BB-Day3

41
Six Sigma Black Belt Training DAY 3

description

Black belt course

Transcript of BB-Day3

Six Sigma Black Belt Training

DAY 3

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Agenda for Day 3

• Analysis Stage• Basics of Variation• Multi-Vari Chart• Component Search• Paired Comparison• FMEA• Sampling Distribution• Hypothesis Testing

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Analyze - Major Activities

• Identify patterns through Data Analysis.

• Analyze the root-cause through Cause Analysis.

• Validate the root cause.

• Implement local actions to eliminate root-cause.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Data Analysis – Tool Selection

CTQ Type Trend Chart Summary Chart

Variable

• Run chart• Xbar and Range chart• Individual and Moving Range chart

Histogram Box Plot Cross-tabulation

Attribute

P - chart NP - chart

Pareto chart Concentration diagram Cross-tabulation

Counts (Defects)

C - chart U – chart

Pareto chart Concentration diagram Cross-tabulation

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Cause Analysis - Tools

• Casuse and Effect Diagram

• Paired Comparison

• Why – Why Analysis

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Analysis Stage - Planning

• Data Analysis– CTQ– Statistical Tools : Trend and Summary– Frequency

• Cause Analysis– CTQ– Tool for understanding Causes – Root cause– Frequency

• Potential Causes - FMEA

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Analysis Stage Planning SheetProcess

Step CTQ

Data Type

Data Analysis Cause Analysis

Tool Frequency Tool Frequency

1 Write Order

Quality Attribute np-chart Daily

Data Errors Defects c-chart Daily

Booking Time Variable I&MR chart Daily

2Assemble Pizza

Assembly Defects

Defects c-chartDaily

Assembly Time Variable I&MR chart Daily

3 PackagePacking Defects Defects c-chart Daily

Packing Time Variable I&MR chart Daily

4Deliver Pizza to Customer

Right Pizza Attribute np-chart Weekly

Delivery Time VariableI&MR chart

Weekly

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Difficulty in Pinpointing Causes

Cause

Cause

Cause

Cause

Cause

CauseCause

(a) Single Cause

(b) One of Multiple Varying Causes

(c) Varying Complex Combination of Causes

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Robustness• Robustness means making the product and process

less sensitive to variation of the key inputs.• Robust products are products that are insensitive

to the manufacturing process, usage conditions, materials and their deterioration.

• Robust processes are processes that are insensitive to materials, environment, and variation in machine parameters.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Tools for Identifying VIPs• Top three causes account to over 80% of the

variation allowed within the specification limits.

• The top three causes are called Red X, Pin X and Pale Pink X (Dorian Shainin)

• The Tools for Diagnosing the top three causes:1. Multi-Vari Chart

2. Components Search

3. Paired Comparisons

• Some times two or all the above techniques are used to sequentially or simultaneously.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Multi-Vari Chart

• A multi-vari chart is a stratified experiment to quantify different sources of variation, namely- positional, cyclical or temporal.

• Multi-vari charts allow the total variation to be decomposed into components including the variation due to time to differences, stream to stream differences, and within-part differences.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Injection-molding Process

• A set of data collected from an injection-molding process making plastic connectors. These connectors are cylindrical in shape.

• Two parts were selected from each of four mold cavities every 3 hours.

• The outside diameter was measured at each end as well as at the middle of each part.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Within-Part Variation

Bottom Diameter

Middle Diameter

Top Diameter No difference

Taper

Thick ends

Thick top

or

Thick center

For an individual part any one of these patterns might result from noise. However, if the same pattern repeats over and over, then with-in part variation is contributing to total variation.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Conducting the Multi-vari Study• Do not make process changes while collecting data.The

sequence should not be randomized.• Multi-vari charts provide a snapshot of how the process is

performing. The single largest source of variation at that time can be clearly identified.

