STAT 8140 STUDY GUIDE – Class 1 I.

II.

Introduction to Six Sigma Setting Six Sigma is about improving operating performance, which translates to saving money. Thus, there is widespread interest by organizations in implementing Six Sigma methods. While the basis of most Six Sigma methods is statistical, it is necessary to learn many non-statistical tools to be effective. For example, many team-based tools are part of the Six Sigma Body of Knowledge (BoK). Since improving quality performance is often a consequence of Six Sigma projects, it is desirable to start a Six Sigma journey by understanding the twentieth century development of quality leading to the birth of Six Sigma in 1986. Lean Six Sigma then developed in the mid-1990’s by integrating Six Sigma with lean Toyota Production System methods. History of Six Sigma Table 1.1 gives a timeline for major events related to Six Sigma (p. 972). Many events are related to quality control and quality improvement. A key learning is that Six Sigma is not a methodology conceived by Motorola and launched into the business environment. Rather, Six Sigma is the result of integrating a century of quality method development into a DMAIC problem-solving framework. This framework was new, but some would say it is an extension of the Deming PDCA problem solving cycle. Thus, Six Sigma is an umbrella encompassing many existing statistical tools integrated into the DMAIC problem solving process. Incorporating lean methods into the “I” or Improve phase of DMAIC is a natural development. Over time Six Sigma will likely change again, perhaps incorporating quality system methods such as Baldrige or ISO 9000 into DMAIC. Some of the key influences on Six Sigma methods are given below (p. 972-973): http://geekswithblogs.net/srkprasad/archive/2003/10/27/276.aspx http://www.syque.com/improvement/a_encyclopedia.htm http://www.businessballs.com/dtiresources/quality_management_gurus_theories.pdf

1) W. Edwards Deming (1900-1993) Deming was an American professor and consultant who is credited with advising the Japanese (see G3 Deming Award) on how to build a quality system. He lectured widely in the U.S. in his 3 or 4-day seminars for management. He developed: • 14 Principles for management (p.974) • Plan-Do-Check-Act (PDCA) improvement cycle • Wide use of SPC • Special and Common Cause variation • Management is responsible for quality • Build quality into a product. He lectured quarterly at Ford in the mid 1980’s and would often start a seminar to upper management with the statement, “Management is responsible for poor quality.” He could be very intimidating. http://en.wikipedia.org/wiki/Edwards_Deming

Opened: 1/14/11 File: S8140SGC1d090410.docx

Page 1 of 10

V. Kane

2) Joseph Juran (1904-2008) 2008) As an engineer/manager, Juran worked for Bell Labs and Western Electric and developed many quality control practices that are used in Six Sigma: • Project focus for quality improvement • Pareto Principle • Management is responsible for quality using Juran’s Trilogy - Quality planning - Quality control - Quality improvement He published a large Quality Control Handbook (1951, last edition 1999) and founded the Juran Institute (1979). http://en.wikipedia.org/wiki/Joseph_M._Juran http://www.skymark.com/resources/leaders/juran.asp

3) Genichi Taguchi (1924 1924- ) Japanese engineer and statistician who developed a fundamental concept used in Six Sigma the loss function. Taguchi chi claimed that loss to an organization or society in general was a parabolic function with minimum loss at a process target value, typically the midpoint of the specification limits.

This idea changes the idea of “meeting specifications”. Merely meeting ing specifications can result in a significant loss, if the process output has a mean near the specification limits. Thus, great effort should be expended to target processes optimally to minimize loss. Taguchi developed two other concepts that are fundamental in Design for Six Sigma (DFS). Products should be made more robust to environmental variables (e.g., temperature, moisture, age, etc.) using a signal-to-noise ratio. Robustness of design is a key goal for DFS using Taguchi methods. These methods use orthogonal arrays in design of experiments to optimize designs. http://en.wikipedia.org/wiki/Genichi_Taguchi wikipedia.org/wiki/Genichi_Taguchi http://www.skymark.com/resources/leaders/taguchi.asp http://elsmar.com/Taguchi.html

4) Phil Crosby (1926-2001) 2001) With the publication of his book Quality is Free (1979), Crosby refuted the then hen prevalent management belief that quality was costly. Much of the cost focus of Six Sigma can be attributed to Cosby’s work. His basic ideas were: • the definition of quality iis conformance to requirements defined by customers • the system of quality is prevention • the performance standard is zero defects • the measurement of quality is the price of nonconformance • quality improvement is a journey • the cost of poor quality (COPQ) is high Opened: 1/14/11 File: S8140SGC1d090410.docx

Page 2 of 10

V. Kane

III.

