Quality Control — Complete
Notes
1. Introduction to Quality
Quality: Degree to which a product/service meets customer
requirements and specifications; "fitness for use" (Juran) or
"conformance to requirements" (Crosby).
1.1 Dimensions of Quality (Garvin's 8 Dimensions)
1. Performance
2. Features
3. Reliability
4. Conformance
5. Durability
6. Serviceability
7. Aesthetics
8. Perceived quality
1.2 Quality Control (QC) vs Quality Assurance
(QA) vs TQM
Term Focus Nature
Reactive,
Quality Control Detecting defects
inspection-
(QC) in the product
based
Preventing
Quality Proactive,
defects in the
Assurance (QA) process-based
process
Total Quality Organization- Strategic,
Management wide culture of continuous
(TQM) quality improvement
1.3 Cost of Quality
Prevention cost: training, quality planning, process
control
Appraisal cost: inspection, testing, calibration
Internal failure cost: scrap, rework (before delivery)
External failure cost: warranty, returns, complaints
(after delivery) — most expensive
2. Statistical Quality Control (SQC)
SQC uses statistical methods to monitor and control
quality. Divided into:
1. Statistical Process Control (SPC) — Control charts,
used during production
2. Acceptance Sampling — Used for lot
acceptance/rejection (before/after production)
2.1 Variation in Quality
Chance (common) causes: inherent, random, small
variations; process is "in control."
Assignable (special) causes: identifiable, large variations
(tool wear, operator error, bad material); process is "out
of control."
2.2 Basic Statistics Used
Mean (x̄), Range (R), Standard deviation (σ)
Process capability depends on the natural variation of
the process.
3. Control Charts
Used to distinguish between chance and assignable causes
of variation, plotted over time with:
Centre Line (CL) = process average
Upper Control Limit (UCL) = CL + 3σ (or 3σ
equivalent factor)
Lower Control Limit (LCL) = CL − 3σ
(3σ limits give 99.73% confidence that points will fall within
limits if the process is in control.)
3.1 Types of Control Charts
Type Data Used For
Measurable data
Variable Continuous quality
(length, weight,
Charts characteristics
diameter)
Detects shift in
— x̄ chart Sample means
process average
—R Detects change in
Sample ranges
chart process variability
Attribute Countable/pass-fail Discrete quality
Charts data characteristics
Varying/constant
— p chart Fraction defective
sample size
— np Constant sample
Number defective
chart size
Constant sample
Number of defects
— c chart size (e.g., defects per
per unit
item)
— u chart Defects per unit Varying sample size
3.2 x̄ and R Chart Formulas
For x̄ chart:
CL = x̄̄ (grand mean of sample means)
UCL = x̄̄ + A₂R̄
LCL = x̄̄ − A₂R̄
For R chart:
CL = R̄ (mean of sample ranges)
UCL = D₄R̄
LCL = D₃R̄
(A₂, D₃, D₄ are constants from standard SQC tables,
depending on sample size n)
3.3 p-Chart (Fraction Defective)
CL = p̄ (average fraction defective)
UCL = p̄ + 3√[p̄(1−p̄)/n]
LCL = p̄ − 3√[p̄(1−p̄)/n] (if negative, taken as 0)
3.4 c-Chart (Number of Defects)
Based on Poisson distribution:
CL = c̄
UCL = c̄ + 3√c̄
LCL = c̄ − 3√c̄
3.5 Interpreting Control Charts — Out-of-Control
Signals
Any point outside control limits
7+ consecutive points on one side of centre line (run)
Trend of 7+ points continuously increasing/decreasing
Cyclic patterns
Points hugging the centre line or the control limits
(non-random pattern)
3.6 Process Capability
Process Capability (Cp): Cp = (USL − LSL) / 6σ
Process Capability Index (Cpk): accounts for process
centering:
Cpk = min[(USL − x̄)/3σ , (x̄ − LSL)/3σ]
Cp/Cpk ≥ 1.33 generally considered capable; ≥ 2 is
"Six Sigma" quality level.
4. Acceptance Sampling
Deciding whether to accept or reject a lot/batch based on
inspecting a sample, rather than 100% inspection.
4.1 Why Sampling (vs 100% Inspection)
Economical, faster, less destructive testing damage,
feasible for large lots; but carries risk of wrong
decisions.
4.2 Types of Sampling Plans
1. Single Sampling Plan: one sample of size n; accept if
defectives ≤ acceptance number (c), else reject.
2. Double Sampling Plan: take first sample; if
inconclusive, take a second sample before deciding.
3. Multiple/Sequential Sampling Plan: several small
samples taken in sequence; decision made based on
cumulative results (uses Sequential Probability Ratio
Test — SPRT).
4.3 Operating Characteristic (OC) Curve
Graph of probability of acceptance (Pa) vs. fraction
defective (p) of the lot. Shows the discriminating power of a
sampling plan.
AQL (Acceptable Quality Level): worst quality level
acceptable as "good"; associated with Producer's Risk
(α) — probability of rejecting a good lot (Type I error),
typically α = 0.05.
LTPD (Lot Tolerance Percent Defective): worst quality
level a consumer will tolerate; associated with
Consumer's Risk (β) — probability of accepting a bad
lot (Type II error), typically β = 0.10.
