Advanced Product Quality Planning

Advanced Product Quality Planning (or APQP) is a framework of procedures and

techniques used to develop products in industry, particularly the automotive
industry. It is quite similar to the concept of Design For Six Sigma (DFSS).

It is a defined process for a product development system for General Motors,

Ford, Chrysler and their suppliers. According to the Automotive Industry Action
Group (AIAG), the purpose of APQP is "to produce a product quality plan which
will support development of a product or service that will satisfy the customer."
The process is described in the AIAG manual 810-358-3003.

History
Advanced Product Quality Planning is a process developed in the late 1980's by
a commission of experts gathered from the 'Big Three' US automobile
manufacturers: Ford, GM and what was then Chrysler. This commission invested
five years to analyse the then-current automotive development and production
status in the US, Europe and especially in Japan. At the time, the success of the
Japanese automotive companies was starting to be remarkable in the US
market.

APQP is utilised today by these three companies and some affiliates. Tier I
suppliers are typically required to follow APQP procedures and techniques and
are also typically required to be audited and registered to ISO/TS 16949.

The APQP process is defined in the AIAG's APQP Manual, which is part of a
series of interrelated documents that the AIAG controls and publishes. The basis
for the make-up of a Process Control Plan is included in the APQP Manual. [1]
These manuals include:
The FMEA Manual
The Statistical process control (SPC) Manual
The Measurement Systems Analysis (MSA) Manual
The Production Part Approval Process (PPAP) Manual

The Automotive Industry Action Group (AIAG) is a non-profit association of

automotive companies founded in 1982.

Main content of APQP

APQP serve as a guide in the development process and also a standard way to
share results between suppliers and automotive companies. APQP specify three
phases: Development, Industrialisation and Product Launch. Through these
phases 23 main topics will be monitored. These 23 topics will be all completed
before the production is started. They cover aspects as: design robustness,
design testing and specification compliance, production process design, quality
inspection standards, process capability, production capacity, product packaging,
product testing and operators training plan between other items.
APQP focuses on:
o Up-front quality planning
o Determining if customers are satisfied by evaluating the output and
supporting continual improvement

APQP consists of five phases:

o Plan and Define Program
o Product Design and Development Verification
o Process Design and Development Verification
o Product and Process Validation
o Launch, Feedback, Assessment & Corrective Action

There are five major activities:

o Planning
o Product Design and Development
o Process Design and Development
o Product and Process Validation
o Production

The APQP process has seven major elements:

o Understanding the needs of the customer
o Proactive feedback and corrective action
o Designing within the process capabilities
o Analyzing and mitigating failure modes
o Verification and validation
o Design reviews
o Control special / critical characteristics

APQP 39 STEP PROCESS

Product Design Process Design Product &
Plan & Define
& Development & Development Process
Programme
Verification Verification Validation
•Design Goals •Design FMEA •Packaging Standards •Production Trial Run
•Reliability & Quality •Product & process •Measurement Systems
•DFMA
Goals Quality System Review Evaluation
•Preliminary Bill of •Preliminary Process
•Design Verification •Process Flow Chart
Materials Capability Study
•Preliminary Process
•Design Reviews •Floor Plan Layout •Production Part Approval
Flow
•Preliminary Listing of
•Production Validation
Special Products and •Prototype Build •Characteristics Matrix
Testing
Process Characteristics
•Product Assurance Plan •Engineering Drawings •Process FMEA •Packaging Evaluation
•Material Specifications •Pre-Launch Control Plan •Production Control Plan
•Drawing & Specification
•Process Instructions •Quality Planning Sign-Off
Changes
•New Equipment, Tooling
•Measurement Systems
& Facilities
Analysis Plan
Requirements
•Special Product and •Preliminary Process
Process Characteristics Capability Study Plan
•Prototype Control Plan •Packaging Specifications
•Gauges & Testing
Equipment
Requirements
FMEA (Failure Modes and Effects Analysis)

Failure Modes and Effects Analysis, or FMEA, is a methodology for identifying

the potential failure modes that a product or process may encounter, assessing
the risks associated with these failure modes, prioritization of these failure modes
according to their urgency, and prevention of the more urgent failure modes, i.e.,
the ones that are most likely to cause serious harm to the company.

