Page 35 - How To Implement Lean Manufacturing

16 Cha pte r T w o

• Cycle stocks (normal pickup volumes)
• Safety stock (to account for internal variations in production)
• Buffer stocks (to account for external variations in demand)

So, for example, when the customer comes to pick up product in a quantity greater
than the scheduled demand, some of the buffer stock is removed. Once removed, buffer
stock removal rules require that immediate corrective action be implemented (likewise,
corrective action is required for the removal of safety stocks). In addition, by using a
kanban system, once the product is picked up by the customer, the kanban cards are cir-
culated and show up at the heijunka board at the production cell. The cards for buffer
and safety stocks are a different color, which signals there has been a change that
requires some countermeasure by the production cell, such as scheduling some over-
time to replenish the inventory. If the kanban cards are picked up each hour, then in two
hours or less, the signal that something unusual has happened is sent to the pacemaker
process. The cell can then implement some countermeasure for replenishment. If that
information goes to the typical MRP system, a countermeasure may be a full week
away. The Lean solution is both more responsive and relies less on centralized planning
functions for daily operations. It should be noted that for production floor scheduling,
the classic models of MRP, MRPII, SAP, ERP, BRP, SCM and any other generic models
are simply inadequate. Many efforts have been made to make them responsive enough,
but they have internal inadequacies when it comes to shop floor planning. In most cases,
the best system is kanban for triggering production. The inadequacies of these scheduling
tools such as MRP, is a topic outside the scope of this book but your experience has
already told you that your scheduling model will never make you Lean. However, if you
wish to read about this, it is covered well in Factory Physics, (McGraw-Hill, 2008) by Hopp
and Spearman.

Managing Quality in MassProd
How quality is managed in MassProd is changing somewhat. A few years ago bad quality
was only a volume and financial issue. If production rates could not be met due to quality
drop out, it was a crisis, but if rates could be met, then the motivation to reduce defects
was one of finances alone, using short-term economics and immediate effects only.
Since the rework of product was commonplace, this system caused little improvement
in quality.
However, more recently, driven by customer demands for a higher-quality product,
all manufacturing facilities in the supply chain have been affected. Now almost all firms
still in business, at least claim to have a continuous quality improvement philosophy.
However, old habits change slowly and most do a better job of talking about quality
than actually delivering it. Typical quality systems rely heavily on inspection done by
humans and have a high dependence on attribute data. In the end, they generally just
sort out the defective product. Frequently the only place good inspection is done is at
final inspection and test.
These systems are not designed to be used for continuous improvement and the defect
data are employed mainly for yield and maybe quality cost data. Sometimes problem
solving is completely absent; most often it is superficial at best. These facilities are not
staffed for problem solving, and so quality problems not only persist, but frequently even
the ones that were “solved,” rise again. Finally, the attitude toward problems, in the typi-
cal MassProd facility is one of trying to avoid dealing with them.