You are on page 1of 43

Six Sigma Approach to Quality

Six sigma is a new approach to process


control first practiced by companies such as
General Electric and Motorola to achieve near
zero defects in processes
Elements of six-sigma
Understanding customer needs well
Appropriate and disciplined use of data and
statistical tools
Statistical analysis and a closer attention to
managing and improving business processes using
a set of tools

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Why near zero defects


Examples of 99% quality
Sl.
No.

Process
Description

Impact of 99% quality

Guest Check
out

Every three days one guest checks out


without paying the bill

Quality of food
served

Every day 20 orders of fried eggs have


human hair in them

Facilities up
keep

Every day 15 tables in the restaurant


have soiled linen

Training & Skill


Development

Every day 250 plates are broken

Order taking &


Delivery

Every day the drinks of 40 guests gets


mixed up

Laundry

Every day 20 guests do not get back the


right laundry

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

How is it different from


traditional approach?
A new metric, Defects Per Million Opportunities, to
predict/assess the quality of a business process
A new methodology, DMAIC (Define-MeasureAnalyze-Improve-Control), to ensure that very
high levels of quality could be assured in the
chosen business processes thereby generating
favourable outcomes to both the business and the
customers
An organizational framework for ensuring the
above outcomes are generated on a sustained
basis

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Defects Per Million Opportunities


(DPMO)
It indicates how many defects a process
generates in a million opportunities
While service industries find it convenient to use
DPMO, manufacturing organizations use Parts Per
Million (PPM). Both are conceptually same.

If in a process
k Denotes the number of opportunities for making a
defect per unit of execution of that process
n Number of units of observation of the process
d Number of defects that occurred in that process
during the observation
The defects per opportunity = d * 1,000,000
k *n

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

DMAIC Methodology
Define

Measure
Analyze

Control
Improve

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

DMAIC Methodology
Define

Define the problem, the requirements, project scope, project


charter
Set goals for improvement

Measure

Identify variables to be measured, the type of measurement


Data collection and synthesis

Analyze

Develop a set of tools for analysis


Apply graphical tools of analysis
Identify possible sources of variation and vital few root
causes
Explore means of eliminating them

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

DMAIC Methodology
Improve

Generating and validating improvement alternatives


Creating new process maps for the process

Control

Develop control plan


Establish revised standard measures to maintain
performance
Develop relevant raining plans to maintain standards

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Organization for six sigma


Process Owner
supervisor or a manager who takes responsibility for
various steps of a process that is responsible for
delivering some output to the customer. It could be
the in a particular work area where the improvement
project has been identified

Team Leader & Members


Team leader (the project leader) and the members
will comprise of the employees in the chosen work
area
They will have day-to-day operational control of
activities
In six sigma, three terminologies are used to indicate these organizational entities. This
includes Master Black Belt, Black Belt and Green Belt. The depth of training and experience
differentiates these three.
Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Organization for six sigma


Six sigma coach
A consultant or a senior person in the organization who
offers expert knowledge on various aspects of six sigma.
This includes statistical tools, process design & analysis,
change management, small group improvement, use of
QC tools for improvement etc.

Sponsor
A member of the senior management who oversees the
overall progress and implementation
Helps the team refine the project scope, sorts out issues
cutting across other parts of the organization, approves
projects and provides the necessary support in terms of
resources

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Statistical Process Control


Collective set of tools & techniques used to
develop a quality assurance system when
business processes exhibit variations is known
as Statistical Process Control (SPC)
Key issues addressed in SPC based quality
assurance system:

How does one ensure that the random events are


indeed rare events?
How do we know whether the observed changes are
due to random variations or assignable causes?

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Common & Assignable Causes


Two types of variations occur in business processes
Chance variations due to common causes
Common causes are those due to random events that
cannot be controlled
Non-random variations due to assignable causes
When observed variations are not statistically found to be
due to random events, it clearly points to the existence
of assignable causes

Errors due to operator skill level differences


Changes in the operating condition of an equipment
Changes introduced in the standard operating procedures

In the case of assignable causes, the system will


drift from desired level of performance over time

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Quality Assurance using SPC


Some terminologies
Designed Standard
Centre of specification
limits (Target)
Upper Specification Limit
(USL)
Lower Specification Limit
(LSL)
(USL LSL): Desired
tolerance
This represents the voice
of the customer

Status of process
Centre of the process
(Process Average)
Upper Control Limit
(UCL)
Lower Control Limit
(LCL)
(UCL LCL): Spread of
the process
This represents the voice
of the process

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Control Chart

A generalised representation
Upper Control Limit (UCL)
Plot of
sample data
Process Average

Lower Control Limit (LCL)

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Setting up a process control


system
Choose the characteristic
for process control

Choose the
Measurement method

Choose an appropriate
Sampling procedure

Choose the type of


Control Chart

Calculate
control limits

Plot the data & Analyse

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Characteristics for process control


Some examples
Sl. No.
1

Type of Applications
Component Manufacturing

Characteristic for Measurement

Conformance of physical measurements of


components and sub-assemblies to specifications
Conformance to operating characteristics of
machines and other resources involved in the
process

Final Assembly

Number of defects in the product


Conformance to test specifications
Number of missing elements

Process Industries

Temperature, Pressure and Heat specifications


Conformance to product specifications
Conformance to equipment specifications
Vibrations and other variations in equipments and
sub-systems
Conformance to specifications of the automation &
control system

Service Systems

Number of defects in various business processes


Errors in processing documents
Conformance to waiting time/lead time related
specifications

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Choosing a characteristic
Examples from service industry

How long does it take to complete the


patient admission process in a hospital?
How dissatisfied were the customers when
they stayed in a hotel for a holiday?
Does our process of understanding
customer requirements needs
improvement?
How capable are my sub-processes in
software development?