The output of an FMEA cycle is the FMEA Table, which documents how
vulnerable a product or process is to its potential failure modes. The FMEA
table also shows the level of risk attached to each potential failure mode, and
the corrective actions needed (or already completed) to make the product or
process more robust. The FMEA Table generally consists of 16 to 17 columns,
with each column corresponding to a piece of information required by FMEA.
The FMEA is a proactive analysis tool, allowing engineers to anticipate failure
modes even before they happen, or even before a new product or process is
released. It also helps the engineer to prevent the negative effects of these
failure modes from reaching the customer, primarily by eliminating their causes
and increasing the chances of detecting them before they can do any damage.
The actions generated by a good FMEA cycle will also translate to better yield,
quality, reliability, and of course, greater customer satisfaction.

There are many types of FMEA, but the most widely used are probably the
following: 1) System FMEA, which is used for global systems; 2) Design or
Product FMEA, which is used for components or subsystems; 3) Process
FMEA, which is used for manufacturing and assembly processes; 4) Service
FMEA, which is used for services; and 5) Software FMEA, which is used for
software. In the semiconductor industry, the Design or Product FMEA and the
Process FMEA are the most frequently-encountered FMEA versions.

Despite the existence of many types of FMEA today, the basic structure and
process for executing them remains the same. Any FMEA process must include
the following steps, information details of which are documented in the FMEA
Table:

1) Assembly of the team;

2) Understanding of the Product or Process to be subjected to FMEA;
3) Breaking down of the product or process into its components or steps
(components and steps are also known as items);
4) Identification and assessment of the following for every item listed:
function(s), potential failure mode(s), failure mode effect(s), failure mode
cause(s), and controls for detecting or preventing the failure mode(s);
5) Evaluation of the risks associated with the failures modes and prioritizing
them according to importance;
6) Implementation of corrective actions to minimize the occurrence of the more
significant failure modes;
7) Reassessment of the product or process by another cycle of FMEA after the
actions have been completed; and
8) Regular updating of the FMEA Table.

The most critical information on the FMEA Table is the Risk Priority Number
(RPN), which is the numerical rating given to the level of risk associated with a
failure mode, and therefore denotes the urgency of addressing that failure
mode.

The RPN is actually the product of three (3) factors, namely, the severity of the
effect of the failure mode (SEV), the probability of the occurrence of the cause of
the failure mode (PF, for probability factor), and the effectiveness of the controls
for detecting and preventing the failure mode (DET). Thus, RPN = SEV x PF x
DET. The SEV, PF, and DET are also documented in the FMEA Table.

The FMEA Table is a living document, constantly changing from the time of its
first release when the product or process is still being designed until its archiving
after the product or process has been obsoleted.

Critical times or events that require an update to the FMEA Table include the
following:
1) when a new product or process is being designed or introduced;
2) when a critical change in the operating conditions of the product or process
occurs;
3) when the product or process itself undergoes a change;
4) when a new regulation that affects the product or process is instituted;
5) when customer complaints about the product or process are received; and
6) when an error in the FMEA Table is discovered or new information that affects
its contents come to light.

.
Table 1. Example of a Simplified FMEA Table

FMEA Methodology?
RETURN
SPC (Statistical Process Control)

Introduction:

Statistical process control (SPC) involves using statistical techniques to measure

and analyze the variation in processes. Most often used for manufacturing
processes, the intent of SPC is to monitor product quality and maintain
processes to fixed targets. Statistical quality control refers to using statistical
techniques for measuring and improving the quality of processes and includes
SPC in addition to other techniques, such as sampling plans, experimental
design, variation reduction, process capability analysis, and process
improvement plans.

SPC is used to monitor the consistency of processes used to manufacture a

product as designed. It aims to get and keep processes under control. No
matter how good or bad the design, SPC can ensure that the product is being
manufactured as designed and intended. Thus, SPC will not improve a poorly
designed product's reliability, but can be used to maintain the consistency of how
the product is made and, therefore, of the manufactured product itself and its as-
designed reliability.
A primary tool used for SPC is the control chart, a graphical representation of
certain descriptive statistics for specific quantitative measurements of the
manufacturing process. These descriptive statistics are displayed in the control
chart in comparison to their "in-control" sampling distributions. The comparison
detects any unusual variation in the manufacturing process, which could indicate
a problem with the process. Several different descriptive statistics can be used
in control charts and there are several different types of control charts that can
test for different causes, such as how quickly major vs. minor shifts in process
means are detected. Control charts are also used with product measurements to
analyze process capability and for continuous process improvement efforts.

