Managing
the IE (Industrial Engineering) Mindset: An investigation of
Toyota’s practical thinking shared among employees
Abstract:
Purpose: The
goal of this work was to investigate the managerial practices of today to
understand if Toyota is sheltering themselves from these newer practices or embracing
them like most believe.
Design/methodology/approach: This
work utilizes a new form of data mining named Latent Semantic Analysis (LSA) to
analyze an organizations ideal management practices.
Findings: This
work shows quantitatively that TPS favors earlier versions of industrial
engineering compared to the optimization techniques available today.
Originality/value: The
use of data mining to analyze organizational management practices.
Introduction
What
is unique about Toyota's system is not particularly any single piece of TPS,
but how the pieces are combined to bring out something new, different and very
difficult to imitate. It is argued in this work that Toyota's management system
is a richly interconnected set of parts and relationships that are more
important than the nature of the parts themselves. This means that even if the
parts themselves can be identified, their relations are often lost, which loses
meaning of the system. It is believed that research in TPS must follow the same
type of systems thinking to discover how TPS emerges from the way the parts are
organized in the system. Holism, rather than reductionism can provide a more
entire solution than a partial one. Historically, practitioners have been
concerned about what Toyota is doing now rather than what was Toyota doing when
TPS did not exist. Pioneers like Taiichi Ohno, the father of TPS and one of his
close friends, Sheigo Shingo, an industrial engineering consultant to Toyota
during the time, are less received and noted for developing TPS.
In
Ohno's book (Ohno, 1988), named the Toyota Production System, Ohno firmly
believed that TPS is simply a form of industrial engineering (IE) aimed at
reducing cost through systematic study. By treating everything as a process,
Ohno and Shingo built the interconnections of TPS one by one, but more
importantly passed on this industrial engineering way of thinking to future
generations. The
purpose of this work is to evaluate and quantify some of Toyota's thinking
styles as it relates to Ohno's traditional view of TPS. It is speculated
that one of the ways Toyota is able to develop such a holistic approach to TPS
is by passing down from generation to generation a type of thinking similar to
industrial engineering. The secondary goal of this work is apply a new form of
management science, named dimensional reduction analysis to highlight and quantify
managerial preferences. The secondary
goal of this work is apply a new form of management science, named dimensional
reduction analysis to highlight and quantify managerial preferences. In this
work the link will between TPS and IE will be established and analyzed to
determine which trend of IE practices are utilized to maintain the TPS
structure.
Literature
review
Ohno
believed that the quickest way to catch up with America was to import American
production management techniques and business management practices. Toyota
studied industrial engineering (IE) which by Ohno's accounts can best be
compared to the Toyota Production System. That is, a company-wide system tied
directly to management to systematically lower cost and raise productivity. Shingo
also viewed TPS as a way of thinking to addresses plant improvement. He
believed that management should possess a set of fundamentals closely related
to industrial engineering as a way to spread and teach the Toyota Production
System.
1 Industrial
engineering
An industrial engineer is one who is concerned
with the design, installation and improvement of integrated systems of people,
material, information and equipment which utilizes specialized knowledge and
skills in mathematical, physical and social sciences together with the
principles and methods of engineering analysis (Salvendy, 2001). Over the years
industrial engineering has drawn upon mechanical engineering, economics, labor
psychology, philosophy, and accountancy in an effort to bring together people,
machines, materials and information (Saunders, 1982). If industrial engineers
had to focus on one aspect of their field it would be productivity or
productivity improvement. That is, the total elimination of waste by increasing
efficiency through cost reduction (Going, 1911).
Industrial
engineering not only covers the technical aspects of systems, but also systems
relating to management. Anderson proposes that industrial engineering is one
the primary drivers for linking the needs of the employers to the needs of the
employees. Employers want industrial peace, reduction of cost, higher
efficiency and improvement in quality. Employees want steady work, higher
wages, better personal relations with their supervisor and good working
conditions. By utilizing industrial engineering techniques, management can
develop, evaluate and improve the wants of both groups (Anderson, 1928).
