Introduction to TQM

Quality thinking

TQM is management and control activities

• Top management leadership

• All employees - all departments, all levels

• Quality Assurance built into products and services   -paraphrased from Japanese Society for Quality

TQM - The "home-court advantage"

In business, an organization should try to get the home court advantage using TQM. The activities for each employee, including top management, should be geared toward the same purpose, which is often referred to as the vision. If people in a company do not have the same purpose, the organization will be weak. And just as an NBA team sometimes loses fans as it loses games, a company that loses its purpose, misses its goal, and makes too many mistakes, will lose customers. If the organization is to achieve its goals, all people related to the organization must have activities for achieving their purposes. For example, in basketball, the fans scream and clap to show their support for their team, and remain silent or boo the opposing team to discourage them. Each player brings the ball from the front court to the back court. The head coach calls time-out and observes the team’s situation. Every activity is for achieving one purpose. The bigger the number of people who have the same purpose, the stronger the chances are of achieving the purpose. If 20,000 employees are committed, the organization will be stronger than if 10,000 are committed.

Many TQM experts use special terms to explain TQM, such as the voice of the customer table, QFD implementation, Cross-Functional Management, and Hoshin Planning (some terms are mis-defined). These terms are confusing to many people, particularly novices, in many companies.  The initial, simplistic step of implementing TQM is to determine the organization’s goals or obstacles according to the dream oriented or problem oriented approach. In Japan, almost every company uses the problem oriented approach during the initial phases of TQM. Everyone, from top management to the line workers in an organization, determines and defines the problems. Each employee defines the problems using a statement beginning with "I don’t like to . . ." This is a key issue. Jobs listed this way by each employee are not only the employee’s problems, but also the organization’s problems. Using an approach like this is good because it not only helps pinpoint problems, it also opens up discussion within the organization.

In Japanese companies that implement TQM and have won the Deming Prize, all employees, including the president or executive director, discuss their problems using methods like the real-life dream and problem-oriented approaches. Many people in this country tell me that this happens only in Japan, and that it cannot happen here. Do you think this is just the Japanese way or that they can do it only because they’re Japanese? Don’t forget that TQM started in the United States, not Japan. TQM’s home court is America. 

Copyright © 1994-96 by Satoshi "Cha" Nakui. All rights reserved. Used with permission.

Diagramming the Home Court Advantage

W. Edwards Deming: (1900-1993) 
The Father of Modern Quality

To achieve this level of performance requires more than a good philosophy - the organization must change its behavior and adopt new ways of doing business. This is what Dr. Deming preached to the Japanese in 1950, and in the 1980s and 90s until his death, in America. Deming's approach were amply summed up in his famous 14 Points. These exhort management to rational action instead of merely sloganizing quality and blaming workers for issues beyond the workers' control. We call this "walking the talk." Deming formulated this into his System of Profound Knowledge™ by which management could change itself only with a view from the outside; the system cannot understand itself.

"We have learned to live in a world of mistakes and defective products as if they were necessary to life. It is time to adopt a new philosophy in America."

Deming based much of his work on earlier work done by Walter Shewhart on statistical quality control (SQC). Shewhart is considered the father of quality control. SQC uses control charts to identify and control sources of variation in manufacturing processes. In TQM, we apply the principles of controlling the quality of machine-based factory operations to controlling the quality of people-based management operations. The principles we will learn are:

• Variation is inherent in all processes - mechanical and human.

• The Plan-Do-Study-Act (PDSA) cycle developed by Dr. Walter Shewhart helps us manage the effects of variation. This is the scientific method applied to problem solving which has us plan and test our improvements, make adjustments, and then standardize them to prevent recurrence. PDSA or PDCA (check) is fundamental to TQM.

• To do PDSA, we must collect data relevant to the process and understand what this data means.

• Understanding this data helps prioritize and direct improvement.

• Improvement increases stakeholder (shareholder, employee, customer, community) satisfaction both now and in the future.

