LSSG Black Belt Training

1
LSSG Black Belt Training
Analyze What are the Key Drivers for
Breakthrough Performance?
2
Qualitative vs. Quantitative Analysis
Qualitative Goal - Identify and prioritize
critical Xs Tools - Pareto Charts, CE
matrix/fishbone diagram, Brainstorming, Detailed
As-Is process, Non-value add analysis,
Benchmarking, FMEA
Quantitative Goal Find Root Causes and
Validate critical Xs. Tools - Hypothesis Testing
Concepts, Multi-Vari Studies, Simple and Multiple
Regression, ANOVA
3
Process Analysis

• Process
• Any part of an organization that takes inputs and
  transforms them into outputs repeatedly
• Process Analysis
• The ability to diagram a process, identify its
  flow, potential bottlenecks, and determine its
  capacity
• Process Flowchart
• The use of a diagram to present the major
  elements of a process.

4
Example of a Process Flowchart
5
Process Analysis Definitions

• Cycle Time - The average time between successive
  completions of units
• Throughput Time - The total time it takes for a
  unit to move through the process from beginning
  to completion
• Throughput Rate - The inverse of cycle time,
  expressed as number of units per time period
• Capacity - The measure of output per unit of time
  when system is busy
• Utilization - The ratio of the time that a
  resource is actually active relative to the time
  that it is available for use
• Continuous Flow - A job performed in a steady
  flow with no wasted motion, no interruptions, and
  no waiting
• Work-In-Process - Any type of unfinished work
  waiting to be performed at a later time
• Bottleneck - The slowest operation in the
  process defines cycle time and capacity of the
  process

6
Process Analysis for Processes without Buffers

• What is the cycle time of this process?
• What is the throughput time of this process?
• What is the throughput rate of this process?

Batch1 0-20 20-50
B Starved for 20 min
Batch2 20-40 50-80
A Blocked for 10 min
Batch3 50-70 80-110
A Blocked for 10 min
7
Process Analysis for Processes without Buffers
8
Littles Law
Throughput Time
Throughput time in WIP
How to reduce Lead (Throughput) Time?
9
Process Analysis with Buffers
Example of Using Littles Law
Bake 100 meat loaves Baking time 60 minutes
Bake 100 meat loaves Baking time 60 minutes
CT 30 min
Individual Pack 100 meat loaves 48 min
CT 48 min
10
FMEA Template
Severity How serious is the failure to the
process to business results? (1 minor, 2-3
annoyance, 9-10 very high/most severe)
Detection Likelihood that a defect will be
detected by controls before the next (subsequent)
process (1-2 very high, 9 very low, 10
absolutely cannot detect)
Occurrence Frequency of failure mode (1
remote, 9 inevitable, 10 certain)
11
Benchmarking
Phase 1
Phase 2
Phase 3
Phase 4
Phase 5
Phase 6
Internal Benchmarking
Reverse Engineering
Different Industry Process Benchmarking
International Benchmarking
Similar Industry Process Benchmarking
Strategic Benchmarking
Standardize internal best practices
Study/ Incorporate best features of competitive
products
Study/ Incorporate best process practices
Study/ Incorporate applicable best process
practices
Study other approaches to strategy
development Leverage core competencies and trends
Identify global strategies, customers, new
partners, products, and processes
12
Problem Solving through Appreciative Inquiry
Appreciative Inquiry extends the first step of
the benchmarking process by focusing on the best
of what is
Source David Cooperrider et all, Appreciative
Inquiry
13
Statistics Roadmap for Analyze(Black Belt Topics)
DMAIC
Find out what is causing the problem
ANALYZE
ID and Confirm with Data
VALIDATE ROOT CAUSES
Is there really a relationship? What to compare?
Amount of risk? Calculation and comparison to
test statistics
HYPOTHESIS TESTS
Measured Data, E.g. time, length and weight
Compare Proportions Count Data Etc
CONTINOUS DATA
DISCRETE DATA
Definitions www.asq.org/sixsigma/terms/a.html
14
Hypothesis Testing Roadmap
15
Testing of Means - Roadmap
Comparing Means
3 or more Factors
2 Factors
1 Factor
1 Sample
2 Samples
2 or more samples
s not known
Indep- endent
s known
paired
1-sample Z-test
Two way ANOVA
ANOVA GLM
One way ANOVA
1-sample t-test
2-sample t-test
Paired t-test
16
Testing of Variation - Roadmap
Comparing Variances
2 Sample
More Than 2 Samples
1 Sample
Test for Equal Variance
1 Variance
2 Variance
Test
Test
Bartletts Test
F- test
Levenes Test
Levenes Test
Descriptive Statistics
17
Testing of Proportions - Roadmap
18
Non-Parametric Testing Roadmap