OVERVIEW OF SIX SIGMA (6σ) FOR PROCESS IMPROVEMENT Presented by: Larry Bartkus of Biosense Webster and Matthew Thompson of FCI Management Solutions
Oct 19, 2015
OVERVIEW
OF SIX SIGMA (6) FORPROCESS IMPROVEMENT
Presented by:
Larry Bartkus of
Biosense Webster
and
Matthew Thompson of
FCI Management Solutions
Goals
g Introduction to Six Sigma Description Philosophies Benefits Origin Mythsg Identify the five steps of DMAIC, the core of Six Sigma, and
the logical flow from step to step.g List tools and concepts useful in each step.g Brief introduction to DMADV
Six SigmagDescription
Six Sigma is a quality program that, when all is said and done , improves your customers experience, lowers your costs, and builds better leaders. Jack Welch CEO GE
Six Sigma is a proven set of tools and tactics used for process improvement, reduction of defects, and improved quality
Six Sigma uses data and statistical analysis to zero in on root causes
Six Sigma can be applied to any process
23456
Defects per
Million opportunities
Defects per Million
opportunities
308,53766,8076,210
2333.4 .
3 to 6 - 20,000 Times Improvement ... A True QFL
Six Sigma Measurement
Six Sigma: What Does it Mean
30%
6%
.6%3.8 10,724 1%
2.5 158,686 16%
99.99966% Good (6 Sigma)
20,000 lost articles of mail per hour
5,000 incorrect surgical operations per week
Two short or long landings at most major airports each day
200,000 wrong drug prescriptions each year
Seven articles lost per hour
1.7 incorrect operations per week
One short or long landing every five years
68 wrong prescriptions per year
99% Good (3.8 Sigma)
Six Sigma PerformanceSix Sigma Performance
Most Businesses Operate at about 3.5 Sigma
Six SigmagPhilosophies
When defects occur look to the process for the cause. Excellent processes will allow average people to consistently
generate superior results.
gBenefits More loyal and satisfied customers (internal and external) Financial savings through improved efficiency and effectiveness Resolution of chronic problems
gOrigin The late Bill Smith, a reliability engineer at Motorola, is widely
credited with originating Six Sigma and selling it to Motorola'slegendary CEO, Robert Galvin.
gMyths
THE DMAIC METHOD
Overview of the DMAIC Method
IMPROVE
CONTROL
MEASURE
ANALYZE
5
2
3
4
DEFINE
1
Phase 1: DefineGoal
g Define the projects purpose and scope and get background on the process and customer
Outputg A clear statement of the
intended improvement and how it is to be measuredg A high-level map of the
processg A list of what is important to
the customer
IMPROVE
CONTROL
MEASURE
ANALYZE
5 1
2
3
4
Bus
ines
sC
ase
Proj
ect
Char
ter
SIPOC
VOC
QFDDEFINE
Project Charter
Define
DefineDefine MeasureMeasure AnalyzeAnalyze Improve/Improve/InnovateInnovate ControlControl
Product FlowSIPOC
Process OutputInput CustomerSupplier
Planning Level Build Plan Doctors
Quality
Engineering
JCIT
CapacityRequirements
Line Design
ComplianceIssues
Consulting
Improved CycleTimes /WIPReductionsReduced QNCs
Improved Efficiency/Improved Compliance
Shareholders
Quality
Regulatory
ReceiveFinished Goods
Trigger
Cut Work Orderto Line
Sequence WO'sThrough
Single Prod.Line
MarketingForecast
SterilizeRelease
Product toFinished Goods
EXISTING PROCESS FLOW
DefineDefine MeasureMeasure AnalyzeAnalyze Improve/Improve/InnovateInnovate ControlControl
VOICE OF THE CUSTOMER
Identify hidden non-compliance issuesWhy NotCampaign
Reduce QNCs
Create simplified process flow.