2 Scientific
management
Scientific management
was never a manner of how much a person can do under a short burst of speed but
instead a safe and comfortable working speed that can be done day after day
(Mogensen, 1935). The long term prosperity of the employer cannot exist unless
it is accompanied by the prosperity of the employee (Gilbreth, 1973).
Scientific management included employees to some degree in decision making.
3 Skill
sets of industrial engineers in the era of scientific management
IEs
were mostly concerned with defining processes, identifying problems and
establishing standard operations (Harrington, 1911). Some of the most practical
techniques employed by IEs was the use of direct observation and work sampling
(Staley & Delloff, 1963). The industrial engineer also was proficient with
the use of charting. By breaking down processes into smaller units, the IE
could analyze work flow by examining process steps visually. Lastly, the IE was
concerned with running trials to test new productivity ideas. By testing
factors one at a time and by sequentially changing those parameters based on
previous trials, the IE could speed up decision making while focusing on
improvement.
4 Industrial
engineering today
Today,
industrial engineering has become a more integral part of the organization.
With the invention of the high speed computer in the 1960s industrial
engineering has evolved into a hard discipline where data can be recalled at
any time and decision making can be improved through the use of models and
simulations (Saunders, 1982). Computers have given industrial engineers the
ability to analyze and optimize complex systems throughout the organization
(Katzell et al., 1977). Industrial engineering offers several sub specialties
such as human factors, job design,
labor psychology and systems engineering.
Now it is not uncommon for IEs to work on planning systems, supply chains,
accounting systems and organizational polices.
5 Skill
sets of industrial engineers in the 21st century
The skill sets of the
modern industrial engineer are much different compared to the days of
scientific management. Most modern IE skill sets emphasize rapid organizational
change instead of spending time stabilizing and documenting current operations.
Techniques such as Process Design and
Re-engineering can result in radical change by focusing on end to end
processes. PDR assumes a clean state change and suggests skipping documenting
existing processes because it limits the vision of the design team with nothing
to be gained (Taylor et al., 2001). A popular tool to aid in the study of
complex organizational factors in PDR is Experimental Designs (ED) and Design
of Experiments (DOE) concepts. These techniques allow industrial engineers to
understand the complexities of the business and interacting factors acting on
and within the organization before leaping towards a new state (Czitrom, 1999).
Six Sigma is a systematic method for
strategic process improvement that relies on statistical methods to make
dramatic reduction in customer defect rates (Tanco et al., 2009). Initially
established by Motorola in 1987, Six Sigma has been extremely popularized as
new form of business management strategy (Jugulum & Samuel, 2008). Lean Sigma is another improvement
methodology that is being employed by industrial engineers. Proponents suggest
that by integrating statistical methods with the ideas of work simplification a
common language can be developed to help organizations be responsive to
changing markets while eliminating defects (Wedgwood, 2007). Six Sigma provides
the detailed statistical study to optimize projects while lean is usually
implemented through a series of short focused kaizen blitz. (Jugulum &
Samuel, 2008).
6 Has
lean followed the trends of industrial engineering?
The lean community
has followed many of the same trends and skill sets as applied by the
industrial engineering profession. Bicheno suggests that a kaizen specialist
should be capable of performing value
engineering in product design and development (Bicheno, 2000). There is work that suggest the kaizen
specialist should be able to perform cellular manufacturing, production flow
analysis and supply chain infrastructure design (Askin & Goldberg, 2002;
Srinivasan, 2004). In these contexts, the kaizen specialist is illustrated as
person that exists within an organization to advance lean concepts in highly
specialized areas single handedly. In other ways the kaizen specialist is also
expected to work with employees utilizing team-based worker participation
activities often referred to as kaizen events. Compared to extreme Taylorism,
where the IE function is responsible for telling workers what to do, kaizen specialist
appear to be much nicer and softer. For example, the work of Martin and
Osterling state that these specialists should be armed with PowerPoint kick-off
material, masking tape, whiteboards, post-it notes and kaizen team t-shirts
(Martin & Osterling, 2007).
