Page 92 - The Real Work Of Data Science Turning Data Into Information, Better Decisions, And Stronger Organizations by Ron S. Kenett, Thomas C. Redman (z-lib.org)_Neat

84 The Real Work of Data Science

The First Industrial Revolution: From Craft to Repetitive Activity

In medieval Europe, most families and social groups made their own goods such as cloth,
utensils, and other household items. The only saleable cloth was woven by peasants who
paid their taxes in kind to their feudal lords. Barons affixed their marks to the fabric,
which came to stand for their levels of quality. While some details differ, the textile
industry all over Europe and China was similar. It was apparently the first industry to
analyze data. Simple production figures, including percentages of defective products,
were compiled in British cotton mills early in the nineteenth century. Quality control
activities generated data that was aggregated in ledgers for accounting and planning pur-
poses (Juran 1988).
The industrial revolution started in England. Richard Arkwright (1732–1792) was an
English inventor and a leading entrepreneur who became known as the “father of the modern
industrial factory system.” He invented the spinning frame and a rotary carding engine that
transformed raw cotton into cotton lap. Arkwright’s achievement was to combine power,
machinery, semiskilled labor, and a new raw material, cotton, to create mass‐produced yarn.
In 10 years, he became the richest man in England.

The Second Industrial Revolution: The Advent of the Factory
During the early twentieth century, a constellation of technologies and management tech-
niques expanded mass production. The internal combustion engine (and the oil and gas needed
to fuel it) and electricity powered the way, the production line formalized the division of labor,
and huge factories were built. The Taylor system, featuring time and motion studies, drove
production tasks and productivity quotas. And companies learned how to manage enormous
factories (Chandler 1993). This was the second industrial revolution.
As one example, Western Electrics’ Hawthorne Works, on the outskirts of Chicago,
employed up to 45,000 workers and produced unheard of quantities of telephone equipment
and a wide variety of consumer products. It was in this environment that Shewhart realized
that manufacturing processes can be controlled using control charts (Shewhart 1926).
Control charts minimized the need for inspection, saving time and money, and delivering
higher quality. W. Edwards Deming and Joseph M. Juran were instrumental in bringing this
approach to Japan in the 1950s. Deming emphasized the use of statistical methods (Deming
1931), and Juran developed a comprehensive management system featuring the so‐called
quality trilogy (Godfrey and Kenett 2007). Like Shewhart, both worked for Western Electric
in the late 1920s.
From a data analysis perspective, attention shifted from inspection to process and the need
to understand variation. Thus, statistical models and probability played a key role.

The Third Industrial Revolution: Enter the Computer
Computers have changed manufacturing in several ways. We’ve picked three to illustrate.
First, computers enabled “mass customization” (Davis 1997). Essentially, mass customiza-
tion combines the scale of large, continuous‐flow production systems with the flexibility of a
job shop. This allows a massive effort, with batches of size one. A call center that employs
screening to route calls to the right specialists is a good example.