• A second study some time later may give different results. For this reason, one may want to repeat Multi-vari study several times, one should also try to identify the second largest difference. It might really be the most important difference.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Component Search

• Unlike Classical DOE, component search actually discovers and isolates unexpected controlling causes, be they main causes or interactions. And it achieves this isolation with much less total testing.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Prerequisites• The technique is applicable, primarily, in assembly

operations.• The performance (output) must be measurable and

repeatable.• The units must be capable of disassembly and

reassemble without a significant change in the original output.

• There must be at least two assemblies or systems with clearly different levels of output.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Four Stages of Component Search

• Ballpark – To determine if Red X, Pink X are among the

causes being considered. It also assures repeatability of the disassembly and reassembly process.

• Elimination– To eliminate all unimportant causes and their

associated interaction effects.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Four Stages of Component Search

• Capping Run– To verify that the important causes are truly

important and that the unimportant causes are truly unimportant.

• Factorial Analysis– To quantify the magnitudes of the important

main causes and their interaction effects.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Paired Comparisons• The technique is used when:

– Components or subassemblies in units cannot be disassembled or reassembled.

– There are several good and a few bad units that can be paired.

– A suitable parameter can be found to distinguish good from bad.

• The technique is applicable in assembly work or processes or in test equipment, where there are similar units, workstations, or instruments, respectively. It is useful in failure analysis also.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Step-by-Step Procedure for Paired Comparisons

1. Select one good unit and one bad unit.

2. Call this pair one. Observe in detail to note differences between these two units. The differences can be visual dimensional, electrical, mechanical, chemical etc. The observation techniques could involve simple or advanced techniques.

3. Select a second pair of good and bad units. Observe and note the differences, as in step 2.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Step-by-Step Procedure for Paired Comparisons

4. Repeat this search process with a third, fourth, fifth, and sixth pair, until the observed differences show a pattern of repeatability.

5. Disregard differences that show contrary directions among the pairs. Usually, by the fifth or sixth pair, the consistent differences will be narrowed down to a few factors, providing a strong clue for the major cause of variation.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.Case Study: The Failed Diode

• A DO-35 diode, used in automobile electronics had reported high failure rate. The results of paired comparisons, when examined under a scanning electron microscope, were as follows:– No flaws observed in Good diodes.

• Following Observation are reported on Bad Diodes1. Chipped die, oxide defects, copper migration

2. Alloying irregularities, oxide defects

3. Oxide defects, contamination

4. Oxide defects, chipped die

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Determining Causes of Differences• In some cases detecting the cause of the difference is

simple.• Many times considerable time and experimentation

are required to isolate and conform the cause.• In order to help determine what is different, samples

should be saved. In addition production records should be saved, operators interviewed, and so on. It is important to act fast so that vital information is NOT lost.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Time Differences

Immediately investigate and save samples for follow-up testing

Control charts to better pinpoint time of changes.

Monitor the process to track inputs

Verify cause by turning off and on

ONE TIME CHANGE

SERIES OF CHANGES

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Stream Difference

Immediately investigate and save samples for follow-up testing

Control chart each stream separately and use methods for a series of time differences

Use assembly/ setup studies, and component swapping studies

Verify cause by turning off and on

ONE TIME DIFFERENCE

SERIES OF DIFFERENCES

CHANGING DEFFERENCES

CONSISTENT DEFFERENCES

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Unit Difference

Run assembly/ setup studies, and component swapping studies

Perform unit comparisons

Verify cause by turning off and on

DISASSEMBLY / REASSEMBLE POSSIBLE

DISASSEMBLY / REASSEMBLE NOT POSSIBLE

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

What is FMEA ?• FMEA is a tool for preventing problems.

– As a tool FMEA is one of the most effective low risk techniques for identifying potential problems and preventing them in a cost effective manner.

• FMEA is a procedure for developing and implementing new or revised designs, processes and services.– A structured approach for prioritizing, evaluating, tracking, and

updating design and process development.