Typical charts and analyses used to monitor and improve manufacturing process
consistency and capability (produced with Minitab statistical software).

Benefits:
 • Provides surveillance and feedback for keeping processes in control
• Signals when a problem with the process has occurred
• Detects assignable causes of variation
• Accomplishes process characterization
• Reduces need for inspection
• Monitors process quality
• Provides mechanism to make process changes and track effects of those
changes
• Once a process is stable (assignable causes of variation have been
eliminated), provides process capability analysis with comparison to the
product tolerance

Capabilities:

• All forms of SPC control charts

o Variable and attribute charts
o Average (X— ), Range (R), standard deviation (s), Shewhart,
CuSum, combined Shewhart-CuSum, exponentially weighted
moving average (EWMA)
• Selection of measures for SPC
• Process and machine capability analysis (Cp and Cpk)
• Process characterization
• Variation reduction
• Experimental design
• Quality problem solving
• Cause and effect diagrams

More on SPC Charts

RETURN
MSA (Measurement Systems Analysis)

1. A MSA study is a method of validating a measurement system to ensure

that it is fit for purpose. A measurement system consists of measuring
devices, procedures, definitions and people. The idea of validating the
measurement system is to minimise controllable factors that could
exaggerate the amount of variation in data.

2. There are two types of measurement data collected by a measurement

system.
a. Variable data is data where an actual measurement is taken, e.g.
length, temperature .
b. Attribute data is data where the measurement result is OK or not
OK. It does not give any information about the quality of the part,
only whether the part is to be accepted or rejected, e.g. a Go/No-go
gauge.

2. A MSA study typically consists of the completion of a Gauge

Reproducibility and Repeatability Study (Gauge R&R Study), where
multiple operators measure multiple parts, multiple times. The gauge
R&R study is carried out blind, so that operators do not know which parts
they are measuring. A MSA study may also include a study on gauge bias
and linearity. Both types of measurement systems can be subjected to a
gauge R&R study.
 3. For variable data measurement systems, analysis of the results from the
study using statistics (Minitab is commonly used to carry out gauge R&R
calculations) is carried out and shows how much variation comes from
differences in the operators, techniques, or the parts themselves. The
gauge R&R makes up part of the study variation, and is made up from
variation due to measuring equipment (repeatability) and variation due to
operators (reproducibility). Typically, the variation due to gauge R&R
should account for less than 30% of the study variation for a variable
study, and should demonstrate adequate discrimination with the number of
distinct categories (NDC) typically greater than 5.

4. Attribute studies are carried out out in a different way, and again it is
common to use statistical software to analyse the results. The attribute
study is looking for agreement between operators and commonly against
a known standard. Typically, agreement should be greater than 95% for
an acceptable attribute based measurement system.

5. Measurement Systems Analysis is a huge subject and the Automotive

Industry Action Group (AIAG) has devoted an inordinate amount of time to
the production of a ‘Measurement Systems Analysis’ manual for guidelines
on techniques to be used in the validation of measurement systems for
use in production processes.

RETURN
SPC Charts

SPC charts are essentially a sophisticated form of Time Series plot that enable
the stability of the process, and the type of variation involved, to be understood.

What do SPC charts detect? …… Changes!

• Changes in process average

A change in process
average: The plot shows a
very clear increase in the
process average, but SPC
Dimension

charts may detect much

smaller changes that are not
obvious to the human eye.
This might be the result of an
uncontrolled change or a
deliberate improvement to
the process.
Time

• Changes in process variation

A change in process
variation: The plot shows a
clear increase in the amount
Dimension

of variation in the process.

Again, this is a very marked
change, but SPC charts can
detect much smaller changes
in variation which wouldn’t
normally be obvious

Time
 • One off events such as “special causes”
One off events (Special
Causes): This plot shows a
process which appears to be
relatively stable and “in
Dimension

control”, with the exception of

two points that are significantly
higher than the rest. These
points are known as special
causes because they fall
outside of the expected
variation range of the process
and are likely to be as a result
Time of a special cause.