What is Six Sigma - DMAIC Six Sigma is a relatively new, since 1986, methodology for improving work processes. (See D4 below for a brief history.) At the core of all Six Sigma methodology is a problem solving process, DMAIC • D – Define: identify the opportunity • M – Measure: evaluate the measurement system, collect data • A – Analyze: use both simple and targeted statistical methods to identify a problem • I – Improve: improve the process • C – Control: put process controls in place to hold the gains All DMAIC activities are conducted as part of a defined project that identifies the problem being investigated and the cost opportunity if the problem is resolved. All project activities are conducted as part of teams of employees, typically from the area with the work process being investigated. Table 1.2 gives typical DMAIC activities that occur during a project.

Opened: 1/14/11 File: S8140SGC1d090410.docx

Page 3 of 10

V. Kane

Table 1.1 Timeline of Changing Quality Systems Start Year 1920s 1926+

Author(s) Shewhart Juran

Label SPC

Unique Characteristics Process Monitoring by Statistical Process Control Top management is responsible for quality, improvement projects, Pareto principle, etc. 1947+ Deming 14 Points Many lectures in Japan addressing: 1) management responsibility 2) understanding variation 1949 Ohno TPS Toyota Production System (forerunner of Lean); uses Kaizen teams, JIT inventory, flows, waste elimination & cycle time reduction. Ideas from Deming and the Ford Production Systems had a major influence. http://www.youtube.com/watch?v=jlcrqb2qMAA 1950+ Taguchi Robust Developed methods to make product and process designs more robust Design to outside influences. 1950 JUSE Deming Award for best quality companies; Administered by Japanese Union Prize of Scientists and Engineers 1951 Feigenbaum TQC Total Quality Control: uses company-wide approach of methods from Deming, Juran, Cosby & Ishikawa 1962 Ishikawa QC Circles Quality Control Circles in Japan – employee groups working on improvements using 7 statistical tools 1979 Crosby Quality is Free book established concepts zero defects goal & cost of nonconformance is large 1980 NBC White Paper Television documentary “If Japan Can, Why Can’t We?” featured Deming and launched American quality revolution. Many Deming lectures follow. 1984 U.S. Navy TQM Total Quality Management company-wide approach to quality: uses Deming, Juran, Crosby & Ishikawa ideas. Based on many TQC principles. http://www.youtube.com/watch?v=Pd_uRGy5RKY 1986 Bill Smith Six Sigma Motorola CEO Bob Galvin launches Six Sigma program. DMAIC problem solving addressing defects and variability using projects involving small teams focusing on internal processes. 1987 NIST MBNQA Malcolm Baldrige National Quality Award launched with criteria similar to TQM 1988 Motorola wins MBNQA, again in 2002 1988 Krafcik Lean Sloan article “Triumph of the Lean Production System” based on TPS principles 1993 Allied Signal (Honeywell) CEO Larry Bossidy launches SS program 1996 GE CEO Jack Welch launches SS program http://www.youtube.com/watch?v=aNMULFcLuIM Companies such as Allied Signal and Maytag combined two Mid 1990+ Lean improvement approaches TPS and Six Sigma. Six Sigma http://www.youtube.com/watch?v=PQspf3q12mo 1999 NIST MBNQA added Education and Health Care categories 2001 U. of WI Wins MBNQA in Education category, graduate (8,000 students) http://www.baldrige.nist.gov/PDF_files/UWStout_Application_Sum Stout mary.pdf 2005 Richland Wins MBNQA in Education category, 2 year (3500 students) http://www.baldrige.nist.gov/PDF_files/Richland_College_Applicatio College n_Summary.pdf Six Sigma Tool/Process (Yellow) Lean Six Sigma Tool/Process (Green)

