Featured Article - December 2005
TPS vs. Lean and the Law of Unintended Consequences
By Art Smalley
Editor's Note: This
article continues our recent focus on "looking Lean" versus "being
Lean". We have discussed this extensively on the blog.
Every
couple of years I am fortunate enough to visit a former colleague’s
Toyota Motor Corporation plant and to tour their facility. This
year I was able to visit an engine plant in West Virginia for an
eventful day of observation. The premise for the visit was to say
goodbye to a former colleague, for whom I briefly worked for Toyota in
Japan during the late 1980’s and early 1990’s. After four years
of advising and directing all production operations in the 1,000
employee producer of over a half million small engines and
transmissions annually, he was returning home to Nagoya.
During
and after the visit I could not help but reflect upon what I observed
at this particular Toyota facility versus what I experience on typical
visits to other plants in North American attempting a lean
transformation. The West Virginia plant has won the Harbor award
for most efficient engine operations in the United States and nearly
meets or exceeds the performance of its sister plants in Japan.
It has no late deliveries and has less than one day of inventory
between it and the customer. Scrap is minimal and customer
defects were in the 50 ppm range. Real productivity was up 10% over the
last year and more improvements were in the works. The plant is
technically a cost center but it contributes to the healthy bottom line
of the company with its overall efficiency. With this performance
level, anyone who visits this stellar facility should come away duly
impressed. But many people walk away from trips to places like
this one and similar Toyota facilities with a mix of awe and utter
confusion. The awe is self-explanatory. The confusion comes
from the fact that the facility looks nothing really like the
stereotypical image of lean that is often fostered in the U.S. by
various well intending parties.
For
starters, the plant staff includes no dedicated change agents or black
belts; there were no value stream maps posted anywhere, nor were there
value stream managers; no small U-shaped work cells; only a small
portion of the plant contained actual standardized work charts; and
many of the daily tracking systems were highly computerized. On
top of all this, there were several automatic guided vehicles, and many
robots that had been installed to replace human operators. A
typical reaction is, “Isn’t this the antithesis of what we’ve been told
to think is Lean manufacturing? Surely if they have added robots
and automated tracking systems but don’t have a value stream map for
every manufacturing line or a standardized work chart on every single
machine then this Toyota plant is not Lean. Has Toyota fallen off
its lofty pedestal?”
The
answer, to put it bluntly, is “No”. Real TPS is not just about
“flow” or “pull production” or “cellular manufacturing” or any of the
other catchy phrases or tools you may frequently hear. For over
fifty years TPS in Toyota has been primarily concerned with making a
profit, and satisfying the customer with the highest possible quality
at the lowest cost in the shortest lead-time, while developing the
talents and skills of its workforce through rigorous improvement
routines and problem solving disciplines. In every piece of TPS
literature from Toyota, this stated aim is mixed in with the twin
production principles of Just in Time (make and deliver the right part,
in the right amount, at the right time), and Jidoka (build in quality
at the process), as well as the notion of continuous improvement by
standardization and elimination of waste in all operations to improve
quality, cost, productivity, lead-time, safety, morale and other
metrics as needed. This clear objective has not substantially
changed since the first internal TPS training manual was drafted over
thirty years ago.
Despite
this consistent message on the part of Toyota, the lines have become
crossed in many Lean plants. In the U.S., becoming Lean appears
to have gone down a path of implementing tools such as “one
piece flow”, “value stream mapping”, “standardized work”, or “kaizen
events”, but results have not always followed. Toyota, by
way of contrast, has stayed focused on its principles and a disciplined emphasis on process improvement to obtain results such as “making a profit”, “reducing lead-time”, “improving
productivity”, “achieving built in quality”, as well as “respecting
human dignity of employees” etc. The difference may sound trivial, but
it is actually significant. In fact, it is likely the main reason
why Toyota has continued to see success on so many dimensions, where
others struggle. This mishap has not occurred intentionally. As
the title of this article indicates, it is something like the ‘law of
unintended consequences’ at work. If we are to be like Toyota in terms of results, and not just look and sound like Toyota in word, then several things need to change in Lean implementation.
There
are several important aspects upon which we should reflect. To a
large extent, the points outlined below contribute the reasons many
have struggled so mightily in their efforts to become Lean, but have
failed to actually become like Toyota
(Price – Cost) x Volume = Profit
Rarely
does anyone at a Lean conference or in a Lean article discuss in any
detail the notion of profits and improvement. It is almost as if
profit is not a proper topic for public discussion, and should be left
for accountants to contemplate behind the scenes. Instead people,
articles, and books opining that lean is all about flow, value, and
customer satisfaction are the norm. Maybe Lean is about these
things, but TPS is not that simple.
