Lean Six Sigma Final

PDCA

Cycle that provides the basis for continuous improvement, persistence is key in continuous improvement

Core method of Lean: Plan-Do-Check-Act

Plan: Devise or revise process components to improve results

Do: Implement the plan and measure it’s performance

Check: Assess the measurements and report the results

Act: Decide and implement the changes needed to improve the processes

• focus on the customer

• attack waste, expose problems, streamline production

• Engage employees at all levels

DMAIC

Core method of Six Sigma Define, Measure, Analyze, Improve, Control

Define: identify customers and their priorities

Measure: determine how to measure the process and how it is performing

Analyze: determine the most likely cause of defects

Improve: identify means to remove the causes of defects

Control: determine how to maintain the improvements

• Focus on the customer

• Improve processes, reduce variation

• Engage Employees at all Levels

Similarities of Lean and Six Sigma

Both emphasize systems thinking and process management

• focus on the customer

• Engage Employees at all Levels

• continuous improvement

• high-level leadership

Six Sigma

philosophy and set of methods companies use to eliminate defects in their products and processes

• seeks to reduce variation in processes that lead to product defects

• NO MORE THAN 3.4 DPMO

Types of Waste

1. Overproduction (production that is more than needed or before it is needed)

2. Waiting (wasted time waiting for the next step in a process)

3. Transportation (unnecessary movement of products & materials)

4. Inefficient Processing (more work or higher quality than is required by the customer)

5. Inventory (excess products and materials not being processed)

6. Motion (unnecessary movements by people or poor ergonomics)

7. Quality Failures/Defects (efforts caused by rework, scrap, and incorrect information

*Non Utilized Talent (under utilizing people’s talents, skills & knowledge) \

• 8th waste

Push Versus Pull

Push System: materials is pushed into downstream workstations regardless of whether resources are avalible

Pull System: material is pulled to a workstation as needed

SIPOC

Suppliers, Inputs, Processes, Outputs, Customers

diagram that outlines processes and gets a sense of process boundaries

focus on aspects of the process that are critical to quality

DEFINE tool

DMAIC Tools

Define: Project Charter, SMART objectives, SIPOC, CTQ

Measure: VSM (current state map), SPC (Control Charts)

Analyze: Hypothesis Testing, Root Cause Analysis, Experimentation, SPC (Control Charts), Ishikawa Diagram

Improve: Pilots & Experimentation, SPC (Control Charts), Set Performance Metics & Target performance levels

Control: Standardization, Visual Management Systems, Error Proofing

Best Practices for charts, figures, and tables

Labeled Axes, Cumulative Percents only go to 100, Chart Title,

• figured should be stand alone (no accompanying text needed to interpret the chart

Appraisal vs. Prevention Costs

Appraisal: after the fact quality management expense

• final product testing

• limited by sampling

• does not improve process capability

• reduce the chances of having an internal failure

• ex. Inspections

• ex. Audits

Prevention: investments before the fact for quality management

• ex. employee training

• ex. Standardization & SOPs

• ex. Preventive maintenance

• ex. FMEA

• ex. Quality planning

Affinity Diagrams

useful for processing brainstorm output

• allows team organization of ideas

• can reduce output of brainstorm into manageable categories

• may uncover previously unrecognized connections

works by organizing ideas into multiple categories which are then labeled with a common theme

Pareto Analysis & Charts

Based on the “Pareto principle” - separate the vital few from the trivial many

Analysis that uncovers relative frequency or impact of various issues

Chart is organized in descending order, most frequent to least frequent, to identify the critical few areas to address first for highest impact

Descriptive Analysis

1. What data is available

• create a table aka data dictionary

2. Describe Data

• Central Tendency (Mean & Median)

• Dispersion (minimum, maximum, standard deviations)

• Correlations (for pairs of continuous variables)

• Missing Data

• Unusual Observations

3. Visualize Data

• create line charts for time series, histograms for distribution, scatter plots for associations

4. Interpret & Report Descriptive Analysis

Measure in DMAIC

Value Add vs. Non Value Add

Value Add: work the customer is willing to pay for

Non Value Add: work that the customer is not willing to pay for and which does not otherwise add value, this is considered waste

Visual Management

process conditions are easy to observe and interpret

• color coding

• labels

• outlines

• shadow boards

5 S

Sort - determining what is necessary and what should be moved or discarded

Set in order - orginaze logically to suit task

Shine - clean

Standardize - use visual management tools to indicate where each item should be stored

Sustain - set routines to make sure new order is maintained (weekly scorecard, routine shine activity)

Value Stream Mapping

• a special type of flowcharting tool used to analyze where value is or is not being added as materials flow through a process

• start by defining value stream boundaries (where map will start/stop)

VSM includes

1. Physical materials flows (activities, waiting/storage)

2. Information Flows (planning, scheduling, quality approvals, shipping approvals)

3. Major suppliers & customers

4. Existing process data (information on cycle times, available operating time, first pass yield rates, scrap/defect rates, changeovers, “pain points” etc.)

