All petroleum companies’ laboratories were busy in the 1950s searching for ways to make unique products from oil. Oil companies were also looking for ways to increase the consumption of oil as there was more oil than there was consumption. This was an interesting new frontier because oil is composed of many different organic molecules which can be broken apart and reassembled into something quite different. Common polymers like polyethylene or polyisobutylene are prominent examples. The properties of these new products were unique at the time and the developers in the laboratory knew more about their potential use than anyone in the converter market. The developers of these new molecules had to instruct potential users on how to mold, cast, or fabricate end-user products.
My
first experience with quality came from the role I played to evaluate our first
production of high-density polyethylene.
The plant struggled with the demonstration run, but large quantities of
the polymer were produced, and I was provided with several hundred pounds for
evaluation in the Products Research Laboratory.
The test samples failed all the tests of durability. Production of that particular polymer was put
on the back shelf for another 20 years. The process that was used in the test
run failed to produce a viable product.
Quality
in these early years had a very narrow definition and was usually defined as
just one parameter. Jokingly, one said about polyethylene that if it's white
and if it floats it must be polyethylene.
As
could be expected, with evaluations of these new products for end uses,
customers began to learn rapidly what raw material characteristics were
important to them for their success in creating new end-use products. The polymer characteristics required for
making film were different from those required to make bottles.
Other
suppliers developed related products and with growing customer interest, plants
began to be built to supply commercial quantities.
Competition
from other labs and their commercial plants plus the growth in customer
knowledge caused a rapid rise in the demand for quality to be defined with
additional parameters and measured by published test methods. Once this change occurred, customers began to
place limits on key variables and added these limits to their purchase
requirements. At first, producers were very secretive about the limits to which
they could produce products suggesting a disparity between what specifications
were required and what the plant could meet.
Competition
and customers’ needs drove specifications ranges to ever-narrower limits, so it
often became necessary to inspect output to ensure quality requirements were
being met. Some customers required tighter specifications than did others, so
products were more carefully inspected for them to earmark product specially
for them. Production that was out of range for most customers was
re-manufactured by recycling with new production at the molecular level. Recycling avoided losing off-specification
production that would otherwise be scrapped, wasting raw material, and labor,
and incurring disposal problems. It didn't take long before it became apparent
that the chemical process required improvement to make it capable of routinely
meeting specifications. The need to inspect
production before shipment ushered in the era of quality and continuous
improvement.
Improving
the process as a means to eliminate quality control by inspection became
obvious but was also recognized as a difficult task due to a number of factors.
The list of factors was long, consisting of process interruptions that
compromised data collection, employee
errors, testing variability, lack of proper equipment for specific steps in the
process, and the inadequacy of pneumatic controls for critical parts of the
process. All the characteristics contribute to an out-of-control process.
Plant
technical service engineers and chemists eagerly accepted the challenge to
improve plant capability but were unsuccessful—the task was too large for the
operating organization to handle.
Those
who worked during this time said that it was as if there was an undefined
struggle occurring between these few process engineers who were looking for a
constructive way of overcoming the operating problems and two opposing forces;
an operating organization that resisted change and a workforce that was
antagonistic. The workforce viewed process engineers as agents of management
and they were treated accordingly.
The
few process engineers were trying to change how operating personnel were
engaged in the process by shifting to a
focus on problem-solving, looking for root causes of operating
problems. The intent was to find some
single cause and correct it to make the process run reliably, but the bulk of
the management was ambivalent. It was apparent that they did not see any case
for changing how the plant was managed. Their unspoken method for overcoming
any of the operating problems was to increase accountability at all levels,
stressing the need for more discipline in operations management. It was not obvious at the time, but plant
problems were generally understood to be the result of errors caused by
careless operations employees, not by defects in the process itself. The
solution to that was to put more pressure on the workforce—creating even more dysfunction.
The
seed for this interest in solving problems came from reading about systematic
problem-solving. A training program for
all engineers on systematic problem solving was conducted but that program was
not effective because faulty data and unskilled treatment of the data did not
support systematic problem-solving.
Furthermore, finding a root cause of a systematic problem was not in the
cards.
One
of the operations supervisors said, “it’s like a war zone out there,” referring
to his morning control room review of overnight operating problems where the
explanation of every upset in plant operations was referred to as a “process
secret” by the workforce. Key logbook
entries were missing, and no one had any understanding of the events of the
previous shift when the upsets occurred.