• FMEA is the diary of the design, the process or service.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Basic Elements of FMEA

Failures

Causes

Detection

Effects Severity

OccurrenceRPN

Effectiveness

S

O

D

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Types of FMEA• FMEA is useful in the following applications:

– System FMEA

– Design FMEA

– Process FMEA

– Equipment FMEA

– Service FMEA

– Project FMEA

• The main deference in the above three applications is the objective

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Types of FMEA• System FMEA

– How can this system fail to do it’s purpose? or

– What are the potential failures in interactions and interfaces of the system?

• Design FMEA– How can this design fail to do it’s purpose? or

– What are the potential failures in usage of this product design?

• Process FMEA– How can this process fail to do what it is supposed to do? or

– What are the potential failures in following this process design?

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Types of FMEA• Equipment FMEA

– How can this equipment fail to do it’s purpose? or– What are the potential failures in usage of this equipment?

• Project FMEA– How can this each activity fail to do it’s purpose? or– What are the potential failures in implementing project

activities?

• Service FMEA– How can this service fail to do it’s purpose? or– What are the potential failures in execution of this service?

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Process FMEA – Pizza Delivery

Process Steps

Write Order

Failure

Incomplete Address

Effects

No delivery

Causes

Key board problem

Phone disturbance

Talking to some one

Current Controls

None

None

Corrective Action

Regular cleaning

Conform

Hands free kit

RP

N

Det

ectio

n

Occ

urre

nce

Se

veri

ty

102

4

6

10

10

10

200

400

600

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Design FMEA – Computer Printer

System (Function)

Paper Feeding(Guide paper, pick one at a time, move smoothly)

Failure

Wrong orientation

Picked >1 papers

Not loading

High Noise

Effects CausesDesign Controls

Corrective ActionR

PN

Det

ectio

n

Occ

urre

nce

Se

veri

ty

Un-intended function

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

System FMEA – Computer Printer

System Interfaces (Function)

Printing Head and Paper Feeding

Printing Head and Ink cartridge

Failure

Printing with no paper

Alignment of out

Effects CausesDesign Controls

Corrective ActionR

PN

Det

ectio

n

Occ

urre

nce

Se

veri

ty

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Equipment FMEA – Fan

Part(Function)

Bearing(Friction free rotation)

Failure

Jam (No rotation)

High friction

High Noise

Effects CausesCurrent Preventive Maintenance

Corrective ActionR

PN

Det

ectio

n

Occ

urre

nce

Se

veri

ty

Un-intended function

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

• Traditional guidelines suggests that failures with high RPN warrant first consideration for analysis and resource allocation for improvement.

• For example consider the following 3 failures:

Failure Modes S O D RPN A 4 5 10 200

B 4 8 2 64 C 9 2 1 18

Traditional Approach for Prioritizing Failures

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Traditional Approach for Prioritizing Failures

• Minimizing the severity of a failure mode (C) or reducing the occurrence of a failure modes (B) are proactive approaches.

• Improving the detection of the failure mode (A) is a reactive approach.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

New Proactive Approach for Prioritizing Failures

Before the organization allocates resources to improve detection, all opportunities for reducing the occurrence and minimizing the effects of the failure modes should be considered.

HighPriority

Medium

Low

0 1 2 3 4 5 6 7 8 10

0 1

2

3

4 5

6

7

8

10Severity

Occ

urre

nce

FMEA Area Chart

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

Strategies for addressing the failure modes

1. Eliminate the cause.

2. Reduce the severity.

3. Reduce the occurrence.

4. Improve the detection

5. Provide a means of detection for the customer during usage.

Six Sigma Black Belt Training - Day3 Rev. 5

2

004

Inst

itute

of

Sig

ma

Lear

ning

Pvt

. Lt

d.

• Two types of Rating Systems are followed:– 1 through 10 (Very popular)– 1 through 5

Rating Severity Occurrence Detection

1 Low Rare Effective

. . . . . . . .

5 or 10 High Frequent No detection

Rating Systems