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Measurement Methods
Attribute Based

simple clustering of the characteristic into a few categories


(such as good or bad)
Two frequently used attribute measures are:
Proportion of defects (denoted as p)
Number of defects (denoted as c)

measurements are easy to make, quick & less expensive,


but reveal very little information about the process

Variable Based

Detailed observation of the characteristic (such as length,


diameter, weight)
measurement will be expensive and more time consuming
but will provide a wealth of information about the process

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Types of Control Charts


Variable based control chart
X bar and R Charts

Attribute based control chart


P chart
C chart

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Criterion
Sub-group size
Sub-group measures

Notations
for control
charts

Number of subgroups
Mean of the
sampling distribution
of the sub-groups

Notation
n
Variable measures
Mean - X
Range R
Attribute measures
Proportion of defects p
No. of defects c
m
Variable measures
Mean - X

Range R

Explanation
Each sub-group consists of n
observations
Based on the n observations, each of
these measures can be computed for
each sub-group

m measurements are used for


establishing the sampling
distribution and control chart
These means are nothing but the
process average (center line) of the
respective control charts formed
using each of these measures.

Attribute measures
Proportion of defects
p

p
m

No. of defects - c

Standard deviation of
the sampling
distribution of the
sub-groups

Attribute measures
Proportion of defects
p

p (1 p )
n

No. of defects - c c

Three standard deviations on either


side of the center line are the control
limits for the respective charts.

X bar and R Chart


Setting control limits
Computations for the X chart
X

Process average or center line:

Upper Control Limit : UCL X X A2 R


Lower Control Limit: UCL X X A2 R
Computations for the R chart
Process average or center line:

R
m

Upper Control Limit : UCL R D4 R


Lower Control Limit: UCL R D3 R
* The values A2, D3 and D4 can be read directly from tables (see table 6.3)

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Coefficients for computing LCL and UCL


in X-bar and R charts*
Sample size (n)
2
3
4
5
6
7
8
9
10

A2
1.880
1.023
0.729
0.577
0.483
0.419
0.373
0.337
0.308

D3
0
0
0
0
0
0.076
0.136
0.184
0.223

D4
3.268
2.574
2.282
2.114
2.004
1.924
1.864
1.816
1.777

Source: Juran, J.M. and F.M. Gryna, (1995), Quality Planning and Analysis, Tata McGraw-Hill, 3 rd Edition, New
Delhi, pp 385.
Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

X bar Chart
An example

X-bar Chart

Mean Diameter (cms)

Sample Means

Centre Line

UCL

LCL

12.50
12.49
12.48
12.47
12.46
12.45
12.44
12.43
12.42
12.41
12.40
12.39
12.38
12.37
12.36
12.35
12.34
1

7
8
9
Sample Number

10

11

12

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

13

14

15

R Chart

An example
R Chart

Mean Range (cms)

Sample Range

Centre Line

UCL

LCL

0.28
0.26
0.24
0.22
0.20
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00
1

Sample Number

10

11

12

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

13

14

15

P chart

Setting the control limits


Computations for p chart
Process average or center line:

p
m

Upper Control Limit : UCL p p 3

p(1 p)
n

Lower Control Limit: LCL p p 3

p (1 p )
n

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

P Chart

An example
p Chart
p

Centre Line

UCL

LCL

0.21

Proportion of defects

0.18
0.15
0.12
0.09
0.06
0.03
0.00
1

6
7
Sample No.

10

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

11

12

C Chart

Setting up the control limits


Computations for c chart
c

Process average or center line: c


m

Upper Control Limit : UCLc c 3 c


Lower Control Limit: LCLc c 3 c

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

C Chart

An example (with an outlier)


c Chart
c

Centre Line

UCL

LCL

24

Number of defects

21
18
15
12
9
6
3
0
1

5
6
Sample No.

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

10

C Chart

Is this process behaving normally?


c Chart
c

Centre Line

UCL

LCL

24

Number of defects

21
18
15
12
9
6
3
0
1

5
Sample No.

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

10

When to stop the process?


15 points in a row in Zone C
+3
Zone A
+2
Zone B
+1
Zone C
Zone C
-1
Zone B
-2
Zone A
-3
Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

When to stop the process?

14 points in a row alternating up and down


+3
Zone A
+2
Zone B
+1
Zone C
Zone C
-1
Zone B
-2
Zone A
-3
Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

When to stop the process?


