting a short stoppage
before the cycle time is lost. Never lose sight of
the parts counter at the end of the process. Some
processes are made up of smaller stations with their
own unique cycle times. The stations form a chain
and the optimum cycle time of the process will be
the cycle time of the slowest station. This is
beneficial because it can be determined which
stations are capable of "catching up" after a
stoppage. If more than one stoppage occurs at the
same time, it may be best to correct them from the
closest to the end of the process and then work
backwards towards the beginning.
This goes against conventional thinking. If a
stoppage occurs down toward the end of the process
and at the same time a machine jams up right in
front of you (while you're at the front of the
process), it may be better to walk down to clear the
other stoppage first. The reason behind this is if
the stoppage down the line is not fixed first, it is
possible the last process will run out of parts.
Each and every 15 seconds is lost forever. By fixing
the stoppages toward the end of the line first, you
have a better chance of parts "catching up" and
there will be no cycle time lost at the last
process.
In some situations any stoppage will shut the
entire process down. In that case I would suggest
placing re-settable trip counters at various
intervals. When a stoppage occurs, the operator can
trip the counter. After a week of production, the
counters will tell you where the most stoppages are
occurring and further evaluation can be planned.
R.O.Q.P. is too product-specific to be discussed
in the article. If your facility doesn't have an
organized quality system, I suggest developing one.
Remember that no system will be any better than
the operators who use it. Machines make parts, but
people make products. Keep an open mind and always,
always, always be creative.
Richard Clark is a consultant from Portland,
IN. For more information about statistical process
control, contact him by e-mail at
rcmetrology@ yahoo.com