Mitigate opportunity for near miss eventsSimplified Product Flow
Eliminate product mix-ups
Compliance to procedures
Consistency between documents
Improved Product Build Documentation
OperationsRapid
Improvement
Documentation mirrors process
Ease of use
Floor associate buy-in to improving the process
Input from floor associates
Reduce process related QNCs by 50%
Reduce near miss eventsfrom 14 in 2004 to 4 (1 per quarter) in 2005
Reduce near miss eventsfrom 14 in 2004 to 4 (1 per quarter) in 2005
Phase 2: MeasureGoal
g Focus the improvement effort by gathering information on the current situation
Outputg Data that pinpoints
problem location or occurrenceg Baseline data on current
process sigmag A more focused problem
statement
IMPROVE
CONTROL
MEASURE
ANALYZE
5
2
3
4
DEFINE
1
BaselineData
Sampling
Gage R&R
PatternsCapability
D M A I C
The system to collect information is already established.
100% of Scrap forms from Work Orders for all products are captured into a Yield Database.
Scrap form (FORM633) has information as Lot Number, Date, Shift, Product Number, Defect Code, Defect Description, Potential Cause, Work Station ID, Catheter Number, Scrap Quantity, and is signed by operator and supervisor / engineer.
Yield Database has information as Product Number, Date, Lot Number, Description, Lot Quantity, Reworks, Scrap in Line, and Scrap in QA.
Documentation Time StudyAnalyze
Minutes per Work OrderMaximum = 26.5 min per W/O
8
32.5 2.5 2.5 2 2 2 2
0123456789
Form
046
Form
335
Form
618
Form
621
Form
617
Form
615
Form
633
Form
616
Form
051
Form
L
e
n
t
h
o
f
t
i
m
e
i
n
M
i
n
u
t
e
s
Burst Tester Gage R&R
AnalyzeMeasure Improve ControlDefine
P
e
r
c
e
n
t
P a r t- to - P a r tRe p r o dRe p e a tG a g e R& R
1 0 0
5 0
0
% Co n t r ib u t io n
% S t u d y V a r
S
a
m
p
l
e
S
t
D
e
v
6
4
2
_S= 3 .4 4 7
U CL=5 .4 1 8
LCL=1 .4 7 6
O p e r a to r 1 O p e r a to r 2
S
a
m
p
l
e
M
e
a
n
1 0 0
8 0
6 0
__X= 8 6 .1 0
O p e r a to r 1 O p e r a to r 2
U CL= 8 8 .8 1
LCL= 8 3 .3 8
O p e r a to r sS e ttin g s
O p e r a to r 2O p e r a to r 1N o m ina lM in im umN o m ina lM in im um
1 0 0
7 5
5 0
OperatorsO p e r a to r 2O p e r a to r 1
1 0 0
7 5
5 0
G a ge na m e : G a ge R &R fo r B urst V a lue s (U n iv e rsa l T ra y )D a te o f study :
R e po rte d by : J . C ha nT o le ra nce :M isc:
Co m p o ne nts o f Va ria tio n
S Cha rt by O p e ra to rs
Xba r Cha rt b y O pe ra to rs
Burst By S e ttings ( O p e ra to rs )
Burst b y O pe ra to rs
Ga ge R& R (Ne s t e d) fo r Burs t V a lue s
The total Gage Repeatability and Reproducibility is 12.38%, which is well below the acceptance criteria of less than or equal to 30% of the Total Variation. Measurement system is acceptable.
The Gage R & R study is able to show that the burst tester is capable of detecting different parts being measured as shown from the data reported by two different operators measuring the samples.
