Total Quality Management

1

• Total Quality Management

2
What is Quality?
3
Old Quality vs. New Quality

• Difference between old quality (Rolls Royce,
  personal banker, ...) and new quality is that old
  was the work of craftsmen and the new is the work
  of a system (Toyota, Big Mac, Boeing Aircraft,
  Disney World, ...). The old is expensive, made
  for the few, using skilled hands, is beautiful
  and functionally based. The new reduces cost,
  made for the many by intelligent minds and should
  drive the economy and make business more
  competitive.

Toyota Commercial
4
Why care about quality

• increase productivity
• expand market share
• raise customer loyalty
• enhance competitiveness of the firm
• at a minimum, serve as a price of entry

5
Achieving high quality Is Difficult

• Only 36 of the firms felt that Total Quality
  programs boosted their ability to compete.
  Arthur D. Little Survey of 500 Firms
• Over 50 of firms rated their efforts D or F
  relative to increasing customer satisfaction,
  increasing market share, or reducing their
  cost. Rath and Strong
• Main Problem Achieving high quality is as easy
  to understand as losing weight and quitting
  smoking and is as difficult to do.
• Steve Schwartz, IBM MDQ VP

6
Why Quality is so difficult to do?

• Quality can only be defined in terms of an agent
  (a judge of quality).
• One has to translate future needs of the user
  into measurable characteristics

7
Service Industries are particularly Difficult

• Reasons
• High volume of transaction
• Immediate consumption
• Difficult to measure and control
• More labor intensive
• High degree of customization required
• Image is a quality characteristic
• Behavior is a quality characteristic

8
Quality Gurus

• Deming The father of the quality movement.
  Scientific approach to quality
• Juran Quality by design
• Crosby Quality is free

9
Demings Seven Deadly Diseases

• Lack of Constancy of purpose
• Emphasis on short term profits
• Evaluation of performance, merit rating or annual
  review of performance
• Mobility of management
• Running the company on visible figures alone
• Excessive medical costs
• Excessive costs of warranty fueled by lawyers
  that work on contingency fees
• Interview with Deming

10
What is TQM??

• The essence of Total Quality Management is a
  common sense dedication to understanding what the
  customer wants and then using people and science
  to set up systems to deliver products and
  services that delight the customer.

Greg Hughes President ATT Transmission Systems
11
Basic Concepts of TQM

• Customer Focus
• Continuous Process Improvement - Kaizen
• Employee Empowerment Everyone is responsible
  for quality
• Quality is free - focus on defect prevention
  rather than defect detection for it is always
  cheaper to do it right the first time
• Benchmarking Legally stealing other peoples
  ideas
• Customer-Supplier Partnerships
• Management by fact..by numbers..by data
  Balanced scoreboard (financial, customer,
  process, learning)

12
Quality in U.S. vs. the Japanese

• U.S. conforming to the requirements at the least
  cost
• Japanese joint responsibility to make the end
  customer happy

13
I met the requirements
Combative non collaborative relationship
14
Creating the Best Vehicle/Systems with All the
People All the Suppliers All the Time
YOU meet the requirements!
Lets create the best Vehicle and Systems
together.
SOR
Partnership - Collaborative relationship
15
Strength of USA vs. Japan
Concept
Good Innovative Ideas
Good Implementation
Strength of Japanese Mfg
Strength of USA Mfg
KAIZEN
Time
Good Ideas, Good Implementation are the goals of
everyone in the automotive industry
16
Seven Basic Quality Tools To improve Process
Quality

• Scatter Diagrams Plot data on a chart no
  attempt is made to classify the data or massage
  it
• Pareto Charts Organize data on a histogram based
  on frequency from most prevalent to least. Help
  identify major causes or occurrences (8020 rule)
• Check Sheets Easy way to count frequency of
  occurrence by front line workers
• Histograms Categorize data is cells and plot
  (see if any patterns emerge)
• Run Charts Plot data as a function of time
• Cause and effects Charts fishbone diagrams are
  used to identify the root causes of a problem
• Control Charts are statistical tools used to
  determine if the variation in results is caused
  by common or special events

17
Failures in O-rings
18
Graph Fit of O-ring failures
19
Full O-ring data including no failures
20
RUN CHART
TRANSACTION TIME
Time of Day
21
Data Collected From Check Sheet

• Time Range (in secs)
• 44-50
• 51-57
• 58-64
• 65-71
• 72-78
• 79-85
• 86-92
• 93-99
• 100-106
• 107-113