7 Is
Toyota sheltered from modern industrial engineering?
IE handbooks today
are emphasizing system optimization, advance computational mathematics and
rapid overhaul within organizations. Interestingly, Toyota’s approach to TPS
appears to be highly shielded from modern trends in the industrial engineering
profession and mainstream business improvement methodologies. In 1935 Sakichi
Toyoda the founder of Toyota developed five basic teachings based on the Toyoda
family work ethic. His teachings emphasized the importance of practicality,
good study habits and healthy homelike work environment. In the 1950s Taiichi
Ohno initiated a new type of production system (i.e. TPS) with an emphasis on
standardization, just-in-time, jidoka and kaizen (Ohno, 1988). Ohno’s shop
floor focus and the idea of testing practical ideas immediately encouraged
learning by getting employees to confirm failure with their own eyes (Ohno,
1988b). Ohno viewed that management should join with subordinates in
experimentation and each supervisor must have the ability to teach. In 2001 Toyota continued this practical
view of TPS when Fuijo Cho then President of Toyota Motor Corporation released
the Toyota Way, a set of managerial values to strengthen the organizational
thinking as it relates to work (Cho, 2001). In 2005 Cho re-issued the company’s 8-step problem solving process
named Toyota Business Practice (TBP) as an effort to share a common way of
thinking about problems in the workplace (Cho, 2005).
Research
approach
The overall approach
in this analysis is to analyze Toyota’s organizational documents by applying
statistical data mining. This work will use Latent Semantic Analysis (LSA) to
study Toyota’s industrial engineering techniques, systems and managerial
practices. LSA is a theory and method for extracting and representing the
contextual-usage and meaning of words and phrases by statistical computation
applied to text (Landauer, 2004). LSA is based on Singular Value Decomposition
(SVD) which is a mathematical matrix decomposition technique using factor
analysis. In LSA repetition does not imply relevance because LSA looks at both
the local weights and global weights and normalizes them.LSA is particularly
not suited for distinguishing similar terms that vary in context. The study of
IE practices can be more quantitative and precise using LSA compared to
traditional techniques. Existing techniques are largely subjective and
qualitative. Current methods rely on interviews that assume that participants’
accounts are a fair reflection of what has actually occurred. Subjects also
tend to inflate the results due to attribution theory (Gerhart, et al., 1999).
Attribution theory is the tendency to make causal explanation about the world
based on individual internal beliefs.
Questionnaires and
surveys are also well accepted techniques for conducting research.
Unfortunately, research shows that subjects are influenced in their responses
by how questionnaires are formatted. Another popular technique is the use of informants (Gardner &
Wright, 2009). An informant is a knowledgeable subject or employee inside the
company that is used to measure the content and quality of the company’s
system. Informants pose many problems in the study and are often a source of
unreliable data. For observation techniques to be accurate, analyst need to
make observations over a wide variety of persons and range of context. There
are also many limitations due to participant observation. Human behaviors are
often emitted at a rapid pace that exceeds our ability to note by informal
means. Research has shown that behaviors are bundles of complex information
(symbols, logic emphasis, warmth, aggressiveness, syntax, humor) and the
personal process of filtration suffers from biased conclusions, which means
that it is difficult to lead to objective evidence (Biddle, 1979). LSA may
prove an acceptable substitute for traditional research techniques; however,
there are noted concerns in this new area of study. First, it is assumed that employees within the organization are
aware of IE expectations and can perform them if asked. Second, that these expectations produce a conforming behavior. Third, it is assumed that these
expectations have been communicated by management throughout the organization.
It is also unknown how management sanctions these expectations, which is the
positive or negative reinforcement for engaging in the desired behavior. These
unknown areas provide many interesting and future opportunities for research. Bedsides
these short falls of studying IE corporate documents, inscriptions do represent
a continuing existence and a more permanent intention of how IE is being used.