Process Capability-Current
I
n
d
i
v
i
d
u
a
l
V
a
l
u
e
4321
1.8
1.6
1.4
_X=1.5955
UCL=1.9437
LCL=1.2472
M
o
v
i
n
g
R
a
n
g
e
4321
0.4
0.2
0.0
__MR=0.1309
UCL=0.4278
LCL=0
Observation
V
a
l
u
e
s
4321
1.68
1.60
1.52
1.81.71.61.51.41.31.2
LSL
SpecificationsLSL 1.13
2.11.81.51.2
Within
O v erall
Specs
WithinStDev 0.116076C p *C pk 1.34
O v erallStDev 0.102049Pp *Ppk 1.52C pm *
Process Capability Sixpack of C beforeI Chart
Moving Range Chart
Last 4 Observations
Capability Histogram
Normal Prob PlotA D: 0.350, P: 0.255
Capability Plot
Process Capability Ring 20 Current Marking Method
0 1 2 3 4 5 6 7 8 9 10
169.70
169.95
170.20
170.45Xbar and R Chart
Subgr
M
e
a
n
Mean=170.1
UCL=170.4
LCL=169.7
0.0
0.3
0.6
0.9
R
a
n
g
e
R=0.3515
UCL=0.9050
LCL=0
0 1 2 3 4 5 6 7 8 9 10
Last 10 Subgroups
169.8
170.0
170.2
170.4
Subgroup Number
V
a
l
u
e
s
169.5 170.5
Capability PlotProcess Tolerance
I I I
I I I
I ISpecifications
Within
Overall
169.5 170.0 170.5
Normal Prob Plot
169.6 170.0 170.4
Capability Histogram
WithinStDev:Cp:Cpk:
0.2076470.800.70
OverallStDev:Pp:Ppk:
0.1911310.870.76
Process Capability Sixpack for Prod. Ring20
DMDMAAII22CC
Phase 3: Analyze
Goalg Identify deep root
causes and confirm them with data
OutputgA theory that has
been tested and confirmed
IMPROVE
CONTROL
MEASURE
ANALYZE
5
2
3
4
DEFINE
1
Process
Analysis
Multi-Vari
Organize
Causes
Hypothesis
TestingRegression
DoE
D M A I C
D-1237-02-SD-1237-01-S
1.34
1.32
1.30
1.28
1.26
1.24
1.22
1.20
1.18
1.16
PART NUMBER
A
v
e
r
a
g
e
p
e
r
W
O
1
2
Multi-Vari Chart for Short in Ring
Shif t
D-1237-01-S D-1237-02-S
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
PART NUMBERA
v
e
r
a
g
e
p
e
r
W
O
1
2
Multi-Vari Chart for No ReadingShif t
D-1237-01-S D-1237-02-S
0.98
1.08
1.18
1.28
1.38
1.48
1.58
1.68
1.78
PART NUMBER
A
v
e
r
a
g
e
p
e
r
W
O
1
2
Multi-Vari Chart for Lasso out of RoundnessShif t
D-1237-02-SD-1237-01-S
1.35
1.30
1.25
1.20
1.15
1.10
1.05
1.00
PART NUMBER
A
v
e
r
a
g
e
p
e
r
W
O
1
2
Multi-Vari Chart for Damaged RingShif t
D-1237-01-S D-1237-02-S
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
PART NUMBER
A
v
e
r
a
g
e
p
e
r
W
O
1
2
Multi-Vari Chart for Damaged Spine CoverShif t
D-1237-01-S D-1237-02-S
1.0
1.1
1.2
1.3
PART NUMBER
A
v
e
r
a
g
e
p
e
r
W
O
1
2
Multi-Vari Chart for Tip going BackwardsShif t
Benchmarking - InternalS
t
r
a
t
e
g
i
c
A
d
v
a
n
t
a
g
e
Advanced ApplicationsHigh Interactivitye.g. online forms
Knowledge Mgt hubPrimary Communications
VehicleContent Rich
Tier 4Transforming
Features
Tier 1Basic Presence
Basic Company
Information
Tier 3Integrating
CommunitiesAdvanced SearchValue-add tools
Moderate Interactivity
PersonalisationExtensive information
Basic Search
Tier 2Searching
Analyze
Regression Analysis
Analyze
Forecasted Demand08/15/05- 07/15/05
M o nth
G
r
a
n
d
T
o
t
a
l
121086420
40000
39000
38000
37000
36000
35000
34000
33000
32000
31000
S 2661.11R -S q 7.0%R-S q(adj) 0 .0%
F itted L ine Plo tG rand T o tal = 3 4329 + 193 .1 Mon th
7% Increase in demand/year
Design of Experiment
AnalyzeMeasure Improve ControlDefine
23 Factorial Design
Factors Low HighTemperature 245 F 260 F
Pressure 55 PSI 90 PSITime 2 SEC 3 SEC
Design of Experiment
AnalyzeMeasure Improve ControlDefine
Hi
Lo0.00000D
New
Cur
d = 0.00000
MaximumBurst Va
y = 94.6429
2.0
3.0
55.0
90.0
245.0
260.0Pressure TimeTemp
[245.0] [75.0] [3.0]
Phase 4: ImproveGoal
gDevelop, pilot, and implement solutions that address root causes.