• Frequency
• 1
• 4
• 17
• 12
• 14
• 19
• 18
• 11
• 3
• 1

22
A Histogram
23
Be careful of Cell Size
24
Pareto Chart (80-20 Rule)
Further info on Pareto Charts
25
Pareto Diagrams

• Purpose
• helps organize data to show major factors
• displays data in the order of importance
• organize based on fact rather than perception
• To construct
• use data from a check sheet or similar instrument
• analyze data to determine frequency
• identify the vital few
• calculate percentages
• add percentages to find vital few (80)
• draw cumulative curve
• Typical Application
• display relative importance of different factors
• choose starting point for problem solving
• monitor success
• identify basic cause of a problem
• use a selling tool to gain support

26
Teller
Processes
Sequence of activities
Fatigue
Too many steps
Training
Control functions
Attitude
Processing Delays
Too much downtime
Not user friendly
Slow response time
Fishbone Diagram aka Cause Effect Diagram
Computers
27
Cause and Effect DiagramFishbone Diagram

• Purpose
• visual display of information to identify root
  causes rather than symptoms.
• To construct
• determine the issue and write problem statement
  in a box to the right of diagram
• find the main causes and write them on branches
  flowing to the main branch (method, equipment,
  people, material, environment, customer
  expectations, money, management, govt.
  regulations)
• identify all possible causes and write them on
  the diagram as sub-causes in each category
• Typical Application
• determine the real cause of the problem
• check the potential effects of a solution
• Fishbone Diagrams Explained
• 5 Whys problem solving technique

28
Mizenboushi and GD3 Concepts
Robust Design - keep Good Designs -
minimize change
Good Design
Find Problems
Prevent Problems
GD3
Good Dissection
Good Discussion
DRBFM
DRBTR
Address any potential issues up stream at Design
Phase
29
Quality Focus At the Design Stage

• Quality from the start
• Directs attention to Change
• Change potential to have problems
• Directs attention to Interfaces
• Most defects occur at the interface

Focus on Change Points Interface Points
30
No change No Problem

• Examples
• Design change
• Packaging environment
• change
• Usage environment
• change
• New manufacturing
• process
• New supplier

Change Points have the highest potential to
introduce defects
31
DRBFM Example

• Tire Pressure Monitoring System
• Changing the sensor from Aluminum Valve to Rubber
  Valve.

• Purely for cost reduction purposes... System
  Performance is the same.

Simple change What could go wrong?
32
Interfaces

• Interfaces (Interfaces where issues can brew
  and surface later)
• Customer to Supplier
• Department to Department
• System Interfaces
• The Crash sensor failure on Honda Minivans

Interface Points have the highest potential to
introduce defects
33
Design Review By Failure Modes (DRBFM)Basic
Concepts

• Before and After Description of the Change
  Point
• Describe the Potential failure modes
• Describe the Design Countermeasures
• Target Testing of the change points and
  Countermeasures Only

Design techniques to uncover defects at the
design stage Up stream
34
DRBFM
DRBTR
Design
Verify/Validate
Design
Changes
Test Result (Change in product due to test
Cracks,Leaks, etc.)
Focus on Implementation
34
35
Where do failures occur

• Design Phase (Suppliers are Up Stream)
• Production
• In the field
• Where is it cheapest to detect failures?
• Example
• Replacing a four crash sensors by a single one ..
  

36
When Failures Occur!

• Why did the failure happen?
• Symptoms vs. Root Causes
• Root Causes (Investigate the whole chain)
• Suppliers/Component failure
• Design
• Manufacturing
• Change management
• Why were not able to detect it?

• Rootcause Analysis
• Why Occurred?
• Why Not Detected?

37
Failure Detection 5Ws-2Hs

• Who
• Where
• When
• What
• Why
• How was the problem found?
• How can we isolate it? Turn On / Turn Off

Rootcause Analysis Methodology
38
Failure Isolation KT Analysis Is - Is Not

• Why is this design and not the other similar
  design
• Why this plant and not another plant
• Why this operator and not the other operator
• Why in winter and not in the summer
• Why this computer and not the other computer
• Why in this model and not in other models

Rootcause Analysis Methodology
39
Finding the root causes of a problem is not Fault
Finding/Criticism.

• To find problems is not fault finding/criticism.
• To find problems is a creative act, same as
  innovation.
• We should never stop at only finding problems,
  but also develop a systemic corrective action
  plan... FIX THE PROCESS that created the problem
  identify detection algorithms
• We never forget that every job should relate
  directly to improving a product. Other jobs are
  nothing but waste, e.g., only to check, to
  inspect, etc.
• Everyone should readily accept help from review
  participants.

40
Summary - Concepts

• Quality all the time by everyone from an end user
  prospective
• Address issues up stream. Address product and
  process defects at the design stage
• Fixing problems usually involves fixing the
  systemic process issues that caused the problem
  Reoccurrence Prevention
• Focus on Implementation
• Focus on Change Points and Interfaces