cqpr 2016 presentation
TRANSCRIPT
Business Excellence Consulting, Inc.Ph: 787.705.7272 – www.calidadpr.com
Application of Six Sigma Tools in the Pharmaceutical Industry
PRChem Annual ConventionAugust 5th, 2016
Ritz Carlton Hotel and Casino, Isla Verde PR
Agenda
I. Six Sigma IntroductionII. Management and Organizational Infrastructure for Six SigmaIII. DMAIC Breakthrough MethodologyIV. Application of Tools: Define PhaseV. Application of Tools: Measure PhaseVI. Application of Tools: Analyze PhaseVII. Application of Tools: Improve PhaseVIII. Application of Tools: Control PhaseIX. Application of Tools for an Annual Product Review
2
This presentation is designed for those persons who are involved in implementing Six Sigma and want to understand its core concepts and benefits.
Provides an overview and orientation of Six Sigma methodologies and organizational requirements.
Audience
3
Course Objectives
ü Describe the Six Sigma Breakthrough strategyü Understand each participant’s role in the Six Sigma
Initiativeü Relate the roles and responsibilities of the
Champion, Black Belt, Green Belt and Team Members
ü Understand the DMAIC methodology and how it applies to improvement projects
4
Section I: Six Sigma Introduction
What is Six Sigma?
3-Sigma vs. 6-Sigma
Six Sigma Objective
What Makes Six Sigma Work?
DMAIC Roadmap
ü Management mandated and directed improvement program focused on breakthroughs in financial performance and customer satisfaction
ü Uses Champions, Black Belts and Green Belts to facilitate change
ü Focused on core business and Customer needs
ü A systematic method for process and product improvement
ü A Greek symbol for measuring performance variation
ü A metric for evaluating performance quality
ü A standard of excellence (3.4 defects per million opportunities)
What is Six Sigma?
6
3-Sigma vs. 6-Sigma
CenterProcess
ReduceSpread
Off-Target Too Much Variation
Centered On-Target
The objective is to understand customer requirements and reduce
process variation and defects
Six Sigma Objective
8
ü Select and work on the most important problems and projects to the business and Customer
ü Assure those projects impact customer satisfaction and financial performance
ü Allocate the time to get the work doneü Have your best people work on themü Provide those individuals with all the training, tools, and resources
they need to make the performance breakthroughü Provide Management direction, support and routine review of
performanceü Require a well thought out, objective and data driven solution ü Verify the dollar savings of your effortsü Sustain the benefits of the solution over time
What Makes Six Sigma Work?
9
Section II: Management Infrastructure for Six Sigma
Interrelationship of Six Sigma Infrastucture
Site Champion and Executive Staff
Hands on Champions
Master Black Belt
Black Belts / Green Belts
All Employees
Interrelationship of Six Sigma Infrastructure
Site Champion and Executive Staff
Functions
Set the vision
Create the mandate for improvement
Initiate and fund the activity
Establish and maintain the reporting structure
Responsibilities (10%+ of time)
Identify Champion/Sponsor in each functional area
Review of projects (members of Governance Boards)
Measure Project results
Training
2 day Champion Training12
Hands-On Champion (Sponsor)
Functions
Communicate the vision
Create the mandate for improvement
Provide direction and remove barriers
Achieve financial results and communicate success
Responsibilities (10%+ of time)
Identify Black Belts and Green Belts
Identify and approve all Six Sigma projects
Part of Governance Board
Measure Black Belt performance
Training
2 day Champion Training 13
Master Black Belt
Functions
Provide technical expertise on Six Sigma and Lean methodology to Black Belts and Green Belts
Work in support of the Black Belt and Champion
Assist in education and training activities
Responsibilities (100% of time)
Work daily with team members, Black Belts and Champions
Participate in the review of projects
Monitor all Six Sigma projects
Training