Among
the first slides I was shown at Toyota during my employee orientation
was the simple equation: Profit = (Price – Cost) x Volume.
In the highly mature, competitive automotive industry, Toyota believes
they cannot dictate price, and the market decides how many units will
sell in the long run. Hence, the only lever they have to control
is cost, so all effort is made to manufacture in such a way as to
reduce cost; otherwise the company will go out of business and not earn
the right to grow and provide higher wages etc. to employees. The
basic, critical assumption behind every aspect of TPS is to
continuously reduce cost. Inventory, scrap and rework, extra
people and resources, downtime, etc… all cost money - and all of these
must be eliminated or minimized. In other words, waste
elimination has a specific purpose, and it not just seven neat categories to be quoted during work or training.
This
point was continually drilled into my head at work until it became
second nature. Toyota is a very cost conscious company that
is very good at designing, producing, and delivering quality products
that meet the needs of its customers. While other individuals
have commented on the need in Lean to have “problem awareness” or
“kaizen consciousness”, the phrase I remember most from my superiors at
Toyota was to have “cost consciousness” and to never waste a
dime. Time and time again, the Japanese proverb “you must wring
water out of a stone if necessary” was repeated in budget meetings and
project review sessions. Unfortunately, these internal processes
that make up the culture of the company, are invisible from the
outside. External consultants and academics cannot see this, and do not
write about it as a result.
Regardless
of your job in Toyota, though, the cost structure is tightly controlled
and subject to harsh review and continual improvement pressure.
This point was very clear to me during my visit to West Virginia
where my colleague explained all the benchmarks between this plant and
a similar engine plant we launched years ago, and in comparison how
fewer support resources were available to him now. Despite the
vastly reduced support levels, he had produced a superior
facility. Somehow this same results emphasis and cost focus needs
to be put into all Lean efforts if American companies are to succeed in
the long run. Anyone who tries to tell you that TPS is all about
“flow” and not about making a profit is sadly mistaken.
TPS Tools Are Just Ways to See Problems – They Are Not Solutions
A
disturbing trend in many Lean Manufacturing efforts I observe pertains
to the instruction given to plants by the head of the company’s lean
program or change agents. The nature of the advice tends to
be along the lines of draw value stream maps, establish
supermarkets, create one piece flow, post standardized work charts on
every machine, create U-shaped work cells, implement kanban, and walk
the plant floor to conduct lean audits every day. These are
examples of Lean phrases recited almost daily and the list does not end
here. While these and similar key phrases from each company’s lean
journey can often be heard in U.S. plants, rarely if ever will you hear
then uttered in this fashion at Toyota.
The inherent problem in all of the above statements is the following:
practitioners have not been required to first determine what the exact
problem is before being instructed to use the tool in question.
When people are told what to do, they are absolved from the critical
responsibility of thinking for themselves and determining why they need
to do it, what is the nature of the actual situation at hand, and
deciding how to best proceed. TPS is not a set of rigid
guidelines proscribing what to do in exact detail in every
situation. In fact the T in TPS has often been said to stand for
“Thinking”, and not just Toyota. The system was developed over
the years by trial and error focusing in on specific problems, making
people challenge the conventional wisdom behind situations, identifying
the root causes of issues and then solving problems in some unique, and
often spectacular, manner. When the countermeasure developed
worked particularly well, or proved to be a very good analytical aid,
it became a standard tool in the Toyota arsenal, and was taught to
other people so it could be used wherever it was
applicable. If TPS is reduced to just applying tools
without equally understanding the problem at hand, it will not produce
as much in the way of significant results.