5. Decision points (where flow spits)

Dirty Dozen

12 Common Causes of Human Factor Error

• Lack of Communication

• Complacency

• Lack of Knowledge

• Distractions

• Lack of Teamwork

• Fatigue

• Lack of Resources

• Pressure

• Lack of Assertiveness

• Stress

• Lack of Awareness

• Norms

Ishikawa Diagram

Fishbone Diagram

1. Identify key performance measure or event of interest

2. Use Categories

• Manufacturing: (People, machines, materials, methods, measurement, environment)

• Transactions: (People, processes, procedures, policies, measurement, environment)

1. Brainstorm under each category

• get as close to variable levels

4. Label each fact/variable as C, N, or X

• C = variables that are already controlled

• N = variables that are not controlled or held constant

• X = variables we believe are key to process performance and which we can influence (x variables may turn into c)

Principles of Cause & Effect

1. Time order Precedence

• X occurs before Y

2. Covariance

• when X moves Y moves

3. Rule out Alternative Causes

• mitigated with experiments

• can’t be statistically calculated

FMEA

Failure Mode Effect Analysis

Failures are prioritized for mitigation based on FREQUENCY of occurrence
and SEVERITY of effects.
When to apply:

• New process/products

• Process/product changes

• Periodically for existing processes/products

• To analyze failures

How to conduct:

Process Function (What is being done)

Failure Mode (How could the function go wrong)

Failure Effects (What is the result of the failure)

Severity

Potential Causes or Mechanisms

Occurrence

Existing Control

Detection

Risk Priority Number

IMPROVE - DMAIC

Improve Best Practices

Regardless of Root Cause Analysis (RCA) method applied, improve should focus on root cause

Validate underlying causes through investigation, observation, and experimentation

• Investigation: Underlying cause validated by consulting with process experts, operators, data collection

• Observation: RCA involves some level of speculation regarding underlying factors. When relevant/adequate data is not available, it should be collected to verify causes.

• Experimentation: Where possible, manipulate factors (X Variables) through controlled experiments to verify cause and improvement

Based on Validated Root Cause:

• Brainstorm potential improvements (use cross-functional expertise and process experts)

• Consider performance objectives (How do we get from X to Z)

• Benchmark improvements and performance

  • Internal benchmarks: What is best practice or performance in your firm for smaller processes

  • External benchmarks: What is best practice or performance in your industry for similar processes

• Evaluate & Mitigate risks

  • Failure Modes & Effect Analysis

Questions to ask to develop evaluation criteria

1. What should the ideal situation look like

2. What are the barriers to implementing a solution

3. How will cost impacts be measured

4. How quickly can an improvement be implemented

5. What are risks of implementation

6. What other processes could be impacted

Define criteria for evaluating improvements:

• Critical to Quality

  • Customer requirements & Customer satisfaction

• Business needs

  • Alignment with strategy

  • Contribution to performance objectives including time, cost, revenue, market share regulatory compliance

  • cost, risk to implement

Organize & define potential improvements:

• Synthesize/group ideas using affinity diagrams or similar processes

• Define Improvement ideas

  • What is it exactly?

  • How should it be measured?

  • What additional data, investigation, experimentation is needed to define the improvement

Start Small

1. Focus improvements on one area or process or one location

2. Carefully consider scope of improvements - quality v quantity

3. Pilot solutions, evaluate performance, seek feedback

4. Evaluate readiness for change

5. Develop change management/implementation plans

Make case for broader implementation

1. consider motivations and incentives of all stakeholders

2. Articulate benefits (individual, team, business unit, organization)

3. Build support among stakeholders

4. Communicate implementation plan, including performance measurement

5. Make performance measures visible

Control Best Practices

Control practices are intended to:

• Sustain improvements

• Evaluate continued effectiveness of improvements

• Identify need for additional problem solving/change

Developing control - questions to ask

1. Measurement - how should performance be measured

2. Standards - what standard should be set for adequate performance

3. Change management - what is the review process for the procedure/standard/process/improvement

Checklist Manifesto

2 Major Challenges to preventing mistakes

1. Lack of Knowledge - either the knowledge doesn’t exist, or we don’t have that knowledge

2. Memory or inattention (distraction) issues - we have the knowledge, but we forget or are distracted/inattentive etc. so we don’t apply the knowledge

What kind of tasks are amenable to checklists

• Routine (we perform it multiple times are there are multiple steps, useful when multiple persons need to execute task repeatedly)

What kind of tasks are not amenable to checklists

• Tasks where there is no realistic/consequential problem if performed incorrectly

• Tasks that have never been performed before

VOC - Voice of Customer

Describes what customers want or expect (their needs, preferences, and quality requirements)

Costs of Control

Prevention & Appraisal Costs

Quality Costs

the total costs associated with ensuring and maintaining product quality, including prevention, appraisal and failure costs (internal + external)

Cost of Failure

External

• Defect gets to the customer

• Costs are higher than internal failures and are hard to measure

• affect liability & reputation/brand

• ex. returns/warranty claims

Internal

• defect found before getting to customer

• ex. scrap

Dimensions of Quality

key aspects customers use to judge a product or service’s excellence

help firms understand what drives customer satisfaction and guide product design and improvement.