Japanese TQM Models

At first, few but the Japanese took Deming seriously. Known for his legendary attacks on executives and compassion for the worker, the Japanese sense of responsibility to one's superiors and subordinates made it easier to accept Deming's message that management's role was to provide the optimal conditions for the workers to do the best job. The Japanese then extended Deming's teaching to many dimensions of management. Here are TQM models from the Japanese Society of Quality and the Japan Standards Association that show how extensive their definition of TQM is. The Japanese identify three major dimensions of TQM: Daily Management, Hoshin/Policy Management, and Cross Function Management. We will study these in detail in the course.  They can be visualized with the following picture.

US TQM Models

Malcolm Baldrige National Quality Award and State versions

Keep US industries competitive in a global economy.

Six Sigma

Hold business accountable for the cost of poor quality by providing a measurement system that ties customer satisfaction with product performance to design quality and process quality performance.

TQM Implementation Model

1. Process Improvement
   Manage variation Fix a known problem: QI Story
   Prevent recurrence: SPC, Kaizen
2. System Improvement
   Six Sigma
   Daily Management
   Quality Assurance
   Theory of Constraints
   Strategic Policy Management
3. Product Improvement
   Quality Function Deployment

Managing variation

Variation is a fact of life. It is random and miscellaneous. Thus, the same process can produce two things that are not alike. In the days of hand-crafted products, this could be accounted for by "fitting" things together. In modern industry where interchangeable parts are assembled into mass-produced final products, controlling variation is critical to customer satisfaction. This is one of the most important tasks a manager faces.

Dr. Walter Shewhart identified two kinds of variation, controlled and uncontrolled, and their characteristics.

Controlled variation:

• stable
• exhibits a consistent pattern over time
• results of a process can be predicted with greater certainty
• process can be improved because outcomes of changes can
  be predicted

Uncontrolled variation:

• changes over time due to "assignable causes"
• can not predict results of process
• process cannot be improved easily since outcome of changes
  are unpredictable

Management's job is to manage variation in order to produce predicable results, such as quality, cost, and production schedules. Since all data sets contain random variation or noise, the noise must be filtered out, otherwise two kinds of mistakes could arise.

1. interpreting controlled variation (noise) as if it were a meaningful change (signal) and expending efforts to "fix" natural variation
2. interpreting uncontrolled variation (signal) as if it were noise and not recognizing when a change has taken place.

To manage the variation in a process, historical data must be analyzed to identify which changes  are noise and which changes are signals. To do this Dr. Shewhart created the control charts. In manufacturing, we use a variety of charts for Statistical Process Control (SPC). 

Deming's Red Bead Experiment

Dr. Deming asks six willing workers to manufacture white beads by scooping them out of a box containing an 80:20 mixture of white to red beads. They are given a target of no more than 3 red beads per day. But the number of red beads produced was as follows:

Willing Worker	Day 1	Day 2	Day 3	Day 4
Mike	8	11	6	7
Leon	14	10	8	11
Karen	7	10	11	5
Bob	11	10	6	10
Melvin	7	12	6	13
Paul	14	7	7	14
Totals	61	60	44	60

Eventually all the Willing Workers lost their jobs and the plant was closed. Why were they fired? Was it their fault? 

Control charts show whether the results were inherent to the process, which the workers did not create and could not control, or whether the results were due to some assignable cause which the workers might have been able to correct themselves. The individuals chart shows the production of each willing worker for each shift. The moving range shows the change from worker to worker.  Based on past performance of the bead plant, the variation in the number of red beads remains consistent from worker to worker, shift to shift.		Individuals Chart  x-bar
		Moving Range  mR

For management, we will look at constructing a chart that tracks individual values. Here are the steps.