Examples of Tool Development in TPS
| Sample problem statement | Historical root cause | Analysis tool or countermeasures developed |
| Poor work motion and material flow in line layouts | Insufficient detail in layout planning or line conversions | Value stream mapping and work motion analysis. |
| Right stamped parts not available when needed despite much inventory | Long change over time | SMED. Analyze and separate internal from external work. |
| Right parts not delivered to downstream when needed | No physical or accurate signal. Push style of production. | Pull system and kanban cards to signal replenishment. |
| High level of scrap and defects | Low process capability | Build in quality at the process. Not through inspection. |
| Low labor productivity in man machine combination areas | One man one machine layout and work assignment | Separate man from machine. Create standardized work. Promote a multi-skilled work force with job instruction. |
The
Figure above is representative of how lean practitioners should think
about implementation. The goal of TPS is not to implement a tool
such as standardized work or kanban. The tools are merely methods
techniques for analyzing and improving situations in a systematic
context. It is critical to have a specific objective in mind when
setting out to make improvements in Lean. In all cases, the
current condition surrounding the situation must be thoroughly grasped
and the actual problem defined. Then the probable root causes
must be isolated, a specific action plan developed, improvement goals
set, and a validation method established. If using an existing
Lean tool helps you to diagnose or accomplish this, so much the better,
but goal is improvement and not mere use of the tool. Failure to
follow this path in Lean implementation diverts the message from “solve the root cause of major problem X” to “apply tool Y in some other area”.
Lots of activity might take place but minimal results will be
realized. Mr. Taiichi Ohno remarked that TPS is very much like
the scientific method of experimentation. When this is not kept
in mind, the result is “push” style Lean (Do as you are told), rather
than “pull” Lean implementation (What is the biggest problem?)
People naturally rebel at the former approach, but will embrace and
take ownership of the latter. The leader or consultant who
properly grasps this principle can be highly effective and will be much
more likely to get results.
Don’t Confuse Local Tools in TPS with Universal Solutions.
I
have a further nagging uneasiness with the prominence of certain lean
tools as they are typically used in the U.S. Chief among them are
value stream mapping, standardized work, kaizen, etc. The problem
is not with any of the tools, per se, as they can be powerful
improvement devices. My concern is with how the tools
tend to be misused. Value stream mapping, for instance, is
perhaps the most widely used tool in lean programs today. The
prevailing assumption in virtually every plant is that a value stream
map must be drawn for each product family, a value stream manager
anointed, and that it will somehow magically reveal all of the plant’s
problems. This practice has become a sort of litmus test for
Lean. If there is no value stream map and an associated tracking
center, then the company is not pursuing true Lean manufacturing.
But there were no value steam maps in the Toyota facility
in West Virginia, nor are there value stream managers. And this
is hardly because Toyota employees are so smart they all carry the
value stream maps around in their heads.
The
reason there are no value stream maps in most Toyota plants is very
simple in hindsight. It was a tool developed primarily as an
analytical aid to look at material and information flow problems in certain processes. In fact, the actual name of the tool at
Toyota is “material and information flow analysis” - not value stream
mapping. A third dimension, human motion, is often added to the mix for
consideration as well at Toyota. As TPS evolved internally and
was rolled out to supplier companies externally a consistent problem
was insufficient investigation into the details of material flow,
information flow, and human motion in the process. A typical
layout drawing, for example, simply does not emphasize these aspects
clearly enough to bring problems to the surface. Once production
starts, it is too late or costly to fix some of these items.
In
response a creative countermeasure was developed that became a
requirement for engineers and others in charge of manufacturing
processes and line conversion work at suppliers. The emphasis was
to draw both detailed standardized work charts depicting operator
motion, and flow charts depicting material storage locations,
scheduling points, and operator work sequence before the start of
production. In other cases, this tool was used externally to find
ways to convert lines to more efficient ones. The key point is
that the tool was created to analyze and solve a specific category of
problems Toyota faced in new production lines and in helping suppliers
implement lean.
From
this fairly specific local origin in Toyota, the tool was slightly
modified (the human motion emphasis was reduced) and popularized in the
U.S. by my good friend and former Toyota colleague John Shook, and his
co-author Mike Rother, in their insightful, best selling workbook “Learning to See”. The title of the work I think is important. Originally the authors had considered titling the workbook Material and Information Flow Analysis for Lead Time Improvement and Work Place Kaizen.
This name, which would have been truer to the original intent of the
material, was changed for marketing reasons to “Learning To See”.
The workbook went on to sell over 125,000 copies, and has affected the
direction of lean efforts in the U.S. more than any single publication.