Upsets invariably produced scrap and off-specification production----these
upsets were costly events, some presenting safety risks. This antagonistic behavior of the operating
crew served to reinforce the conviction of management that all the problems of
the plant were due to operator performance either intentionally or
unintentionally.
It
was obvious that there were operating problems caused by the engineering design
of the plant. The process engineers
suspected this additional likely cause because even they, with their knowledge
of the process, had difficulty establishing stable operations. Most of these design problems were assumed to
be normal because the recurrence of the problems was part of the work of making
the product for years. The instability
of the process was the cause of inspecting production before shipment, so it
was clear that our objective was to improve the stability of the plant. We all knew that inspection to ensure the quality
of production never works 100% of the time.
This
nearly unbearable situation lasted for several years during which time
equipment sabotage, intentional errors in the execution of routine duties,
management directives that were threatening, frequent customer complaints of
poor quality, and a poor safety record were normal. It’s a wonder that no one was seriously
injured, and the plant was not destroyed.
During
this time of conflict, someone had alerted the local newspaper of the plight of
the shift worker and a full-page article based on a night shift interview with
the operating personnel appeared in print, much to the surprise of local
management. The featured article was like confidential information leaked to
the press. The workforce saw success; management was embarrassed. The print
copy even made it to the New York corporate offices.
The
attempt by these few process engineers to tackle process improvement at the
lowest level in the organization was not specifically blocked by management but
was completely stalled by the war between management and the workforce. Neither
side trusted the other. There was little time left in a full workday to improve
anything when all efforts were devoted to repairing damage and reestablishing
operations. In addition, the ability of
the organization to tackle a system problem of this magnitude was insufficient.
This
contentious environment was nudged down the path of reconciliation by the
change in a few mid-level plant management positions. The new leadership had the philosophy of
focusing on the process and not on personalities. This shift in management
style happened by chance and definitely not by design. Occasionally change that occurs by chance results
in improvement, but don’t count on it.
The
changes were timely as the plant was beginning to export products to locations
where some customers were known for their attention to the use of process data
for quality improvement by implementing projects to change the process. Japan especially was getting to be known for
its success in improving the quality of products in response to difficulties in
world trade due to poor quality.
Japanese companies were leading the way in quality improvement using
proven principles adopted from the teaching of Dr. Deming from the USA. That
Japan was leading in adopting quality as a strategy was a paradox because the
broad theory that an improvement in quality would lead to lower cost and higher
productivity all the while increasing worker and customer satisfaction
originated in the USA with leaders like Dr. Deming, along with Dr. Juran and
others many years earlier. These methods were assumed to have been developed in
Japan, not the United States because Japan popularized the use of these
management techniques to improve quality and productivity. At the time, our interest in these methods
was increased from reading accounts of successes in process operations in Japan
as the next level of thinking beyond systematic problem-solving. We were gaining theoretical knowledge through
our study but were not yet applying this knowledge to plant operations.
Still,
few seemed interested in why the Japanese recognized a quality problem and
those in our industry did not. The widespread interest in quality in
manufacturing in Japan could be easily explained by noting that Japanese
company top management viewed improvement as critical to their survival. They adopted quality improvement as their
main strategy as they were convinced of the relationship between improving
quality and all the other outcomes of consisting of customer satisfaction,
worker pride, cost, and productivity.
The apparent reason for their adoption of quality as a business strategy
was that they had a case for action. When we saw little interest in this
approach in our company, we asked ourselves could it be that management here
did not sense a crisis because profitability was satisfactory, and they were
comfortable with the status quo? In the
jargon of the time, there was no case for action here.
The
attention to quality and process improvement received a boost from an
unfortunate production error.[1] High product temperature at the time of
packaging was known to have an adverse effect on product use by the customer.
Lacking noninvasive temperature sensing tools,
dial thermometers were stuck in the solid product as it moved down the
production line to determine if the temperature was acceptable to proceed with
packaging the product. This worked fine
for a while, then, a customer in Japan received their shipment of product with
a thermometer still stuck where the employee left it, having neglected to
remove it. Regrets were expressed in a
response to their written complaint, but they were not satisfied with the
response and replied by telling management that was not nearly as devastating
as the product which they received that contained some worker’s coveralls. Since the plant boasted that contamination of
the product was at the parts per million level, both of these events were most
embarrassing.
This
product contamination with a worker’s coveralls could not occur by chance but
had to occur from some prank, so it was viewed as an act of revenge even though
relations between management and the workforce were improving at this time.
These
contamination events caused two outcomes.
