Ayres cites Meister8 who gives the example
of a large US automaker who, in the course of producing an
auto, provides some 3 billion opportunities for human error per day
mistic assumption of 1 serious error per million operations this
manufacturer could expect to see about 3000 serious undetected
production flaws per day- or about 1 in 3 cars. This data from
Meister is circa 1982, and surely the situation has improved a lot,
but the potential for problems is sobering. As a contrast, defect rates
for Japanese manufacturers were traditionally two orders of magnitude
better. The development of measures of product quality and
process capability, Cp and Cpk for example, are a result of efforts to
track product quality. We will discuss these later.
We will not focus here on the organizational issues of manufacturing
of precision products. But, we must keep in mind the importance
of quality of the final product. The critical elements of McKeown
above address the reduction of uncertainty at the interfaces between
processes and products. In fact, that is the primary objective of the
precision manufacturing engineer- reduction of uncertainty. With
this, the determinism championed by Donaldson and Bryan will be
more economically realized.
Manufacturing processes (so-called unit processes) date historically
back to some 4000 BC where hammering of metals for jewelry or
simple tools were practiced. For reference, Kalpakjian9 summarizes
the developments of these processes over time. Gradually, over the
first five or six millennia, the use of tools of higher sophistication,
1.4 Historical developments in manufacturing
1.4.1 Background
in the assembly operations (remember n x m) alone. With an opti