Constructing the individuals control chart (X-bar mR Chart)

1. Collect data. Always show the data with your charts.
2. Calculate and plot the moving range (mR). The moving range is simply to distance between two successive data points. So if the data points are 5, 2, 7, 6, the range is 3 for the first two data points, 5 for the second and third points, and 1 for the third and fourth points, and so on. The moving range is 3, 5, 1, etc. Plot these on a chart and connect the data points with a line. This is the beginning of the mR chart.
3. Calculate the average moving range (R-bar) by summing the moving range numbers and dividing by n-1. The average moving range in the above example is (3+5+1)/3 = 3. Draw and label R-bar on the mR chart.
4. Calculate the Upper Control Limit (UCL) for mR by multiplying the average moving range by 3.27. Draw and label the mR UCL on the mR chart.
5. Plot your original data points and connect with a line on a separate X chart. Calculate the average of X and draw in as the center line (CL).
6. Calculate, draw, and label the Upper and Lower Natural Process Limits of X.
   Upper Natural Process Limit = X-bar (average of X) + 2.66 x (R-bar)
   Lower Natural Process Limit = X-bar (average of X) - 2.66 x (R-bar)
7. Show both charts together (X-bar mR Chart) to interpret the results. Individual points outside the control limits indicate assignable cause. If the process is stable, points will be inside control limits. There are additional evaluation tests for stability given in MJ2.
8. Only a stable process can be improved with predictable results.

Fieldwork [due next class] (5 pts. max.)

It is your third year as production supervisor of a local plant. You have been called to the plant manager's office to explain why your July level of in-process inventory of 28 units has exceeded your target of 19.7 units, and in fact is at an all time high. 

This is 42% above target, and a further review of your performance this past year shows that you are 12% above last July and 9.6% above target for this year already.

To better understand the situation, you gather some historical data since you were hired (table below). First, construct one individuals control chart (X-bar and mR) per above steps based on the past two full calendar years. Then, in order to determine if the current year work-in-process is in control, continue plotting the current year's inventory levels on the same chart. Do the current year's inventory levels fall within the historical control limits of the first two years?  Has something outside the process changed? What does this say about the process? What needs to be changed in order to see an improvement?

In-Process Inventory [Wheeler 1993]
Year	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
1	19	27	20	16	18	25	22	24	17	25	15	17
2	20	22	19	16	22	19	25	22	18	20	16	17
3	20	15	27	25	17	19	28

The purpose of this assignment is to have you make an X and mR chart and analyze the results so that you will be able to teach this to plant and labor workers. Please do this chart by hand, not in an SPC program or spreadsheet,  using the constants given above.

Extra Credit (5 pts. max.): Pick a controversial issue, and use control charts to validate whether the current status represents a "signal" meaning that the system has changed, or the current status is within historical limits. Ex. A study of "foot & mouth disease" occurrence in England last year showed that the number of new cases reported each day, once false cases were subtracted, showed no significant increase in the number of reported cases for several decades. You might wish to validate such controversial issues as global warming, student test scores, stock market changes, etc.

Optional resources

Software:

Variation. Great program to demonstrate the problem of trying to control variation with an uncontrolled process. Two versions: vari.exe and tennis.exe

Funnel. In addition to the red bead experiment, Dr. Deming also showed the futility of trying to drop a round ball through a funnel and hit a target consistently.

Books:

Deming, W. Edwards. 1982, 1986. Out of the Crisis.  MIT-CAES. ISBN 0-911379-01-0

Deming, W. Edwards. 1993. The New Economics: For Industry, Government, Education. MIT-CAES. ISBN 0-911379-05-3

Inamori, Kazuo. 1995. A Passion for Success: Practical, Inspirational, and Spiritual Insights from Japan's Leading A Passion for Success: Practical, Inspirational, and Spiritual Insights from Japan's Leading Entrepreneur. McGraw-Hill Inc. ISBN 0-07-031784-4

Juran, J.M. ed. 1995. A History of Managing for Quality. ASQ Press. ISBN 0-87389-341-7

Kohn, Alfie. 1993. Punished by Rewards: The Trouble with Gold Stars, Incentive Plans, A's, Praise, and Other Bribes. Houghton Mifflen Co.   ISBN 0-395-71090-1

Wheeler, Donald J. 1993. Understanding Variation: The Key to Managing Chaos. SPC Press. ISBN 0-945320-35-3

Wheeler, Donald J. and David S. Chambers. 1992. Understanding Statistical Process Control. SPC Press. ISBN 0-945320-13-2

Instructor's classroom slides

Student: tqmintro.pdf

Faculty and Professional: tqmintro.ppt

© 1993-2004 Glenn H. Mazur. Comments may be addressed to glenn@mazur.net