One is that the customer in Japan requested an in-person explanation of
how this contamination was allowed to happen and then describe the process
changes to prevent a recurrence. The operations manager traveled to Tokyo and
met with the Japanese customer’s management to provide details on the cause of
the product quality event, but unknown at the time, the presentation failed to
adequately convince them that there was a process-based solution in place. The
customer accurately perceived that a focus on the process was not yet an
attribute of our organization.
The
second outcome was the loan of a plant process engineer from a Japanese
affiliate to help teach our plant process engineers to use basic statistical
methods to improve the control of the process and the resultant quality. This
was a clear example of management leadership assisting the plant’s technical
organization in its efforts to improve the process. These two events were
opportunities to learn what Japanese industry thought were important methods to
manage production. They were convinced
that a focus on process improvement was a key strategy for their business.
Learning
from these two outcomes occurred only in a few spots in the organization.
A
few advances in quality and process improvement occurred at the workforce
level, but some significant setbacks occurred at the same time. Some employees voluntarily participated in
quality improvement projects as a result of the leadership of three first-line
supervisors and made notable contributions to quality improvement. Much of this was lost over the next year due
to personnel changes and organizational structure changes that did not support
the need for quality improvement, confirming that there was still no case for
action to improve and that command/control management methods were firmly
entrenched.
The
focus on operations by the Japanese engineer now on loan was on the statistical
treatment of operating data. The core of
his influence was to use operating data to separate common causes of plant
variability from special causes. His work showed that the process was often not
in control. Many operations personnel were then trained in the practical use of
statistical treatment of data demonstrating that the use of basic statistics to
identify problems would not only create a more uniform product but also make
their work easier.
As
a shining outcome of this training, one shift operations employee brought to
his workstation the software and his laptop computer from home to monitor one
dimension of product quality and found that in the past, a large percentage of
his time was spent adjusting the control settings on the equipment when in fact
the process was not stable. From his use
of practical statistics, the process operator learned when to adjust the
controls of the equipment and that making adjustments when not warranted would
only create more variability. This was an example of using technology from
outside our work sphere to improve operating stability, and voluntarily
introduced as a new method of operations control. This shift worker probably saw himself as an
outlier among his peers, as he subsequently resigned from his position and left
the company.
Work
to establish the stability of the process by focusing on the equipment now took
on more importance. The quality
advocates were pleased but not astonished with this worker’s efforts because
they had high expectations that the talent of the workforce would be evident if
allowed to develop; after all, the workforce hiring criteria was high. However, once hired, there was no strategy
for identifying the skills and talents of the individual and engaging that
individual in the improvement in quality.
Two
major capital-intensive modifications were made to the plant significantly
improving the stability of operations.
These two projects were the first major capital investments purposely
aimed at improving stability in the long history of the plant. One was computer control and the shift to electronic
instrumentation and the other was a complete redesign of the packaging section
of the plant. These two projects demonstrated process improvement and
contributed to improving the capability of the process to meet specifications
and lessen the dependence on inspection.
National
interest in operating excellence ushered in an era of quality management under
the guise of improving quality. The industry was flooded with written material
from theorists on improvement techniques and philosophy.
Managers
formed Quality Councils to focus on goals and their translation throughout the
organization, accompanied by mission and vision statements, change management programs, company-wide
seminars, awards programs for waste reduction, etc. Senior management made a bold move and formed
a company-wide select team to identify key projects for improvement and to
promote quality as a business strategy.
However, for the most part, the effect of these programs never made it
to the operational level of the company.
One
awards program that did make it to the workforce was to recognize work teams
for the lowest level of scrap production; but that ended in total failure
because to be the best at meeting the low scrap goal, those operating the
process began to return the contaminated product to the prime product
flow. This generated both low scrap
levels and as well as many customer complaints.
A principle was easily recognized from this program; be aware of
unintended consequences coupled with the principle of testing a change before
implementing a change, recognizing that not every change results in an
improvement.
The
company-wide spirit was notably increased but had little effect on identifying
projects for improving the basic process capability of various processes. Most of the energy in these management
initiatives was created by well-worded inspirational writing on the subject of
quality and operating excellence. These fashionable activities swept the
industry. Those on the shop floor ignored most of this activity.
Meanwhile,
a major example of a positive quality focus was the construction and staffing
of a new plant apart and separate from the main facility, purposely planned,
and executed in a manner that avoided carrying over any old customs of how to
manage plant operations. The new plant
performance was better than anticipated.
