Page 238 - How To Implement Lean Manufacturing

216 Cha pte r F o u r tee n

In this process line, there was a solder “touch-up” station—actually, the plant had several,
including the rework stations. Here, an operator would make minor repairs using a solder
iron. This station had a high fallout rate. To better understand the process, a DOE (Designs of
Experiments) was performed and it was found that the solder quality and solder iron tem-
perature were the two key process characteristics. The proper solder was purchased and the
solder iron temperature was monitored and placed on SPC. An Xbar-R chart was used. Sub-
groups of three were gathered every hour and plotted. The chart quickly showed the appear-
ance of a special cause after about 20 hours of operation. A quick analysis determined that flux
and residue had been built up on the iron and additional mechanical cleaning was instituted.
It was made part of the standard operating procedure and at the start of each shift the opera-
tors all cleaned their solder irons. Also, at this time the sampling frequency was changed from
hourly to every two hours.
Some minor problems were encountered, but FTY increased by about 4 percent at
this station. (Please recall that one station was at 88 percent FTY—this was it.) Over the
next two months, sampling was changed to every four hours and yield climbed another
2 percent, though we are not exactly sure why—I have always suspected it was due to
the Hawthorne Effect.

The Hawthorne Effect
At the Western Electric plant in Cicero, Illinois, from 1927 to 1932, studies were done
on workplace conditions and worker performance. In one series of tests, it was
(Please recall that one station was at 88 percent FTY—this was it.) Over the ext two
hypothesized that worker productivity would improve if workplace lighting was
months, sampling was changed to every four hours and yield climbed another 2 per-
improved. They improved the lighting and, not surprisingly, worker productivity
cent, though we are not exactly sure why—I have always suspected it was due to the
rose. Someone questioned the validity of this testing, so they performed another test.
Hawthorne Effect.
In this one, they told the workers that the test was such a success that they were going
to increase the lighting even further. They made their changes and worker productiv-
ity rose once again. However, instead of raising the lighting levels they actually had
lowered them; leading the researchers to conclude that there were other dynamics to
consider. This and other experiments led them to develop what has become known
as the “Hawthorne Effect.” One definition of the Hawthorne Effect is:
People singled out for a study of any kind may improve their performance or behavior,
not because of any specific condition being tested, but simply because of all the attention
they receive.

But then we noticed a common occurrence on
many of the SPC charts. Somewhere between 45 and
90 days of operation, the process would go out of con-
“Cease dependence on
trol. Upon investigation, we found that the solder
ins pection to achieve quality.
irons just wore out. It seemed the tips would change
Eliminate the need for inspec- metallurgically and not be able to hold the tempera-
tion on a mass basis by build- ture. So we implemented a program to change out the
ing quality into the product in tips. This was done to all solder irons during the
”
the first place. monthly plant preventive maintenance (PM). The tips
—(Point 3) were only a few dollars a piece and even though we
W. E. Deming changed them out prematurely, we did not want to
scrap even one electronic control unit (ECU) since
each was worth about $65. Shortly after this process change, the solder iron SPC changed