Page 36 - How To Implement Lean Manufacturing

Lean Manufacturing and the Toyota Pr oduction System 17

Compare all of these efforts to supply quality to the Lean facility whose primary
job, like the MassProd model, is to protect the customer. However, quality management
runs much deeper. First and foremost, rather than just inspect and sort the product, the
process is scrutinized deeply so it will improve and the need for both process and final
inspection will be dramatically reduced. The amount of emphasis placed on process
control in the Lean facility is staggering when compared to a typical non-Lean plant.
The result is that the Lean plant develops much more robust processes with more stable
cycle times and improved process and product quality. This isn’t rocket science, simply
good old-fashioned hard work with a deep-seated belief in process management.
In the Lean plant, the primary purpose of inspection data, which is largely variables
data rather than attribute data, is for problem solving. Real problem solving is deep and
practiced by all. Jidoka is used in all steps of the manufacturing process, not just at final
test. The underlying principle of jidoka is that no bad parts are allowed to progress in
the production cycle; even if it means shutting down the production line and ceasing
production until the root cause of the problem is found and removed. Using this tech-
nique, it is clear that quality is equal to production. The second underlying principle of
jidoka is that it is a continuous improvement process. Poka-yoke are used extensively and
quality problems become immediately obvious through the use of andons and other
forms of operational transparency. The uniqueness of Lean, in the area of quality, lies in
four areas. Note that the first three of these distinctions are cultural, rather than technical:

• First, a quality problem is not just a reject, it is failure of the system, which is
owned by all.
• Second, the quality problem is not bad news, rather it is often good news,
signaling a weakness that can now be understood and corrected, thus leading
to a more robust system, rather than be ignored and forgotten only to reappear
again.
• Third, everyone participates in the technical solutions to problem solving.
• Fourth, and finally, the system uses a system of tools such as poka-yokes to attain
100 percent inspection.

How Cycle Time Variations Are Managed
In MassProd, cycle time variations are not considered a problem. They are seldom quan-
tified, hence they are largely unknown and ignored. Average cycle times are understood,
but to maintain average rates, large volumes of inventory are held between stations. As
long as average cycle time is maintained, the variations do not affect the overall produc-
tion, but only at the cost of huge inventory volumes (for an example of this, fast-forward
to Chap. 18 and perform the Dice Experiment). Frequently, without understanding the
damaging effects of cycle time variations, efforts will be made to reduce inventories. This
is almost always met with a significant drop in overall production rates.
It is extremely common in MassProd facilities that even when individual stations
perform at design rates, on average, the overall lines do not perform at the design rate.
This is caused by the interaction of variation and dependent events that occur on the
line. (This is completely explained by the experiment in Chap. 18.) In the Lean facility
there is always a significant effort to optimize the cycle times, including significant
efforts to reduce the variation in the cycle times. In the Lean facility, both the average
and the variation of the cycle times, are known, understood, and managed.