Opened: 1/14/11 File: S8140SGC1d090410.docx

Page 4 of 10

V. Kane

Table 1.2 Six Sigma DMAIC Activities Define – Establish commitment to key organizational problem 1. Work with management to identify key/strategic problem area 2. Collect preliminary data (KOVs) to frame problem and define likely scope 3. Complete Project Charter and get management approval 4. Develop project metrics and reporting requirements with management 5. Survey existing SOP’s and compare to actual practices. 6. Benchmark relevant areas (product, process, metrics, etc.) 7. Establish breakthrough goals for KOVs 8. Establish working team and get time/resource commitment 9. Assemble and analyze available data related to problem definition 10. Construct process and system maps related to KOVs 11. Identify critical operations/constraints in process 12. Identify initial KIVs 13. Develop Pareto Charts to narrow focus to critical few areas Measure – Start targeted data collection and analysis 1. Prioritize KIVs and KOVs 2. Assemble and evaluate SOPs in critical areas (gaging, set-up, etc.) 3. Validate prioritized measurement systems and improve as required 4. Establish baselines for KIVs and KOVs 5. Evaluate process capability and natural tolerances Analyze – Identify sources of failure and variation in KOVs 1. Identify potential causes of failure and variation 2. Correlate KIVs and KOV variation and establish relationship Y = f(X) 3. Screen KIVs to critical few causes of variation in KOVs Improve – Target high impact areas and expand process possibilities 1. Validate causes of KOV variation 2. Establish and validate improvement actions 3. Revise critical SOPs to optimize performance 4. Expand possibilities of customer-focused improvements Control – Establish System to hold gains in performance 1. Implement Error Proofing where possible 2. Revise SOPs and establish charts to define and monitor performance indicators 3. Develop on-going management feedback for monitoring performance

Opened: 1/14/11 File: S8140SGC1d090410.docx

Page 5 of 10

V. Kane

IV.

Quality Movements of 20th Century 1) Total Quality Control (TQC) This term was first used by Feigenbaum in his book Quality control: Principles, practices and Administration (1951). The book incorporated many then current methods into one of the first overall quality systems. http://www.tenstep.fr/01_Publique/90.3_EspaceDeLaQualite/Anglais/700ArmandVFeigenbaum.pdf

2) Toyota Production System After World War II Japanese industry was largely destroyed. Taichii Ohno and Shigeo Shingo developed The Toyota Production System. They incorporated many ideas from the Ford Production System and TQC. During this period Deming is credited as having great influence on the Japanese development. Six TPS rules: • Do not send defective products to the subsequent process • The subsequent process comes to withdraw only what is needed • Produce only the exact quantity withdrawn by the subsequent process • Equalize production • Kanban is a means to fine tuning • Stabilize and rationalize the process http://www.strategosinc.com/toyota_production.htm

3) Total Quality Management (TQM) The exact origin of TQM is debatable. It is a general term incorporating Japanese methods from the 1950’s in what was called Companywide Quality Control (CWQC) with American TQC and teaching of experts Deming, Juran and Cosby. Criteria for MBNQA used many TQM ideas. http://managementhelp.org/quality/tqm/tqm.htm

4) Six Sigma In 1987 Motorola’s CEO, Bob Galvin, began speaking of a new methodology, Six Sigma and launched an internal quality program “The Six Sigma Quality Program”. He credited Bill Smith with developing the methodology. In 1988 Motorola won the Malcolm Baldrige national Quality Award. Six Sigma started with a focus on reducing defects of all types to a 6σ defect rate, 3.4 defects per million operations. In the early 1990’s other companies started adopting Six Sigma. Jack Welch, CEO General Electric, began speaking widely on the value of Six Sigma to save money. In 2002 Motorola’s Commercial, Government & Industrial Solutions Sector won the Baldrige award again. http://www.pqa.net/ProdServices/sixsigma/W06002009.html http://www.businessballs.com/sixsigma.htm

V.

Lean Manufacturing The originator of the Toyota Production system (TPS) was Taichii Ohno. Shigeo Shingo is credited with developing 5S methods that become part of TPS. Because TPS has a focus on kaizen, continuous incremental improvement, it was natural that Six Sigma companies in the mid-1990’s started using lean method (see F below). This incorporation led to what is now called Lean Six Sigma. Today learning lean methods is considered an important element of learning Six Sigma. http://www.strategosinc.com/just_in_time.htm http://www.superfactory.com/content/timeline.html

Opened: 1/14/11 File: S8140SGC1d090410.docx

Page 6 of 10

V. Kane

VI.