The demonstrated performance of the new plant regarding safety, customer
satisfaction, and cost of operations was recognized as outstanding. The new plant’s success was so valued that
the plant manager who promoted quality and process improvement from its
inception, on his planning to retire, held off his retirement to be sure that
his replacement had the right philosophy on employee involvement and the spirit
of open management so that the plant would not see a setback. This may have been the only time that an
outgoing manager had any influence on choosing his successor.
This
new plant brought to light several key principles on how a company can adopt
quality improvement as one of its key strategies. One key principle was to view
the company not only as a structure of authority but also as a structure of
interdependent processes. Even though
the company business was based largely on processes in the plant, very few
managers could relate to the processes involved at their level. In this new, remotely located plant,
employees were so tuned into the concept that processes exist at all levels in
the organization, one technician said during a worksite meeting on quality, “managers continue to urge us to study and
improve the processes for which we are responsible, and someday, we will ask
you, our managers, what processes you own and manage and what are you doing to
improve them. You would be wise to have an answer ready for that day”. Many
off-site managers did not get the point made by the technician, but a few did
and were awakened by the comment.
This
new installation demonstrated that all levels of the organization can and
should be involved in continuous improvement of the process to satisfy customer
needs, lower costs, and make work easier for everyone. The role of management here is to lead the
way for improvement work at all levels.
This
concept raised the question of the leadership for improvement at levels above
that of the plant management. The
routine relationship between the plants and management levels above the plant
was concentrated on reporting the operational problems of the process,
including safety events. During this
time, the erratic operation of the process was deemed to be the result of
incapable employees and their immediate supervision. The net result of this relationship was
frustration at the lack of expectation of a plan for correcting the systematic
problems of the operations. It appeared
that the communication of plant performance to headquarters management was to
assure that no manager was guilty of not knowing the performance of the plant,
rather than not knowing how to improve the plant. Still no case for action.
Leadership
on quality at the top dropped dramatically with the next shift in assignments
and the special quality development team structure was shifted with a change in
personnel.
The
quality development group which was formed to report to the management
committee now reported to the executive vice president for a period of several
years. As part of organizational changes, the development group was moved to
report to one of the product vice presidents, and then, because none of the
product vice presidents knew how to direct and lead what seemed to them like a controversial
company-wide effort to improve manufacturing processes, the group was
reassigned to the Human Resources Department where innocuity was assured.
Eventually
and slowly, brick by brick, the dismantling of the quality thrust had
begun. The buildup of the philosophy of
quality and continuous improvement to its modest extent after more than
twenty-five years took less than one year to effectively dismantle the progress
back to a state that existed many years earlier. The effort to add quality and continuous
improvement to the strategy of the whole company ended when the development
group was assigned projects unrelated to quality and continuous
improvement.
The
decrease in attention to adding quality improvement as one of the business
strategies now took on the characteristics of “just another program” in the
eyes of the organization.
From
the reporting of operations problems and the apparent lack of interest in process
improvement to correct those problems, the next question focused on
measurement. The hypothetical question
was asked, “what key measures of process performance were reported
statistically to top management on a real-time basis?’ There were none. That’s
because the measurement system at the working level consisted of events whereas
the measurement system at the top level of the organization involved key
parameters including money. There was
never a connection between the two systems of measurement, and we considered
this to be a fault.
Another
hypothetical question: What might the interest in quality and process
improvement be if the top level of management had a real-time meter of the cost
of poor quality in the operations? The emphasis on improved process stability
would jump overnight, it was concluded.
The
logic drove the thought processes right back to fundamentals. If management wants to see improvement,
measure the condition—but measure it at its source in terms of what
matters---and that is the company’s key parameters including money. Goals translation failed because it was
translated down the organization and arrived on the workforce floor in terms
that meant little to them. Performance measurement translated up to top
management in the language of money would be a success. Measurement of the
condition is critical because the leadership response to the need for process
improvement is driven by the incentive to improve and without measurement the
level of needed improvement is unknown.
Parallel
thinking applied here: what might have been the interest in measurement if this
dilemma of lack of measurement of performance was documented and leaked to the
local newspaper? Just like twenty years before, the story might make it to the
Corporate Offices.
Here’s
what we learned from recalling the 25 years of effort, all condensed into a few
items which describe the important and necessary conditions within an
organization for it to function with quality as one of its organizational
strategies---none can be ignored at the risk of others. But be aware, that the
following are the necessary conditions for success but getting there still
depends on management’s willingness and ability to provide the leadership.
