Scribd
Upload
50 views

Six Sigma

This document provides an overview of six sigma concepts and methodology. It discusses key figures in quality management like Joseph Juran and William Deming. It also explains six sigma principles and definitions, the DMAIC process for improvement projects, capability analysis, and quality tools like fishbone diagrams.

Uploaded by

MANOUJ GOEL
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
50 views

Six Sigma

This document provides an overview of six sigma concepts and methodology. It discusses key figures in quality management like Joseph Juran and William Deming. It also explains six sigma principles and definitions, the DMAIC process for improvement projects, capability analysis, and quality tools like fishbone diagrams.

Uploaded by

MANOUJ GOEL
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 4

Six sigma notes

Continuous improvements - Thro learning and feedback(reactive)


Continual improvement proactive approach challenge process

Quality of the process is defined by creating excitement in the customer that they get what
they expect
P/E is perception to expectation ratio
93% revenue comes from repeat customers, a good P/E (greater than 1) is important
1970 IOS international organization for standardization was incepted post equipment
failure at WW2. This was later renamed to ISO.
Y=5X where y is the cost of repair/redo and x is cost of the material
Higher inspection actually means lower quality, but is portrayed as higher quality and
customer is charged

Joseph M Juran Father of quality management


JUSE Japanese Union for Scientists and Engineers provides for the Deming award (also
called total quality management award).
o Unlike other awards minimum of 800 points need to be taken , no relative grading
o Highest JUSE awards held by INDIA
o Equivalent award in the US is Malcolm Balridge award
William Edwards Deming was hand in hand with Juran in creating Japans success
Walter A Shewhart- Father of statistical quality control
1975-1977 Motorola bankruptcy (previously top 10 in NYSE for 85 years)
Bill Smith Father of 6 sigma

Lessons for quality

Understand customer needs


Never outsource understanding customer needs

Understand customer needs - > translate to product features -> product design - > process
design -> key parameters (a measurable parameter without which the process fails, has
major influence) this becomes control parameter
Best examples of 6sigma in India IRCTC, Sundaram fasteners TVS, GE finance, GE noiceless
motors for washing machines

Quality control and quality improvement are different Former is to ensure Mu is within
tolerance, bring them back to desired limits, whereas latter is Improving quality by reducing
the Mu (average of defective pieces) or reducing std deviation

COPQ- Cost of poor quality (30% of revenue, need to be reduced)


3 sigma 66,000 defects per million (99.73%)
4 sigma 6210 d/mn
6 sigma 3.4 d/mn (99.99967%)
Since 6 sigma is expensive, companies in general stop at 4.5 sigma (99.997%)

6 principles of 6 sigma : Genuine focus on customer, decision by fact (data), Action is where
the process is, Proactive management , boundary less collaboration, promote risk & tolerate
failure

When does problem qualify for 6 sigma

When process deliverable doesnt meet customer reqirement


Problem must be chronic and not sporadic (Use CAPA for sporadic)
Problem must not have a known solution
Must be cross functional in nature (if its interdept do Kaizen)

Task happens within a dept whereas process is across depts.

DMAIC Define Measure Analysis Improve Control

Team composition for DMAIC Avg size 5, memebrs must be respected by the depts.,
should be self motivated, must be SMEs (Subject matter experts), STARS 40%, RATS 60%.
AVg duration 4 months 8 hr/ week

Day-2

Lean manufacturing started with FORD Motors, later manifested by Toyota


Takt time is the time available and not time taken for a process, eg. If 100 units are to be
made in 8 hours, process needs to be reorder to suit 100/8 units/hr

3 types of activities value added, Non value added, Value enabling

VA customer ready to pay; NVA like multiple documents review, signtatues etc which
can be reduced or eliminated, VE support activities like O&M, HR

If any of the following activities are there, its a sign of NVA

1. Transportation
2. Inventory
3. Motion (human movement)
4. Human
5. Waiting
6. Overproduction
7. Over processing
8. Defects
Efficient do the right way; Effective - do the right thing
Process time : Time taken actually for the VA activity; Lead time is time in between process
(waiting) NVA or VE; Process time : lead time = 0.05 : 1; Tool used to identify these two is
Value Stream mapping

Push system, Pull system, push & pull system Lean is always Pull; this required Kanban
cards

Define

Define Identify neutral customers, collect Voice of customers, Make an affinity diagram to
categorize
On these, make the KANO model (must be, more then better, delighter). Must be is the
most critical/ The 2 most critical needs becomes the CTQ (Critical to quality)

Meetings

Agenda point for the meeting


Punctuality
Minutes of meeting
Team leader must summarize
RACI chart (responsibility, accountability, consutancy, information)

Charter

Problem statement : define the problem in %, define the change attempted, should NOT have a
cause & solution, no scope for blame, scoping must be manageable

Define Milestones

Map the process

SIPOC model (supplier input process output customer) defines the scope, identify the team

DMAIC Measure (collect data)

Data is of 2 types discrete or continuous , former when we count; latter when we


measure, discrete take the units of whats being measures but continuous has SI units

Check if data to be collected is qualitative or quantitative. If former then do AAA, latter GRR
AAA (Attribute agreement analysis) we are measuring against a standard (Kappa should be
greater than 0.9 for acceptance )

GRR Check contribution variation if less than 10 study accept data; if greater than 10,
check % total study variation. This must be less than 30

We check for repeatability and reproducibility


Define sample size = [(1.96 x sigma)/Precision]2
(thumb rule 10 in 1000)
Sampling must be randomized or stratified. Former is when population data is
homogenous, latter is when heterogeneous (keep collecting till u encounter
homogenous)

Defective vs defects

Defectives = P= (no. of defectives / no. of units sampled)


PPM = P x 106
With PPM, check sigma table for sigma level.

For Finite defects; DPO = total no of defects/ (sample size x opportunities per unit)
DPMO = DPO x 106 , then check sigma table for sigma level

For infinite defetcs; DPU = total no of defects / sample size


DPM = DPO x 106 , then check sigma table for sigma level

Now next, with the sample do the Normality test (Anderson darling test). If p value greater than
0.05 is accepted and lesser than 0.05 is not acceptable, check for stability (if process is within 3
sigma it si stable)

Capability of process : (USL Mu) / 3 sigma or (LSL Mu) / 3 sigma whichever is lower

Brainstorming : Has a facilitator, scribe and the 6 sigma team

Uncontrollable cause: Causes within ones circle of influence can be controlled, within
circle of concern cant be
Direct improvement causes : Excuses like not havig equioment, training etc. satisfy few
points based on budget
Potential root causes : Fish bone diagram (Ishikawa) 5M1P (material, machine, method,
measurement, mother nature, people). These causes are further dicidde into events and
theories.
o Events : Occurences/instances/ facts that caused problem. To validate do
PERETTO analysis
o Theories : Forecasts/ hypothesis do hypothesis testing

You might also like