Page 215 - How To Implement Lean Manufacturing

Constraint Management 193

early gains that were achieved through breaking the system constraints. To a large
extent, this was how they “harvested so much low-hanging fruit.”

When Is It Best to Address Constraints?
Very often, during the review of the foundational issues, constraints will be found and
highlighted. It is also, in this timeframe, that the constraints should be aggressively
attacked and removed. First, it will make money for the facility and you will be able to
accrue huge early gains. And second, it is more efficient to achieve the appropriate pro-
cess flow rate, stabilize the flow, and then work on quantity control techniques like
kanban. However, everyone should constantly have an eye out for constraints and
should be attuned to removing them. After all, once the system constraint is broken, the
system’s profitability is vastly improved.

How Do We Spot the System Constraint?
This is the simplest of questions. Look for the inventory build-up. It will always be in
front of the constraint.
What if there is inventory between all process steps? It will then be the largest pile
of inventory. What if there is inventory galore, as is so often the case? Then do a process
study to calculate the time for each step. The step with the longest cycle time is usually
the constraint. I say usually, not always. It is possible that a given process step may not
have the longest average cycle time, but if it has huge variations in the cycle time, this
step may be the system constraint.
Often when a multistep process has a lot of WIP (work in process), it may experi-
ence a broad array of problems. One of which is that it may operate at a cycle time that
is well above its potential, thus producing at a rate well below its design. This is espe-
cially true if it is dominated by manual operations. In this case, unless there is some
clear way to measure and control the cycle time, the system constraint may bounce
from one station to the next and the whole line will under-produce. This is common-
place (see Chap. 16). Consequently, at the end of the day when the production quota
was not met, no one can point to any specific problems that the line experienced. Often,
what follows is some of the most elaborate, but incorrect, rationales for why the pro-
duction goal was not met.
The answer, of course, is first understand and reduce the variation and then get rid
of the WIP inventory so you can “see” the process. This will allow you to find the con-
straint and work on it—if necessary.
Almost always, once the inventory is removed, the entire process, often every step,
will speed up. To most people’s amazement, the quality will improve as well. It is not
logical to most that this should occur—that is, a
simultaneous improvement in both quality and
rate. However, it is a metaphysical truth, and also a Point of Clarity When we
very common occurrence, that production will rationalize, we generally just
speed up and quality will improve once the inven- create “rational lies.”
tory is eliminated.

Why Ohno Does Not Even Talk about Constraints
If this constraint stuff is so important, why is there nothing of it in either Ohno’s writ-
ings or those of Shingo? There are two explanations. First, Ohno and Shingo are so good