How Do SPC Charts Detect Changes?

1. The performance of the process is plotted as a time series plot

2. The level of variation in the process
3. Control limits (see below for an explanation) are drawn on the plot based on
the variation method
4. Each data point on the chart is then assessed against a number of ‘tests’
5. If the test are ‘broken’ then the process is not in control and needs
investigation or adjustment

Control Limits:
These are limits that are calculated using a statistical measure of variation, either
the “standard deviation” (σ) or the “range” (R). They represent the limits of the
process, such that all process data should fall within these bounds.

Control Limits

The ‘Tests’:

A process is out of control if:

o Any point is outside the control lines
o There is a run of 7 points above or below the mean line
o There is a run of 7 points going in the same direction
 o Any obvious non-random pattern (e.g. saw tooth shape)

There are more complex tests (carried out by software such as MiniTab) which
are designed to find much finer process changes but are not of use on the shop
floor.

SPC for In-Process Control

This historical data is used to Each future point will be

construct the control limits plotted here as is occurs

Time

As each result occurs it is plotted on the SPC chart and assessed using the tests
above to detect any change in the process. This gives real time information to the
operator of the process who may make adjustments based on data rather than
opinion or ‘gut feel’.

RETURN
PPAP: Production Part Approval Process

1. The purpose of PPAP is to determine if all customer engineering design

record and specification requirements are properly understood by the
supplier and that the process has the potential to produce product
consistently meeting these requirements during an actual production run
at the quoted production rate.

2. Typically, a Customer will ask for PPAP in the following situations:

a. A new part or product (i.e. a specific part, material, or colour not

previously supplied to the customer).
b. Correction of a discrepancy on a previously submitted part.
c. Product modified by an engineering change to design records,
specifications or materials.
d. Significant process changes as detailed within AIAG PPAP Manual
Fourth Edition (e.g. supplier change, new tooling, etc).

2. PPAP has 19 elements/requirements, and a typical submission would

feature the production of a pack of documentation which provides
objective evidence to show that the requirements detailed in paragraph 1
have been met. The customer would review the pack and master sample
parts and either approve or reject the PPAP. If the PPAP is approved, this
is permission to ship production parts to the customer facility. If the PPAP
is rejected, dependent on the nature of the rejection, parts would either be
supplied under waiver, or with a serious issue, the customer may not grant
permission to ship product. Whenever a PPAP is rejected, a corrective
action plan would need to be submitted to the customer to show how a
supplier plans to achieve full approval.

3. When PPAP is called up on a purchase order, production parts may not be

shipped to the customer without a customer approved PPAP or a waiver
from the customer granting permission to ship in lieu of PPAP approval.

RETURN
FMEA Methodology

Fundamental to a Process FMEA is the production of a process map that charts

the constituent discrete steps or parts of the relevant process.
In the case of a Design FMEA the Bill of Material is used (BOM)
For each discrete step identified above, an analysis is made of all conceivable
failure modes.
For each failure mode the Root Cause(s) are then identified.
This is undertaken in close association with the Process or Design Owner
(Brainstorming and the 5Y’s techniques are usually applied.)
For each Root Cause an assessment is then made of the risks involved in terms
of:
a. Its Occurrence
b. Its Severity
c. Its Detection
d. Its Extent (environmental only)
Each risk element is then ranked using the relevant tables to a ranking number
between 1 and 10.
(1 meaning Virtually Zero Risk, 10 meaning Extreme High Risk)
Each of these Rankings are then multiplied together to give an overall risk rating.
This composite rating is called the Risk Priority Number or RPN.
The RPN’s can then be prioritised (highest first) so as to ascertain the most
significant areas for corrective action.
By closely reviewing the individual RPN factors strong guidance is naturally given
as to the actions necessary to reduce the overall RPN.
The conclusions from this exercise (there may be more than one corrective
action possible) are then assessed as to outcome on the overall RPN from which
an actionable corrective action can be taken.
The process can then be repeated for the next highest RPN and so on.
Equally once a FMEA has been produced, investigations of any individual failure
mode or root cause can be individually assessed and the effect of any specific
corrective action evaluated.
Process FMEA Rankings
SEVERITY