Outputg Identification of
planned, tested actions that should eliminate or reduce the impact of the identified root causes
CONTROL
MEASURE
ANALYZE
5
2
3
DEFINE
1
Solutions
FMEA
Pilot
Implem
en-
tation
IMPROVE4
D M A I C
Improvement team created with members of Quality, Production, Engineering and R&D areas to propose and evaluate ideas.
Brainstorming tool was used to gather ideas on how to solve the problems identified.
Ideas were evaluated per following criteria: feasible, high impact, easy, low cost, and quick.
Following tables summarizes solutions agreed by consensus of the improvement team.
D M A I C
Add a new hole to apply PU to bond components once they are in place.
PU is applied before the components are in place and is partially removed during installation of components.
Lasso to tip method. Defective bonding of Compression Coil and Lasso Stem
Design a new fixture to hold Nitinol during burnishing.
Current burnishing method is not ergonomic causing associates make defects after some time.
Prep. lasso stem method. Poor burnishing of Nitinol wire.
Train associates on visual acceptance criteria.
Rejection of good product exposes it to damage during rework operations.
Some acceptance criteria are not clear for associates.
PROPOSED SOLUTION
BRIEF DESCRIPTIONPROBLEM
Solutions
FMEA
Pilot
Imple
men-
tation
Solutions
FMEA
Pilot
Imple
men-
tation
RISK 12. Support Plan not in Place
Description IntegWare does not want to put Biosense Webster in a position of being overly reliant for day-to-day level 1 support of the system after go-live. IntegWare typically suggests that we assume the role of a second level support for a limited number of hours per month. IntegWare suggests it is typical for Day-to-Day (level 1) support will be a highly resource consuming task for three-six months after go-live. To use IntegWare for this might be an ineffective usage of the customer's resources. The project ownership should be transferred to the customer if at all possible.It would be a better usage of the customers resources to ask IntegWare to implement higher value-added initiatives in PDM (such as developing PART/BOM configuration management features, JDE integration, etc.)
MITIGATION To mitigate this risk, we have discussed a support concept at BWI. This will be a combo function between Doc services (they will handle basic application admin functions) and IM who will handle IT functions.Still need to setup support contract for IntegWare to support system.Celeste has added to the schedule a support contract for level 2 support from IntegWare.
RISK LEVEL HIGH
UPDATED 5/2/03
ASSESSING RISKAn Example:
AnalyzeMeasure Improve/ ControlDefine
Estimated Cost Benefit
100101
CMM vs. Estimated Cost Avoidance
$0
$1
$2
$3
$4
$5
1 2 3 4 5
M
i
l
l
i
o
n
s
Capability Maturity Model (CMM) Level
C
u
m
u
l
a
t
i
v
e
C
o
m
p
l
i
a
n
c
e
R
i
s
k
(
$
)
Cumulative Risk ($) Compliance Cost
Risk reduction= $1.7 Million
1.6 2 .5
Pilot/Pilot Plan
AnalyzeMeasureDefine Control
Used Japan Custom line as a pilot- Medium size production line- Minimize the production Impact
Kanbans Cards
Signals, Lights
Performance Measures
Improve/
Phase 5: ControlGoal
g Use data to evaluate both the solutions and the plansg Validate that all changes adhere to
all operating company change control, GMP, and compliance requirements gMaintain the gains by
standardizing processesg Outline next steps for on-going
improvementOutput
g Before-and-After analysisgMonitoring systemg Completed documentation of results,
learnings, and recommendations
IMPROVE
CONTROL
MEASURE
ANALYZE
5
2
3
4
DEFINE
1C
ontrol
Standardize
Document
Monitor
Evaluate
Closure
D M A I C
All new tooling, processes, clarifications, and visual aids documented in PIs.