• Performance

• Features

• Reliability

• Conformance

• Durability

• Serviceability

• Aesthetics

• Perceived Quality

Variable Types

• Categorical (Qualitative)

  • Nominal (e.g., gender)

  • Shirt Size (e.g. S, M, L).

• Numerical (Quantitative):

  • Discrete (counts)

  • Continuous (measurements).

Conformance Quality

Those that must be present for a product to be considered “good enough” to enter the next process or to be sold. Will meet all specifications.

Conformance Quality Is:

• Determined by the producer to meet customer requirements are

• the degree to which process outputs meet specifications

• typically includes attributes that ensure the product meets regulatory requirements

Specifications

documented target values or standards that define how a product or process should perform → set acceptable limits or tolerances for CTQs to ensure consistency and customer satisfaction.

Process control vs process capability

Process control

• focuses on monitoring and maintaining process stability over time using control charts to detect unusual variation.

Process capability

• measures how well a stable process can produce output within customer or design specifications.
  

CTQ Attributes

key measurable characteristics of a product or process that directly affect customer satisfaction → translate customer needs (VOC) into specific, actionable performance requirements.

Statistical Process Control (SPC)

uses statistical methods, mainly control charts, to monitor and maintain process stability over time

• helps detect unusual variation early, prevent defects, and ensure consistent, predictable quality in products and processes.

• provides a leading indicator of process performance

testing and inspection in quality management

examining products or processes to verify they meet specifications and quality standards

• detect defects or deviations before products reach customers, ensuring reliability and compliance.

Operational Definition

clearly describes how a variable or process will be measured or identified so it can be observed and repeated consistently

• remove ambiguity so everyone measures & interprets data the same way.

X-bar chart

Centerline: mean of means

LCL & UCL: calculated based on sample

Signal: mean of sample observations

Optional: specification limits (LSL & USL – determined by producer for conformance quality)

R Bar Chart

Centerline: mean of ranges

LCL & UCL: calculated based on sample

Signal: sample ranges

Optional: specification limits rely on the assumption of normality

Rules for interpreting X & R Bar Chart

Does the process violate an LCL or UCL?

• if yes then process is statistically OOC and there is likely a specific reason (non random error)

Rule 1: 10 or more consecutive samples above or below the centerline indicate OOC,

Rule 2: non-random patterns in chart suggest special cause variation & should be investigated

if any of these criteria are met the process is NOT STABLE aka OOC

Current vs Future VSM Maps

Current - how process operates in current moment

Future - how process should ideally operate after improvements are implemented

Control Chart Assumptions

Control charts assume a stable, consistent process in which data are independent, normally distributed (for variable charts), measured reliably, and collected in rational subgroups so that only common-cause variation is present when establishing control limits.

Corrective Action

Actions taken after root cause analysis to address the underlying cause(s) of issues, intended to prevent future, similar occurrences

Verifying Root Cause

Pilot Testing and Experimentation

•Control as many contributing factors as possible

•Vary X variables, measure changes in Y (outcome)

•Ideally, hypothesis tests to demonstrate the statistical effect of change

•Small scale pilot allows adjustment of improvement, employee feedback prior to broad implementation

Standardization

Objectives:

• Sustain improvements

• Evaluate continued effectiveness of improvements

• Identify need for additional problem solving/change

Key Principles:

• Standards define what standard should be set for adequate performance

• Procedures outline how a task must be done:

  • Include standards for acceptable quality (completion, accuracy, etc.)

  • Incorporate change management - regularly updated for consistency, changes in practice & requirements

• Checklists:

  • Not for training, but for use by those who already have competency

  • Selective use depending on situation

• Job aids:

  • Memory aids for those who have been trained

  • More detailed than a checklist

  • Often implemented when tasks are completed infrequently

• Tools include: SOPs, job aids, training, and visual management systems

Error Proofing (Poke Yoke)

Objectives:

• Verify root cause (evidence, logic, hypothesis testing)

• Prevent errors from occurring or detect them before they cause problems

• Part of the Control phase tools

General Principles & Examples:

• Physical design constraints: Dedicated material hoses have unique fittings that cannot be used for other materials because connectors do not match (prevents wrong material use)

• Checklist verification systems:

  • 2-person checklist verification on the floor

  • 1-person checklist verification in control room by radio

  • End of checklist review

  • End of shift review with incoming shift

Root Cause Analysis

Objectives:

• Solutions should NOT be generated or evaluated until Root Cause Analysis is complete (exception: low risk, quick improvements)

• Identify key performance measure or event of interest

• Use Ishikawa Diagram (Cause & Effect) with categories appropriate for your context

• Label factors as C (controlled), N (not controlled), or X (key variables we can influence)

• Work from factors "close" to performance or event outward to get to variable level if possible

General Principles:

• Root Cause Analysis occurs in the "Analyze" phase of DMAIC

• Generate as many meaningful ideas for each category

• Facilitator must clarify meaning of terms and supporting evidence

• Facilitator must keep group from jumping to solutions

• Validate underlying causes through investigation, observation, and experimentation

• Once underlying causes are identified, verify with data/evidence, test through experiments, and investigate cost impact before implementing solutions