Page 220 - How To Implement Lean Manufacturing

198 Cha pte r T h i r tee n

For this corporation, new production lines were designed by the home office engi-
neering staff and then sent to the facility to debug them and bring everything up to
speed. Consequently, this facility did not have a full engineering staff.
This production line was a straight flow line with a conveyor. The conveyor would
advance and then remain stopped for the 14-second work cycle time; the transportation
time between work stations was two seconds. The appliances were mounted on rotat-
ing tables to facilitate construction, and the tables were fixed to the conveyor. There was
no new technology in this line, but the design cycle time was considerably shorter than
prior designs. It stood out in the facility as being different and placed significant pres-
sure on everyone, especially the materials delivery staff. But the demand for this prod-
uct was high and the management wanted to only invest in one set of equipment—hence,
one high-velocity line.
Earlier, we had done significant training in statistical problem solving at the facility
and conducted two waves of Greenbelt training, as well as one wave of Blackbelt train-
ing. Following this, Greg, the general manager asked us to review the operation of the
Gamma Line. He said he was in a hurry and wanted us to evaluate why the labor effi-
ciency was so low. Labor efficiency was one of their most important plant metrics.
Greg explained the situation. The line had been placed in service over three months
before and had never achieved design rates. As the demand ramped up, they had to
schedule Saturdays and even some Sundays to meet demand. That was why the labor
efficiency was low. In addition, the union was becoming a significant obstacle. From the
beginning, the union was against the design. The shop steward claimed the short cycle
times placed too much pressure on the workers. Greg had the same concern.
We were not familiar with the metric of labor efficiency, so we asked about this
metric and the 0.85 standard. It was explained thusly: It is the allotted labor, which is
based on the cycle time and design line staffing, divided by the actual hours worked for
all the hourly workers. Each line is calculated based on only their direct labor head-
count for those on the line, including those not working. The 0.85 factor was to account
for labor that was scheduled and working but not producing due to machinery failures
or anything that would cause production to be reduced. We asked if the production
losses included quality dropout, stock outs, machine downtime, and not performing to
cycle time. They said that was the concept and the minimum standard was 0.85. Any-
thing over 0.85 (or 85 percent) was gravy. On this line, the best they had done was just
recently when they struggled to get to 70 percent.
Without knowing it, their labor efficiency was a type of OEE. Not as powerful and
not as usable as OEE, but a very similar concept. The unfortunate part of this metric was
twofold. First, their objective was to reach 85 percent. If they got there, they would be
happy. There was no strategy to go beyond the 85 percent. Second, this labor efficiency
was affected by all the aspects that affect OEE, such as quality, machine availability,
material supply, and cycle-time stability—to name the major issues. Yet they framed it
as a labor issue. Quite frankly, it was everything but a labor issue.
A problem exists when a facility is managed as only a cost center. Clearly their only
effort was to make sure they did not use too much labor. But as we will see, labor was
not their problem—they had neither begun to understand, nor begun to attack their real
problems. It shall be shown that their true problems were waste and variation—which
is just waste by another name.
Greg went on to explain some more history of the line. He said that upon startup,
unlike prior startups, tremendous scrap was generated, over 30 percent. When they