Black Belt Training, 2 years minimum as a Black Belt and a Masters Degree in a related field
14
Black Belts and Green Belts
FunctionsBlack Belts 100% dedicated to process improvement, Green Belts 20%Work on improvement projects with other Green Belts and Team MembersAchieve financial results for each project
Responsibilities (100 – 20%+ of time)Use the DMAIC methodology to create breakthroughs in performanceReport progress to the ChampionHold team meetings and provide project managementWork with the Master Black Belt
Training20 day Black Belt Training 10 day Green Belt Training
15
All Employees
Functions
Work to achieve excellence in daily work
Participate in Lean Six Sigma projects
Support team activities
Identify opportunities for improvement
Responsibilities (5-10%+ of time)
Participate in Lean Six Sigma activities as requested
Complete action items as assigned by the team
Attend team meetings
Training
1 day Lean Six Sigma Awareness training 16
Section III: DMAIC Breakthrough Methodology
Methodology
Process
Roadmap
DMAIC Methodology
18
DefineIdentify Project, Champion and Project OwnerDetermine Customer Requirements and CTQsDefine Problem, Objective, Goals and BenefitsDefine Stakeholder/Resource AnalysisMap the ProcessDevelop Project Plan
MeasureDetermine Critical Xs and YsDetermine Operational DefinitionsEstablish Performance StandardsDevelop Data Collection and Sampling PlanValidate the MeasurementsMeasurement Systems AnalysisDetermine Process Capability and Baseline
AnalyzeBenchmark the Process or ProductEstablish Causal Relationships Using DataAnalysis of the Process MapDetermine Root Cause(s) Using Data
ImproveDesign of ExperimentsDevelop Solution AlternativesAssess Risks and Benefits of Solution AlternativesValidate Solution using a PilotImplement SolutionDetermine Solution Effectiveness using Data
ControlStatistical Process ControlDetermine Needed Controls (measurement, design, etc.)Implement and Validate ControlsDevelop Transfer PlanRealize Benefits of Implementing SolutionClose Project and Communicate Results
DMAIC Roadmap
Section IV: Application of Tools -Define Phase
Project Charter
- Problem Statement
- Scope
- Goal Statement
- Resources
- Benefits
Project Charter Example
ü Problem
ü Objective & Scope
ü Goal/target
ü Resource requirements
ü Financial benefits
ü Strategic benefits
ü Team resources
ü Project approval
In this step you will establish direction for the project by documenting
Título del Projecto: Responsable: Espónsor del Projecto:
Caso de Negocios: (Resumen corto detallando los beneficios del proyecto, incluyendo la parte económica)
Escriba el caso de Negocios:
Descripción de la Oportunidad: (Detalle la oportunidad que el proyecto quiere mejorar)
Describa la Oportunidad:
Alcance del Projecto: (Defina los limites del proyecto)
Escriba los límites del Proyecto
Metas del Poyecto: Especifica Medible Alcanzable Relevante
Escriba las Metas del Proyecto
Miembros del Equipo (Listado de los miembros y sus funciones)
Listado de miembros del equipo aqui
Milestones/Deliverables: (Fecha que en que esperamos completar cada etapa)
D - M - A - M - C -
Project Charter
ü A description of the issue or problem
ü What, when, where and how the problem occurs
ü What is critical to quality (CTQ) from the customer perspective? What is nonconforming to specifications?
ü Define the boundaries of the problem, how big is the problem?
ü What are the consequences of the problem to the Company and or Customer?
ü Do not state a solution when describing a problem
Description of the “Pain”
Project Charter:Problem Statement
22
ü What process will the team focus on?
ü What are the boundaries of the process we are to improve? Where does the process begin? Where does it end?
ü Which customers will be affected / included
ü What is outside of scope for the team?
ü What constraints or timelines must the team work under?
Customer or Suppliers Site or Line Process
Process 1
Process 2
Process 3
Project Charter:Scope
Goal Statement Example:
Reduce Packaging Line #2 scrap by 33% at the end of 3Q 2016.