Additional rules

One point beyond Zone A


Nine points in a row in Zone C or beyond
Six points in a row, steadily increasing or
decreasing
Fourteen points in a row, alternating up and
down
Two out of three points in a row in Zone A or
beyond
Four out of five points in a row in Zone B and
beyond
Fifteen points in a row in Zone C
Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Predictive capability of processes


Which process is better?

Spread of a
process is
indicative of
its capability

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Predictive capability of processes

Offset

Which process is better?

A process that is
aligned closer to
the desired
target is likely to
be more capable

Process B

Process A

LSL

Target

USL

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Process Capability
Process Capability is defined by the spread of
the process
Potential capability (Cp) is defined as the
ratio of the difference in specification limits
to the process spread
Cp =

Specification Range (USL LSL)

Pr ocess Capability
6

Actual capability (Cpk) takes into


consideration the extent to which the
process has deviated from the desired target
Cpk =

Pr ocess Centre LSL


Min

USL Pr ocess Centre

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Process Capability & Defects


Process Capability Index
(Cpk)

Total Products outside twosided specification limits

0.25

453,255 ppm

0.50

133,614 ppm

0.60

71,861 ppm

0.80

16,395 ppm

1.00

2,700 ppm

1.20

318 ppm

1.50

7 ppm

1.70

0.34 ppm

2.00

0.0018 ppm

Source: Quality Planning & Analysis, Juran & Gryna, Chapter 17, 3e
Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Six sigma quality

A graphical representation

Cp

(USL LSL)
2 (USL LSL) 12 A spread of 6
6

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Acceptance Sampling
A method in which the decision with regard to
acceptance or otherwise of a lot of incoming material
is based on the observations made using a sample
Setting up a system for acceptance sampling
requires addressing the following questions:
What is the size of the sample required for assessing the
quality?
What should be the criteria for acceptance of the lot?
Is it enough to have just one sample or is it desirable to go
for multiple samples
How do we assess the impact of the choice of the above
parameters on the quality level in the long run?

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Sampling Plan
A sampling plan describes
the Lot size (N)
size of the sample to be assessed (n)
the acceptance number (c): the number of defects
permissible in the observed sample

Therefore, a single sampling plan is denoted by

(N, n, c)

In the case of multiple sampling plans, for each


of the plan, the sample size (n1, n2 etc.) and
the acceptance number (c1, c2 etc.) needs to
be specified
Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Acceptance Sampling
Operating Characteristics

Acceptable Quality Level (AQL): the percent defect that the


buyer is willing to tolerate in the lot delivered by the supplier
Lot Tolerance Percent Defective (LTPD): the worst quality
beyond which the manufacturer is not willing to accept the
incoming lot
Producers risk: the risk supplier faces in getting his/her lot
rejected when indeed it was conforming to the agreed terms
Consumers risk: the risk that a manufacturer faces of
accepting a lot with poor quality beyond his/her tolerance
level
Operating Characteristics (OC) Curve: A graphical
representation of a sampling plan that reveals the above
aspects of a sampling plan

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Operating Characteristics Curve


Example 18.5
1.000
Producers
Risk 0.900

OC Curve for n=500, c=2

Probability of acceptance

0.800
0.700
0.600
0.500
0.400
0.300
0.200

0.20

0.18

0.16

0.14

0.12

0.10

0.09

0.08

0.07

0.06

0.05

0.04

0.03

0.02

0.01

0.00

Consumers 0.100
Risk
0.000

Percent defective

AQL

LTPD

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Statistical Process Control


Chapter Highlights

Six sigma is a new approach to process control based


on as set of principles that enable organizations
improve their quality to near zero defects level
Six sigma quality control differentiates itself from the
traditional quality control methodology on the basis
of three features:
A new metric, Defects Per Million Opportunities, to
predict/assess the quality of a business process,
A new methodology, DMAIC (Define-Measure-AnalyzeImprove-Control)
An organizational framework for ensuring the above
outcomes are generated on a sustained basis

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Statistical Process Control


Chapter Highlights

All processes exhibit variations on account of two causes


Common causes of variations happen on account of
random events
Assignable causes introduce variations that can be
detected and eliminated

Setting up a process control system involves


choosing a characteristic to control
identifying a measurement method and
developing a relevant control chart

A plot of the data helps a quality manager to assess if


the process is in control and also helps her to detect any
impending shift in process parameters

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

Statistical Process Control


Chapter Highlights

Superimposing the specification limits on the control


chart helps an organisation to assess the process
capability and the likely number of defects that the
process will produce. It also helps to target for a sixsigma quality improvement programme.
Acceptance sampling is used to accept a lot of items
based on the observed number of defects in the sample
drawn out of the lot
The performance of an acceptance sampling plan could
be judged by developing an Operating Characteristics
(OC) curve for the plan
An acceptance sampling design results in a certain
amount of risk for both the customer and supplier

Mahadevan (2010), Operations Management: Theory & Practice, 2nd Edition, Pearson Education

You might also like