Research
methodology
A
document-term(s) matrix was created
from numerous Toyota documents; such as the Toyota Way, The Toyota Business
Practices, the team member basic training manual, the team member handbook,
role of the supervisor, standardized work training manuals, process and system
kaizen manual and problem solving for managers. Also tables for text corpus
properties of the documents are used in the matrix and the themes selected in
the study. Next, the Singular Value Decomposition (SVD) algorithm is used to reduce the document-term(s) matrix.
Interpretation
of results in latent semantic analysis
The
overall goal in LSA is to map the dominate semantic themes in a reduced
dimensional space. The reduced dimensional space represents all word and
document vectors in the semantic space or text corpus. This work will map the strength (i.e. magnitude) of each
word vector and its ranks to illustrate the level of dominance throughout the
document collection. Rank 1 (lower order) is the most dominant rank followed by
rank 2 and so on. The singular value matrix indicates through a scree plot (not
shown) the optimal rank. Ranks beyond the “k” value are less dominant.
Results
Figures
attempt to describe how a manager from Toyota is expected to apply some of the
same industrial engineering logic or thinking styles in the workplace. (For
Detail analysis of figures go to http://www.jiem.org/index.php/jiem/article/view/293/252
)
Discussion
The
figures attempt to evaluate how managers at Toyota are trained in assessing the
workplace. More specifically, when a manger begins to question or speculate
organizational activities or functions, there are two general approaches
managers are trained to assess the current situation. A rank 1 data point in
quadrant four would indicate that a manager's starting point in the
investigation is to gather information by searching through a variety of
different sources such as reports, shift logs, e-mails and databases. Data
points in quadrant two would mean that a manager's first instinct into
inquiring and assessing a workplace issue or problem is to see the problem
first hand, meaning that it is more important to judge the situation personally
without trying to make assumptions or inferences about actual situation.
Results
suggest that Toyota strongly favors the scientific management approach in
assessing workplace problems. Figure 4 substantiates quantitatively that Toyota
trains it managers in genchi genbutsu which
in Japanese means go and see.
Managers are expected to go to the source to find facts to make correct
decisions. This data also implies that management should not rely on other
employees’ interpretation of the problem. It could also be argued that Toyota's
scientific management approach for information gathering is out of date,
unpractical and too slow for modern business. . Not all andons are problem
solving or genchi genbutsu visits by a manager. Most andons are answered by a
team leader to adjust or correct the process. Yet, when problems do reach a
point of escalation or severity, managers at Toyota are expected to see the
problem first hand. If the problem solving team did not actually see or witness
the problem first hand, the countermeasure trial would not be approved. This is
important because data does not replace facts gathered through genchi genbutsu.
Again, like the Gilbreth's pointed out some 50 years ago, many charlatans got
into time and motion study without understanding the purpose of scientific
management. Like lean, most practitioners place value in obtaining speedy
results rather than achieving consistency, predictability and stability. Toyota
views that stability (i.e. standardization) is the prerequisite for kaizen. Toyota's
preference on standardization compared to BPR demonstrates quantitatively that
managers follow an older view of the IE profession.
Conclusion
This work uses a new
method for analyzing managerial practices. LSA was used to mathematically
describe Toyota's approach towards industrial engineering. The other goal in
this work was to understand the similarities or differences of industrial
engineering towards Toyota’s management style. Early work indicates that Ohno
and Shingo modeled much of their thinking towards the earlier versions of
industrial engineering, namely scientific management. Up to this point, most
work describes Toyota or lean methods as modern IE techniques. This research
shows quantitatively that Toyota’s managerial practices of today very much
resemble the IE profession in the early 1900s. It is argued that Toyota has
remained successful applying TPS because they have sheltered themselves from
modern IE influences that seek to raise the competence of a single employee
(i.e. industrial engineer) rather than the basic thinking skills of all
employees. Lastly, this work illustrates that Toyota has been successful
passing down scientific management practices from generation to generation. Interestingly,
Toyota’s drive to “get back to basics” continues and remains a constant force
among managers to prepare future generations of Toyota employees.