ü Describe, in measurable terms, what success will look like when you’ve solved the problem
ü Include a statement of the performance level that will satisfy your CTQ and the time frame in which you plan to implement the improvement. State targets and tolerances where appropriate.
Project Charter:Goal Statement
24
Determine the Stakeholders
Who will need to be involved and at what level for this project to be successful?
Determine Team Members
Who will need to be on the team and at what % of their time?
Determine any Capital or HR Requirements
What materials, testing, software programming, outside services or equipment will be needed? Are those costs acceptable to management?
Other Support to be Successful
Project Charter:Resources
25
ü Benefits may include any of the following:ü Hard dollar savings (cost reduction)ü Soft dollar savings (cost avoidance)ü Improved Customer satisfactionü Improvement in the performance metrics (cycle time, yield,
defects, etc.) ü Improved job satisfaction (reduction in frustration)ü When considering benefits… what must change for the
savings to occur?
Project Charter:Benefits
26
27
Section V: Application of Tools -Measure Phase
Measurement System Analysis
Process Capability
Graphical Tools:
- Histogram / Dotplot
- Boxplot
- Pareto- Scatterplot
Measurement System Analysis
29
30
Elements of Variation
◆ Each operator measures the same part several times.◆ Data must be balanced; that is, operators must measure
each part the same number of times.◆ Units must represent the whole range of expected
variation. It is recommended to select parts within the specification range, that is, parts from the lower specification limit, parts from the upper specification limit, and parts within the specification limits.
◆ Operators must measure the parts randomly. They must not know which part number they are measuring at any given time.
31
Gage R&R Basics
32
Gage R&R Metrics
◆ Repeatability—The variation caused by the instrument. It is the variation observed when an operator measures the same part repeatedly using the same measurement instrument.
◆ Reproducibility—The variation caused by the measurement system. It is the variation observed when several operators measure the same part using the same instrument.
◆ Percent precision-to-tolerance (% P/T)—The percentage of the specification tolerance occupied by the measurement system variation.
◆ Percent gage R&R (% R&R)—The percentage of the total variation occupied by the measurement system.
33
Performing a Gage R&R
34
Performing a Gage R&R
Process Capability
35
Analyzing Process Capability
Capable Process
Incapable Process 36
Process Capability Indices
37
Interpreting the Process Capability Indices
38
Process Capability Example
39
Process Capability Example cont.
40
Individuals
Moving Range
Process Capability Example cont.
41
Process Capability Example cont.
42
Process Capability Example cont.
43
Graphical Tools
44
Describing the Data:Histograms
45
Describing the Data:Dotplots
46
Comparing Groups:Boxplots
47
Prioritizing Our Actions:The Pareto Diagram
48
Analyzing Relationships:The Scatterplot
49
Section VI: Application of Tools -Analyze Phase
Hypothesis Tests:
- 1 sample t, 2 sample t, ANOVA
- 1 sample sign, Krustal-Wallis
- F test, Levene
◆ A hypothesis test is a method of making decisions using data from our processes.
◆ In statistics, a result is called statistically significant if it is unlikely to have occurred by common causes of variation alone, according to a predetermined threshold probability called the significance level.
51
Overview
◆ If the p-value is less than the required significance level, then we say the null hypothesis is rejected at the given level of significance. – Under this scenario, we might accept the alternate
hypothesis. ◆ If the p-value is not less than the required significance
level, then the test has no result. – The evidence is insufficient to support a conclusion.
52
Overview
◆ To summarize, here is a slogan you can use to remember when to reject or fail to reject the null hypothesis (H0):
“If p-value is low, the null hypothesis must go.”
◆ In other words, when the p-value for the null hypothesis is lower than a predetermined value, then the null hypothesis must be rejected.
53
Overview
◆ When dealing with hypothesis testing, there are four decisions that can be made.