Unfortunately the object of what the workbook urges the reader to see is not as clearly communicated in
the catchier title – and here is where the law of unintended
consequences kicks in. The book is about learning to see what is
primarily a material and information flow problem, or essentially
elements of the JIT pillar of Toyota’s production system (flow, takt
time, level, and pull production). By design it doesn’t even
attempt to address the topic of Jidoka for example which Toyota
considers an equally if not more important support pillar than JIT or
equipment stability. The technique used in the workbook simply
measures the overall manufacturing lead-time versus production value
add time. Everything non-value adding (i.e. the waste) is to be
eliminated and answering seven specific questions outlined in the
workbook will help you accomplish some of this goal. Overall,
however, when the 4M’s of manufacturing (man, machine, material, and
method) are considered you’ll realize that this tool mainly considers
the material (and information) flow component. The other 3M’s are
much less emphasized and one other important M – metrics - is expressed
chiefly in terms of lead-time and value-add time. This is fine
for Toyota. Internally they well know the limits of the tool and
understood that the it was never intended as the best way to see and
analyze every waste or every problem related to quality, downtime,
personnel development, cross training related issues, capacity
bottlenecks, or anything to do with profits, safety, metrics or morale,
etc. No one tool can do all of that. For surfacing
these issues other tools are much more widely and effectively used.
Unfortunately,
the average user of the workbook tends to copy the pattern expressed in
value stream mapping regardless of the nature of their manufacturing
problems. The unintended consequence of the success of the method
has been to convince many people that it is a universal tool for
identifying all problems in manufacturing operations. Marketing hype
helps reinforce this notion. “Just draw a value stream map and it will
show you all your problems to work on” is a popular refrain that I hear
quoted in companies attempting lean. This guidance however biases
companies with major quality, downtime, or factor productivity problems
to deemphasize them since those items are not surfaced well using the
method and questions outlined in value stream mapping. The tool
just does not frame these problems well by design. Couple this
effect with the fact that most lean efforts already have a
disproportionate bias towards the concept of “flow”, and there is a
recipe for inherent danger. For example instead of learning to
see what is truly broken in their processes companies wind up typically
focusing on a particular subset of operational problems chiefly that of
flow and lead-time related issues.
There
are a host of similar problematic examples I could mention with other
tools, such as standardized work, U-shaped cells, kanban, etc. They are
all good tools but have inherent limitations and specific applications
in Toyota. Companies struggle when they take any of these tools
that have a specific and local purpose in TPS and attempt to apply it
as a more universal tool or solution for their respective
situation. Inevitably this conveniently ignores other, often more
serious, manufacturing problems and fails to deliver the desired
results.
The Importance of Developing People
Another
critical difference between Toyota plants and other ‘Lean’ plants is
skills development. Despite common knowledge of its importance,
this topic is typically relegated to the HR department or someone else
in training and development. Implementing lean requires a
specific set of skills and experiences. Most companies go to
great lengths to hire or create special change agents, responsible for
directing and facilitating lean program implementation. On the
surface, this practice seems to make sense. There is, after all,
a small, elite group of specialists in Toyota filling similar roles,
although there are perhaps only 50 of them world wide in a company of
over 200,000 employees, and they are very senior experts in TPS for the
most part.
The area of
skills development that I am most concerned with, however, is not that
of the change agent specialist, but that of the natural operations work
team leader. Although less prestigious than the TPS specialists,
development of work team supervisors in Toyota is considered an
equally, if not more important, topic merely because there are tens of
thousands of these individuals. Specifically, it is these
manufacturing leaders that are the main focus of training efforts in
Toyota since they lead the daily work areas, and they directly and
dramatically affect quality, cost, productivity, safety, and morale of
the team environment. In most companies implementing Lean the
reverse set of priorities is true. Emphasis is put on developing
the specialist, while the supervisor skill level is expected to somehow
develop over time on its own.
Focusing
development efforts on the supervisor goes all the way back to Toyota’s
early days, when they introduced several management training programs
in the company known collectively as Training Within Industry
(TWI). These programs focused on the development of the
production supervisor and his or her ability to instruct, improve, and
lead their work team. Three specific courses known as Job
Instruction, Job Relations, and Job Methods were used to provide
supervisors with skills needed to develop their teams and solve
problems in the work area they managed. To a very large degree,
these programs are still the back bone of the supervisor development
program in Toyota, and a big part of why the company is so successful.
The
contents of these courses are still highly relevant to the success of
Lean or TPS today. Everyone managing people at Toyota is expected
to not only have knowledge and proficiency of their job, but to teach,
improve, and solve work team related issues in a standard and expedient
manner, as well. These skills are not trivial. There are 10 hour
training courses for each of these areas, as well as in other topics in
Toyota. Space limitations in this article preclude describing
Toyota’s entire supervisor training program, but it is important to
understand the fundamental point that, while having trained specialists
in staff roles, can be helpful, Lean success is more directly related
to a company’s ability to elevate the Lean skills of operations
managers and supervisors. Within Toyota, the internal focus is on
achieving results through natural work teams and not via the efforts of
change agents around them.