Key Areas of Focus for Lean Six Sigma The term Lean was added to Six Sigma during the mid 1990’s. This was a natural addition with key relevance to the DMAIC Improve phase. During this phase the problem being addressed is being “improved” sometimes using lean tools such as error proofing. However, there is another synergy between the two methodologies that is not as apparent. In the define phase, it is not always apparent what problem should be addressed to reduce cost or improve quality. Lean approaches and metrics can be used to start a Six Sigma project. Some common examples of lean approaches used to help define Six Sigma projects are: 1) Waste Elimination – a fundamental principle of Lean is the elimination of waste, also called muda in Japanese. The key is that waste is defined in an operational manner as seen in table 1.2.

Waste Element

Table 1.2 Lean Seven Categories of Waste Business Outcome

Examples

1. Defects

• Unauthorized system and application changes. • Substandard project execution.

2. Overproduction

• Unnecessary delivery of low-value applications and services.

3. Inventory (Excess)

• Server sprawl, underutilized hardware. • Multiple repositories to handle risks and control. • Benched application development teams.

4. Waiting

• Slow application response times. • Manual service escalation procedures.

5. Transportation

Higher capital and operational • On-site visits to resolve hardware and software expenses. issues. • Physical software, security and compliance audits. Lost productivity • Fire-fighting repeat problems within the IT infrastructure.

6. Motion (Excess)

7. Over processing (Non-valued added processing)

• Reporting technology metrics to business managers.

Poor customer service, scrap, rework, errors Business and IT misalignment, Increased costs and overheads: energy, data center space, maintenance. Increased costs: data center, energy; lost productivity

Lost revenue, poor customer service, reduced productivity.

Miscommunication.

Adapted from Wikipedia

2) Flow Focus – One of the 7 simple statistical tools is a process flow diagram. Lean has a focus on flow of material, information and work. It starts at the work level with a Process Flow Diagram. At the organizational level, a Value Stream Map is used to map the global product value added process from raw material, to product production, to customer receipt. 3) Cycle Time – How long does it take to accomplish work defined in a work process flow diagram? Time is money; reducing cycle time reduces cost. Many Six Sigma projects result in improving the flow of work processes. Customer requirements are defined and questioned, steps are eliminated, simplification is a focus, and controls put in place. All this effort starts with a process flow diagram, which defines where cycle times should be measured. At the lowest level, machine cycle time can refer to the machine time required to produce a part. At a higher level, product development cycle time refers to Opened: 1/14/11 File: S8140SGC1d090410.docx

Page 7 of 10

V. Kane

the time it takes to develop a new product from customer requirement to customer receipt of a product. Cycle time can be important for any work process. 4) Wait Time – Use of cycle time is not the only time-related variable used in lean projects. Customer wait time is a typical measured variable quantifying possible customer dissatisfaction. 5) Pull/Takt Time – One waste to be eliminated in Table 1.2 is over production, making too many products beyond customer requirements. This results in a waste of excess inventory. Use of takt time was developed to produce only the amount of product required by the customer. To define takt time, let Ta = total available time for production and Nc = number of units demanded by the customer then Takt Time T =

்ೌ ே೎

is the definition of takt time. For example, suppose there are 8 hours available in a shift or 480 minutes. There are 80 minutes of lost production time for lunches, breaks, meeting, training, etc. Thus, Ta = 400 minutes. Suppose the customer demand is Nc=200 units per day. The takt time is 2 minutes. Thus, the production work processes should be set-up to produce a unit every 2 minutes. If the current process cycle time is 1.5 minutes, inventory will build creating waste. People and equipment should be balanced to create a 2 min. cycle time. If the cycle time is 2.4 minutes, steps must be taken to reduce the cycle time to 2 minutes to satisfy customer requirements. The use of takt time requires detailed knowledge of the production process and customer requirements. Most organizations do not have this understanding and generate waste. Using takt time it is possible to produce exactly what the customer needs, eliminating waste. Thus, the producer is tied to customer demand, the customer pulls product from the produce eliminating waste. In the same way, the producer’s suppliers are tied to the volume of the producer, producing only what the producer needs. Thus, the producer pulls from its suppliers. The Kanban card is the inventory signal card used to communicate a pull requirement. The total system is called Just in Time inventory management. 6) Kaizen – The use of kaizen is from the Toyota Production System (TPS) where small employee groups would work on continuously improving a work process. Employee teams working on improvement are a fundamental element of Six Sigma; even before Lean. Thus, the marriage of Lean and Six Sigma was natural. 7) Seven Simple Statistical Tools – The employee teams working on kaizen activities in TPS were trained to use basic statistical tools: • Process Flow Diagram • Control Charts • Check Sheets • Histograms • Scatter Charts (x,y plots) • Pareto Diagrams • Cause and Effect Diagram (Fishbone Chart) These tools are used in many Six Sigma programs by employee teams.