Creation of MOPXXX with all the quality criteria for Variable Lasso Deflection and Contraction performance.
MOP003 updated with reference to new created MOPXXX.
PU quality criteria documented in WSS001 and available in QA workbenches.
D M A I C
Yields Chart is published every week.
Work Orders Built
Yield Goal
Scrap Produced
Yield Actual
I-MR Charts of Burst Tester
AnalyzeMeasure Improve ControlDefine
Burst Values
I
n
d
i
v
i
d
u
a
l
V
a
l
u
e
302928272625242322212019181716151413121110987654321
115
110
105
100
_X=105.88
UCL=113.76
LCL=97.99
Observation
M
o
v
i
n
g
R
a
n
g
e
302928272625242322212019181716151413121110987654321
10
5
0
__MR=2.97
UCL=9.69
LCL=0
Universal Tray Packaging (CA Sealer)In-H20
Burst ValuesI
n
d
i
v
i
d
u
a
l
V
a
l
u
e
302928272625242322212019181716151413121110987654321
110
105
100
95
_X=100.72
UCL=109.68
LCL=91.76
Observation
M
o
v
i
n
g
R
a
n
g
e
302928272625242322212019181716151413121110987654321
12
8
4
0
__MR=3.37
UCL=11.01
LCL=0
Universal Tray Packaging (Juarez Sealer)In-H20
Lessons Learned
Culture shock Resources Scope creep / Identifying the critical few Management buy-in of change of policies Time management & deliverables challenges J&J Goals too high for small operating company (CMM
goal of Level 3 per J&J Corp Strategy) Difficult to measure compliance
Project Closure
gImprovement must be continuous, but individual initiatives and project teams come to an end.gLearn when its time to say goodbye.gEffective project closure weaves together the themes
of: Project purpose. Improvement methods. Team skills and structures.gDevelop managerial systems to capture learnings and
enable the organization to address system issues.gDocumentation and recognition are two critical aspects
of project team closure.gCelebrate!
DMAIC Pathway
New product Development - DEx Roadmap - DMADV
DEFINE Develop the Business Case, Scope & Charter the Project MEASURE Gather & Quantify Design Inputs (Customer, Technical, Business,
Regulatory) ANALYZE Develop and Investigate Conceptual Designs DESIGN Develop Detailed Product/Service/Process Designs VERIFY/VALIDATE Confirm design outputs meet design input requirements and
ensure specifications conform with intended uses and users; Scale-up manufacture and release the product or scale-up and implement the new process; Finally, transfer to process owners
Verify/ Verify/ ValidateValidateDesignDesignAnalyzeAnalyzeMeasureMeasureDefineDefine
Opportunity
Define
Measure
Analyze
Design
Verify/Validate
Transfer
DEx MethodologyIDTask Name1 ID Customers2 ID Needs3 ID CCRs4 Review5 Develop Concepts6 High-Level Design7 Capability8 Design Review9 Develop Details
10 Simulation11 Cost Analysis12 Design Review13 Procurement14 Implementation
SepOctNovDecJanFebMarAprMayJunJulAugSepQtr 4, 1999Qtr 1, 2000Qtr 2, 2000Qtr 3, 2000
Business CaseGoal StatementProject Plan
Opportunity StatementProject ScopeTeam Selection
Team Charter
Generation 1Generation 2Generation 3
Vision
Product/ServiceGeneration
Technologies/Platforms
Theme 1
Need 1 Need 2
Theme 2
Need 3 Need 4
Need 5
Theme 3
Need 7
Need 8
1
CustomerRequirem
ents(V)
3Characteristic/Measur
es (how)(I)
Target goals
7Correlatio
n (I)
2
CustomerRating(B) (V)
4
Relationships(What vs.