4.4 Risks in Sampling
Risk Description Symbol
Producer's Risk Good lot rejected α
Consumer's Risk Bad lot accepted β
4.5 AOQ and AOQL
Average Outgoing Quality (AOQ): expected quality
level of outgoing lots after inspection/rescreening of
rejected lots.
AOQL (AOQ Limit): maximum value of AOQ curve
— worst possible average quality regardless of
incoming quality.
5. Total Quality Management (TQM)
5.1 Principles of TQM
1. Customer focus
2. Continuous improvement (Kaizen)
3. Employee involvement/empowerment
4. Process-centered approach
5. Top management commitment
6. Fact-based decision making
7. Integrated system across the organization
5.2 PDCA Cycle (Deming Cycle)
Plan: identify problem, plan improvement
Do: implement on small/trial scale
Check: study results, compare with expected
Act: standardize successful changes or revise
5.3 Deming's 14 Points (summary themes)
Constancy of purpose, adopt new philosophy, cease
dependence on mass inspection, don't award business on
price alone, improve constantly, institute training, institute
leadership, drive out fear, break down departmental
barriers, eliminate slogans/targets without methods,
eliminate numerical quotas, remove barriers to pride of
workmanship, institute education, take action for
transformation.
5.4 Quality Gurus (Comparison)
Guru Key Contribution
W. Edwards PDCA cycle, 14 points, statistical
Deming control
Quality trilogy (planning, control,
Joseph Juran
improvement); "fitness for use"
"Zero defects," "quality is free,"
Philip Crosby
conformance to requirements
Guru Key Contribution
Cause-and-effect (fishbone) diagram,
Kaoru Ishikawa
quality circles
Genichi Taguchi's loss function, robust design,
Taguchi design of experiments
Armand
Total Quality Control concept
Feigenbaum
6. Seven QC Tools (Basic Tools of
Quality)
1. Check Sheet — structured form for
collecting/recording data
2. Histogram — frequency distribution of data, shows
process spread/shape
3. Pareto Chart — bar chart ranking causes by frequency
(80/20 rule: 80% of problems from 20% of causes)
4. Cause-and-Effect (Fishbone/Ishikawa) Diagram —
identifies root causes; common categories: Man,
Machine, Material, Method, Measurement,
Environment (6M's)
5. Scatter Diagram — shows relationship/correlation
between two variables
6. Control Chart — as discussed in Section 3
7. Stratification / Flowchart — separates data by
source/category, or maps process steps
7. Six Sigma
7.1 Concept
A data-driven methodology aiming for near-perfect quality:
3.4 defects per million opportunities (DPMO),
corresponding to a process spread of ±6σ within
specification limits (with 1.5σ shift allowance).
7.2 DMAIC Methodology
Define — problem, customer requirements, project
goals
Measure — current process performance, collect data
Analyze — root cause of defects
Improve — implement solutions
Control — sustain improvements, monitor with control
charts
7.3 Sigma Levels
Sigma Level DPMO
2σ 308,537
Sigma Level DPMO
3σ 66,807
4σ 6,210
5σ 233
6σ 3.4
7.4 Roles in Six Sigma
Champion, Master Black Belt, Black Belt, Green Belt,
Yellow Belt (based on training/leadership level).
8. ISO Quality Standards
ISO 9000 series: Quality Management Systems (QMS)
ISO 9001: requirements for QMS certification
ISO 9004: guidelines for performance
improvement
ISO 14001: Environmental Management System
ISO 45001: Occupational Health & Safety
Management
ISO/TS 16949 (now IATF 16949): Automotive
industry QMS
Benefits of ISO certification: customer confidence,
standardized processes, access to global markets,
continuous improvement framework.
9. Reliability & Maintenance-related QC
Concepts (brief)
Reliability: probability that a product performs its
intended function for a specified time under stated
conditions.
Bathtub Curve: failure rate vs time — three phases:
1. Infant mortality (decreasing failure rate —
manufacturing defects)
2. Useful life (constant failure rate — random failures)
3. Wear-out (increasing failure rate — aging/fatigue)
MTBF (Mean Time Between Failures) and MTTR
(Mean Time To Repair) used to assess availability:
Availability = MTBF / (MTBF + MTTR)
FMEA (Failure Mode and Effect Analysis): systematic
method to identify potential failure modes, their
causes/effects, and prioritize using Risk Priority
Number (RPN = Severity × Occurrence × Detection).
10. Quick Revision Table
Concept Formula / Key Point
x̄ chart limits x̄̄ ± A₂R̄
R chart limits D₄R̄ (UCL), D₃R̄ (LCL)
p-chart limits p̄ ± 3√[p̄(1−p̄)/n]
c-chart limits c̄ ± 3√c̄
Process
Cp = (USL−LSL)/6σ
Capability
Producer's Risk α (reject good lot)
Consumer's Risk β (accept bad lot)
Six Sigma
3.4
DPMO
PDCA Plan-Do-Check-Act
Define-Measure-Analyze-Improve-
DMAIC
Control
RPN (FMEA) Severity × Occurrence × Detection
Notes for Exam Prep
Practice drawing and labeling control charts (x̄-R, p, c)
with correct limit formulas.
Be able to sketch the OC curve and explain
AQL/LTPD, Producer's/Consumer's risk clearly.
Memorize the 7 QC tools with one-line use-cases —
frequently asked as short-answer/definition questions.
Compare QC vs QA vs TQM and the quality gurus'
contributions — common table-based question.