Dr.
Deming’s point about top management involvement was beginning to make sense.
Looking
back over 25 years of effort in the field of quality and continuous improvement
the structure and the capability of the organization were overlooked as contributing
factors to the inefficacy to tackle process improvement projects. Perhaps this is a common fault in that not
much is known about the subtle effects of organizational structure on
productivity. Halfway through this
journey of 25 years, the organization was restructured adopting a matrix design
where the lead was given to product lines.
When this shift occurred, a major effect occurred: near minimization of
process technology in the matrix structure coupled with a drift to avoid
looking outside our industry for innovative solutions to process problems. With this structural change, the new matrix
management lacked process knowledge and interest, and the technology resources
were lost. The core process part of the
organization was left in a stranded condition.
This
distillation of the events would suggest the following conditions necessary for
an organization to operate with one of the strategies being quality
improvement.
·
All
members of the organization must understand the operational definitions of a
few keywords used in quality improvement.
New jargon should be minimized but some words must be defined---one such
word is QUALITY.
·
New
processes introduced to the organization must be defined in their technical
terms; input and output definitions and measurement methods, controls
necessary, and services needed. Process
and design engineers must deliver a turnkey operation to the operations
organization. Business specialists and staff groups must follow the same rigor
in instituting new business processes.
·
The
performance of processes must be measured. The units of measurement must be
tailored to the part of the organization—such as process variance at the
working level, and financial at the upper level. Without accurate measurement,
there can be no organized effort to make improvements. Measurement systems must be capable in and of
themselves and have meaning at every level in the organization. Process
performance measures must be able to be converted into relevant terms at any
level in the organization thus assuring one language of measurement. The
Information Technology function now has a more distinct role in process
improvement, beyond that of mere automation.
·
Managers
perofrming as leaders must formulate plans for improvement, develop projects
for improvement and dedicate resources to improving process performance to
achieve stable processes. In a large organization, this will require a shared
process improvement function because of the many common processes. Continuous improvement successes must be
applied across the organization.
·
All
members of the organization must know how to use the relevant tools and
methodology for improving processes because their leader will expect them to
participate in improving their process. The use of the tools of improvement
includes the concept of testing a change before full implementation because not
every proposed change will result in an improvement. Everyone must be alert to the pitfalls of
changes that could contribute to lowering performance.
·
There
must be a constancy of purpose set by leaders at every level. Assignments at every level must be consistent
with the philosophy of constancy of purpose.
Leaders must be confident and able to demonstrate that constancy of
purpose directed at the quality of output assures not only customer
satisfaction but pride in workmanship both of which make other management
programs unnecessary and avoids some which may be destructive. Especially think
about safety, environmental control, and employee morale, all of which fall
under the umbrella of quality and continuous improvement.
·
The
organization must be capable of celebrating its successes and learning from its
failures. This means that the whole organization can celebrate a success that
occurred in another part of the organization because everyone sees value in
system improvement no matter where it occurs.
Failures in one part of the organization must play a role in avoiding
repetition in another part of the organization.
·
The
structure of the organization must not, in its zeal to focus on the customer,
neglect the focus on processes, for it is the effective process that best
serves the customer. Furthermore, the
organization must not be complex and should not have more layers than necessary
to manage the processes. The span of control can be increased with a better
focus on the process and the customer.
Excessive layers in the organization lead to bureaucracy and add
distance between leaders and workers.
If all this seems profound, why did we
fail to develop the leadership in management to adopt quality and continuous
improvement as one of the strategies of the company?
The simple answer may lie in the
conversation with an associate recently.
We raised the question of why many governmental and industrial
organizations appear to approach improvement haphazardly. The answer from Ken was quick and to the
point: “It’s all about measurement; we
engineers measure everything and measurement is a key component of the
scientific method for making improvements.
This method drives action.”
We now conclude that we failed to develop
and adopt a measurement system to reveal the cost of poor quality and make that
measurement the driving force for leadership in process improvement. We came close at one time about halfway
through the journey, by developing manager’s quality stations where results
were on display, but the measures were not relevant. We chose the few measures that were
acceptable because our quality stations were visible to visitors. The cost of poor quality could not be on
public display.
Alternatively, providing an internal
weekly report to top management showing the costs of poor quality for the last
week would have been effective.
The chapter on the history of a valiant
effort on quality was opened and closed for the last time.
[1]
Anything that can go wrong will eventually go wrong, an adage dating back to
1866 turned out to be the case here.
No comments:
Post a Comment