How)(I)
6
5
I
m
p
o
r
t
a
n
c
e
(
V
)
How importantTargets/Specs(B)(I)Technical Evaluation (B)
Measures(Hows)
C
u
s
t
o
m
e
r
n
e
e
d
s
(
W
h
a
t
s
) House Of
Quality #1
Criticalprocess characteristics
(Hows)
M
e
a
s
u
r
e
s
(
W
h
a
t
s
) House Of
Quality #2
Creative Techniques
Experimentation
Proto
typing
Customer Trials
Design PrinciplesBen
chmark
ing
TRIZ
White Red Black Blue Green Yellow
Start With What You
Know
Get Other Perspec-
tives
Non-Linear Thinking Build on Ideas
Combine Ideas Compare
FAULT TREESYSTEM
RELIABILITY PREDICTION
SystemFailure
PumpP1
Fails
PumpP2
Fails
PumpV1
Fails
PumpV2
FailsR = RBil X (1 - (1-Rp)2) x
(1 - (1-Rv)2)
PumpsBilge Valves
P1
P2
V1M
V2M
BilgeFailure
PumpFailure
ValveFailure
Time (t)
H
a
z
a
r
d
R
a
t
e
-
h
(
t
)
Random FailuresEarly
FailuresWearou
t Failures Proce s s 2
Proces s 1
Proces s 3
Worke r 1Ins pe c tor 1
Worke r 2Ins pe ctor 2
Worke r 3
Inspe cto r 3
Re corde r 3
Re corde r 2
Re cord e r 1
Card
Ins pe c t 1
Inspe ct 2
Inspe ct 3
Card Drop Shop - Proces s Mode l
Re corde rs
Worke rs
Ins pe ctors
DevelopProcess
Control Plan
ErrorProofing andContingency
Planning
DetailedDesignReview
Prepare PilotTest Plans
Act Plan
Check Do
Start
Type of data
?
Counting items with an
attribute or countingoccurrences?
Equalsamplesizes
?
Equal opportunity
?
Continuous
Yes
No
Yes
Rational Subgroups
Discrete
Yes
No
No
Do limits look right?
Try individuals charts
Need to detect small shifts
quickly?
Individual measurements or subgroups
?
Do limits look right?
YesNo
Either/Or
No
Yes
Individualmeasurements
OccurrencesItems withattribute
p chart
Try transformation to make data normal
no chart u chart c chart X, R or s chart
Individualscharts
EWMAchart
Unstable (Not in control)
Stable (In control)
Within spec limits
Upper Spec
Lower Spec
Upper Spec
Lower Spec
UCL
LCL
UCL
LCL
Outside spec limits
Upper Spec
Lower Spec
Upper Spec
Lower Spec
UCL
LCL
UCL
LCL
Type of Design
P
r
o
d
u
c
t
s
S
e
r
v
i
c
e
s
D
e
s
i
g
n
-
P
r
o
d
u
c
t
i
o
n
-
D
e
l
i
v
e
r
y
P
r
o
c
e
s
s
e
s
S
u
p
p
o
r
t
i
n
g
P
r
o
c
e
s
s
e
s
Type of Design
Design Elements
P
r
o
d
u
c
t
s
S
e
r
v
i
c
e
s
D
e
s
i
g
n
-
P
r
o
d
u
c
t
i
o
n
-
D
e
l
i
v
e
r
y
P
r
o
c
e
s
s
e
s
S
u
p
p
o
r
t
i
n
g
P
r
o
c
e
s
s
e
s
1. H\W Products
2. S\W Products
3. Services
4. Analyses
5. Information Systems
6. Processes and Methods
7. Human Resources
8. Site \ Facilities
9. Equipment \ Tools
10. Materials \ Supplies
11. Sales \ Marketing
12. Other Non-Technical
Question & Answer