◆ Two of those decisions are correct decisions, while two of them are wrong decisions.
◆ In statistics, those wrong decisions are called type I error and type II error.
54
Overview
55
Overview
◆ One-Sample t-Test– Comparing one average to a target value
◆ Two-Sample t-Test– Comparing the averages of two populations
◆ One-Way ANOVA Test– Comparing the averages of two or more populations, using
just one factor
56
Comparing Means
Three months ago, a company implemented a new CAPA investigation process after a massive training. The company wants to know if the training is paying off, that is, if the average closure time for the investigations has decreased significantly. Before training, average closure time was 20 days.
57
One-Sample t-Test Example
58
One-Sample t-Test Example
A company wants to compare how long it takes (in hours) for each shift in the quality control laboratory to perform an analytical test on a specific product. There are two shifts: shift 1 and shift 2. A given test is selected, and the times to complete that test are collected and analyzed with statistical software.
59
Two-Sample t-Test Example
Two-Sample t-Test Example
NOTE: A test for variances must be performed prior to the two-sample t-test
61
Two-Sample t-Test Example
A company wants to compare the performance of three shifts in terms of the manufacturing cycle time to determine if there are significant differences in their averages. After verifying the assumptions related to normality of the data and equal variances, the box plots shown in the next page were developed and a One-Way ANOVA performed.
62
One-Way ANOVA Example
63
One-Way ANOVA Example
64
One-Way ANOVA Example
Although it was proven that the average manufacturing cycle time for each shift was not statistically different, the performance among the shifts does not appear to be the same. For that reason, the company developed a quality index to measure their performance. The index is based on a scale of 1 to 5, where 1 means poor performance and 5 means excellent performance.
65
One-Way ANOVA Example
66
One-Way ANOVA Example
67
One-Way ANOVA Example
◆ One-Sample Sign Test– Comparing one median to a target value– Non-parametric equivalent to the one-sample t-test
◆ Two-Sample Mann-Whitney Test– Comparing the medians of two populations– Non-parametric equivalent to the two-sample t-test
◆ Krustal-Wallis Test– Comparing the medians of two or more populations, using
just one factor– Non-parametric equivalent to the ANOVA test
68
Comparing Medians
The median cycle time for investigation has been 25 days (50% of investigations needed more than 25 days to be completed). After new root cause analysis training, three months of data were used to evaluate the improvement (if any).
69
One-Sample Sign Test Example
70
One-Sample Sign Test Example
A pharmaceutical company wanted to determine if there are significant differences in the hardness of the tablet they manufacture using two different pieces of equipment. In order to analyze this, a period of one year of production was collected from each machine. After performing a normality test, it was noticed that data do not follow the normal distribution. The histograms in the next page show a p-value for the Anderson-Darling normality test of 0.0000 for each machine, indicating that the data do not follow the normal distribution.
71
Two-Sample Mann-Whitney Test Example
72
Two-Sample Mann-Whitney Test Example
73
Two-Sample Mann-Whitney Test Example
A pharmaceutical company wanted to analyze three different suppliers of a fluid used in their manufacturing process in order to select one of them. A critical characteristic of the fluid is viscosity, measured in centipoise (cP). The specifications for the viscosity of the fluid range from 2.4 to 3.6 cP. Data from the three suppliers were analyzed and a histogram and descriptive statistics of them generated.
74
Krustal-Wallis Test Example
75
Krustal-Wallis Test Example
76
Krustal-Wallis Test Example
◆ F-test– Comparing variances of two populations
◆ Bartlett Test– Comparing variances of two or more populations
◆ Levene Test– Comparing variances of two or more populations when
they do not fit a Normal distribution
77
Comparing Variances
A company wants to analyze two different adhesives for the transdermal patches they manufacture. Adhesives on a patch generally help maintain contact between the transdermal system and skin surface. The adhesiveness of the patches is critical in the drug delivery mechanism, its safety, product quality, and efficacy. As such, a good adhesive should easily adhere to the skin with an applied finger pressure and be tacky enough to maintain a strong holding force. The adhesive should also be easily removed from the skin without leaving a residue. The specification for the adhesiveness of the transdermal patch ranges from 300 to 3500 g/system.