Opened: 1/14/11 File: S8140SGC1d090410.docx

Page 8 of 10

V. Kane

8) Visual Management (5S) – A general objective of TPS is to use visual management tools (e.g., colors, andon boards, etc.) to control the workplace. A formal 5S process is used to attain this objective: • Sort – Eliminate unnecessary items • Straighten – Organize and label using visual tools • Shining – super clean • Standardizing – common practices throughout • Sustain – put process in place to maintain http://totalqualitymanagement.wordpress.com/category/management-of-process-quality/toyotaproduction-system-management-of-process-quality/

VII.

Quality Awards As Table 1.1 indicates, quality awards have played an important role in promoting a quality focus within organizations. This interest has fostered the development of quality control methodology. Management often uses a quest for these awards as a methodology to motivate improvement within an organization. Knowledge of the awards and differences between the awards is an important element of understanding Six Sigma. 1) Malcolm Baldrige National Quality Award (1987) http://www.nist.gov/baldrige/

This award (p. 984) is sponsored by the U.S. Department of Commerce, National Institute of Standards and Technology (NIST). 2) Shingo Prize for Operational Excellence (1988) http://www.shingoprize.org/htm/about-us/the-shingo-prize

This is a prize (p. 985) administered by the School of Business at Utah State University. The focus of the award is implementing the Lean principles of Shigeo Shingo. 3) Deming Prize (1950) http://www.juse.or.jp/e/deming/index.html

This award is sponsored by the Japanese Union of Scientists and Engineers (JUSE). The focus is on implementing TQM throughout the organization. Most recipients have been Japanese companies. 4) ISO9000 (1987) http://www.iso.org/iso/home.html

These are a set of standards for an organization administered by the International Organization for Standardization (p.982). ISO 9001:2000 combines the three standards 9001, 9002, and 9003 into one, called 9001. The intent is to provide a framework for attaining and maintaining quality products. Many organizations require their suppliers to attain ISO9001 certification. Some areas addressed include: • set of procedures that cover all key processes in the business, • monitoring processes to ensure they are effective, • keeping adequate records, • checking output for defects, with appropriate and corrective action where necessary, • regularly reviewing individual processes and the quality system itself for effectiveness; and • facilitating continual improvement.

Opened: 1/14/11 File: S8140SGC1d090410.docx

Page 9 of 10

V. Kane

VIII.

Notes for Book Reading 1) Lean Six Sigma – Most practitioners use the LSS terminology for the combination of the 1989 Motorola Six Sigma with the TPS principles of lean. Our text uses the acronym S4/IEE for Smarter Six Sigma Solutions/Integrated Enterprise Excellence (p. xxxi). The author explains that this enhanced version incorporates Total Quality Management (TQM), Malcolm Baldrige Assessment Criteria, Shingo Prize and ISO9000. Since this is non-standard terminology, we will use “Lean Six Sigma” (LSS) to denote the procedures that are studied. For student reading, simply make the mental substitution S4/IEE = Lean Six Sigma. 2) Cost of Poor Quality (COPQ) – There are many hidden quality costs (p.5-7). These costs are considered costs that are the 90% of the iceberg below the water: • Lost management time • Lost opportunity • Rerun or rework • Maintenance • Lost goodwill • Hidden factory repair/rework 3) Key Principle: All work is a process (p.10-11)

Figure 1.1 Work Process Input-Output Model KPIV

Inputs

X

KPOV

Process Operations

Outputs

Y = f(X)

Y

4) Benchmarking (p. 16) – The general concept of benchmarking is determining what other organizations do to see if it can be used. Do not “reinvent the wheel” is the goal of benchmarking. “Not invented here” often stands in the way. There are several types of benchmarking: • Metric Benchmarking - similar organization metrics such as defect rates, customer satisfaction, etc. • Dissimilar Organization Benchmarking – similar metrics or processes • Process Benchmarking – Common business processes such as product development, customer complaint management, etc.

Opened: 1/14/11 File: S8140SGC1d090410.docx

Page 10 of 10

V. Kane