Final Advice
If
there is just one simple piece of advice that I would give those
implementing Lean manufacturing today it is the following; put a stop
to the dogmatic routine of using any single tool (value stream maps,
kaizen events, or any other) and expecting that it will highlight or
solve all the problems in the facility. The reality concerning actual
TPS could not be further from the truth. While value stream mapping for
example can be a great high level analysis tool that can show many
problems in flow, lead-time, where inventory piles up in the factory,
and what might be causing on-time delivery problems in production,
these may or not be the main problems facing your operations.
Certainly these are not the only problems facing a facility.
Therefore, different and better ways to see problems and to solve them are required. In other words customers, executives and
shareholders care more than just about the flow of product through the
plant. They care about profits, quality, downtime, productivity,
safety, and many other key metrics. These should be the concerns
of everyone implementing Lean, as well.
From
my own years of experience and discussions with numerous Toyota
managers, I have learned the hard way that the only proper approach to
implementing TPS is to reflect long and hard on Mr. Taiichi Ohno’s
original advice. Ask what the greatest point of need for
improvement is and start from there. If it is quality then figure
out where the majority of the defects are occurring, why they are
occurring, fix them and prevent recurrence immediately. If the
problem is low productivity, then analyze jobs for non value added
versus value added time, figure out the points of the greatest amount
of waste and eliminate it. If the problem is on-time delivery,
then figure out what products are late, why they are late and fix the
root cause. If there is too much inventory and poor flow in the
plant, then by all means, draw a value stream map and get about fixing
the associated points in the process! Over course TPS is a
comprehensive system, and basic problem solving alone will not make the
company Lean. If Lean is implemented without a rigorous problem
solving focus, however, then I guarantee that you will not achieve
lasting results.
If the law
of unintended consequence runs true to form, then someone who reads
this article may run out and throw away some or all the Lean tools in
their corporate toolkit. Of course, nothing like this should
happen. TPS is a complete production system, comprised of many
methods and techniques, and not limited to mere problem solving tips.
What I have hoped to emphasize however is the following two
items. First you need to learn to use the right tool for the right
problem or you will struggle. Secondly, in all implementation
efforts the concepts of TPS and discipline of problem solving cannot be
separated if you aim to generate any real success. Toyota did not
even bother to name its production system for over 20 years after its
initial pilot line experimentation or write down the tools in a
handbook. Instead they focused on making real, quantitative
improvements in line with the core principles and metrics, and the
tools were deployed or invented along the way. Everyone should
keep that order in mind during their own implementation journey.
In
conclusion I would like to echo a couple of comments that senior Toyota
executives in Toyota are fond of making. First, no one ever said
that the business of making improvements is easy. It is
strenuous, time consuming work and requires dogged persistent if you
want to be any good at it in the long run. If the situation were
otherwise, then other companies would have caught up to Toyota by
now. Also, executives frequently point out that the T in TPS
actually should stand for “Thinking”, and this too requires hard work
in order to achieve any substantial process improvement. This is,
in fact, the very essence of the ‘Toyota Way’. In other words, do
not simply put opinions on the table or throw a universal tool at your
problems and then accept whatever happens. Instead, demand that
problems be brought to the surface, challenge all of the conventional
notions of manufacturing wisdom, and assure that root causes are
promptly corrected. Anything less is just another type of waste
in your factory and will not lead to improved profits, added value, or
enhanced customer satisfaction.
Art Smalley is the author of the Lean Enterprise Institute workbook Creating
Level Pull: a lean production-system improvement guide for production
control, operations, and engineering professionals, which
received a 2005 Shingo Research Prize. Art learned about lean
manufacturing while living, studying, and working in Japan for 10 years
as one of the first foreign nationals to work for Toyota. He spent the
majority of his Toyota career helping the company transfer its
production, engineering, and management systems to facilities around
the world. After leaving Toyota, Art became director of lean production
operations at Donnelly Corp., (now part of Magna Inc.), a tier one
automotive supplier with more than 15 plants in North America and
Europe. Art subsequently joined McKinsey & Company, where he was
the firm's leading expert in lean manufacturing. He currently aids
companies implementing lean through Art of Lean, Inc..