78
F-test
79
F-test
80
F-test
Let us suppose that in the previous example the manufacturer would like to test a third adhesive. Since the F-test can only compare two variances at a time, we need to perform a Bartlett test.
81
Bartlett Test Example
82
Bartlett Test Example
83
Bartlett Test Example
84
Bartlett Test Example
A pharmaceutical company wants to evaluate the time it takes their analysts to perform a certain laboratory test. They have noticed too much variation in the time it takes the various shifts to complete the test and would like to investigate the reason for the variation. In order to start the analysis, the time it took each analyst to perform the laboratory test (in hours) is collected. Then, the data are segregated by shift.
85
Levene Test Example
86
Levene Test Example
87
Levene Test Example
88
Most Commonly Used Hypothesis Tests
Section VII: Application of Tools -Improve Phase
Lean Tools
Control Charts
◆ Value stream mapping◆ Minimization of non-value added activities◆ Set-up reduction (SUR)◆ The use of standard operating procedures◆ Total productive maintenance◆ Poka-yoke techniques to prevent or detect errors ◆ Workplace organization (5S approach)◆ Just-in-time principles◆ Continuous flow manufacturing concepts 90
Lean Techniques
Control Charts
91
Section VIII: Application of Tools -Control Phase
Rational Subgroups
Nonrandom Patterns
Variables Control Charts:
- IMR, Xbar & R, Xbar & s
Attributes Control Charts:- p, np, c, u
◆ One of the most important concepts in control charting is the rational subgroup.
◆ A rational subgroup is just a subset of a group intended to be as homogeneous as possible. In that way, we will be able to compare the variation within each subgroup and compare the variation between subgroups over an extended period.
◆ But how do we determine the appropriate subgroup size? That will depend on the process being studied. 93
The Rational Subgroup
94
Nonrandom Patterns
◆ Test 1—One point more than 3 sigma from centerline◆ Test 2—Nine points in a row on same side of centerline◆ Test 3—Six points in a row, all increasing or all decreasing◆ Test 4—Fourteen points in a row, alternating up and down◆ Test 5—Two out of three points in a row more than 2 sigma from
the centerline (same side)◆ Test 6—Four out of five points in a row more than 1 sigma from the
centerline (same side)◆ Test 7—Fifteen points in a row within 1 sigma from the centerline
(either side)◆ Test 8—Eight points in a row more than 1 sigma from the centerline
(either side)95
Nonrandom Patterns
◆ Individuals and Moving Range Chart» Subgroup size = 1
◆ X-bar and R Chart» Subgroup size from 2 to 5
◆ X-bar and s Chart» Subgroup greater than 5
96
Variables Control Charts
97
Variables Control ChartsA
vera
geR
ange
◆ p Chart: percent defectives◆ np chart: number of defectives◆ c chart: number of defects◆ u chart: avg. number of defects per unit
98
Attributes Control Charts
99
Attributes Control Charts
100
Attributes Control Charts
101
Attributes Control Charts
Example of p Chart
Section IX: Application of Tools for an Annual Product Review
Descriptive Statistics
Process Capability
Hypothesis Tests
103
Process Monitoring for an Annual Product Review (APR)
104
Process Monitoring for an Annual Product Review (APR)
105
Process Monitoring for an Annual Product Review (APR)
106
Process Monitoring for an Annual Product Review (APR)
107
Process Monitoring for an Annual Product Review (APR)
108
Process Monitoring for an Annual Product Review (APR)
Thanks !!!
109
Business Excellence Consulting Inc.
110
Business Excellence Consulting Inc.
111
Business Excellence Consulting Inc.
112
www.calidadpr.com