Rankin Description Criteria

g
Failure without Warning
Highly Product is hazardous for production personnel,
10
Hazardous Failure affects safe product operation or non-
compliance with government regulations
Failure with Warning
Product is hazardous for production personnel,
9 Hazardous
Failure affects safe product operation or non-
compliance with government regulations
Major production disruption,
8 Very High High rework or scrap potential,
Loss of primary function
Minor disruption to production
7 High Probable sort and scrap procedure.
Primary function is impaired
Minor disruption to production,
A portion of the product may have to be scrapped
6 Moderate (no sorting)
Products primary function continues,
Secondary function is impaired
Minor disruption to production,
High rework potential,
5 Low
Product’s primary function continues,
Secondary function is impaired
Minor disruption to production,
Sort and rework potential
4 Very Low
Product is non-conformant,
some Customer dissatisfaction expected
Minor disruption to production,
A portion of the product may be reworked
3 Minor
Product is non-conformant,
some Customer dissatisfaction may occur
Minor disruption to production,
A portion of the product may be reworked
2 Very Minor
Product is non-conformant,
Customer dissatisfaction is unlikely

1 None No Effect
RPN process = S x O x D
Process FMEA Rankings
OCCURANCE

Rankin Description Failure Rate

g
20% Probability
10 Very High:
(200,000 PPM)
Failure is
almost
10% Probability
9 inevitable
(100,000 PPM)

5% Probability
8
High: (50,000 PPM)
Frequent
Failures 2% Probability
7
(20,000 PPM)

1% Probability
6
(10,000 PPM)
Moderate:
0.2% Probability
5 Mild level of
(2,000 PPM)
failure
0.04% Probability
4
(400 PPM)
Low:
0.0067% Probability
3 Isolated
(67 PPM)
Failures
Very Low:
0.0007% Probability
2 Infrequent
(7 PPM)
Failures
Remote:
> 0.0007% Probability
1 Failure is
(> 7 PPM)
unlikely

RPN process = S x O x D
Process FMEA Rankings
DETECTION

Rankin Description Criteria

g
Non- Current Control either does not exist or will not
10
Detectable detect the Cause of Failure or its Failure Mode

Current Control is Very Unlikely to detect the

9 Very Remote
Cause of Failure or its Failure Mode

Current Control is Unlikely to detect the Cause of

8 Remote
Failure or its Failure Mode

Current Control has a Very Low Capacity to detect

7 Very Low
the Cause of Failure or its Failure Mode

Current Control has a Low Capacity to detect the

6 Low
Cause of Failure or its Failure Mode

Current Control has a Moderate Capacity to

5 Moderate
detect the Cause of Failure or its Failure Mode

Moderately Current Control has a Moderately High Capacity

4
High to detect the Cause of Failure or its Failure Mode

Current Control has a High Capacity to detect the

3 High
Cause of Failure or its Failure Mode

Current Control has a Very High Capacity to

2 Very High
detect the Cause of Failure or its Failure Mode

Highly The Design Control will almost certainly detect the

1
Detectable potential Cause of the Failure or its Failure Mode

RPN process = S x O x D
Design FMEA Rankings
SEVERITY

Rankin Description Criteria

g
Failure without Warning
Highly
10 Product is unsafe for customer use,
Hazardous
or violates government regulations
Failure with Warning
9 Hazardous Product is unsafe for customer use,
or violates government regulations

8 Very High Product’s primary function fails

7 High Product’s primary function is impaired

Product’s primary function is unaffected,

6 Moderate
Secondary function fails

Product’s primary function is unaffected,

5 Low
Secondary function is impaired

Product is non-conformant, some Customer

4 Very Low
dissatisfaction expected

Product is non-conformant, some Customer

3 Minor
dissatisfaction may occur

Product is non-conformant, Customer

2 Very Minor
dissatisfaction is unlikely

1 None No Effect

RPN design = S x O x D
Design FMEA Rankings
OCCURANCE

Rankin Description Failure Rate

g
20% Probability
10 Very High:
(200,000 PPM)
Failure is a
Continuous
10% Probability
9 Problem
(100,000 PPM)

5% Probability
8
High: (50,000 PPM)
Frequent
Failures 2% Probability
7
(20,000 PPM)

1% Probability
6
(10,000 PPM)
Moderate:
0.2% Probability
5 Sporadic
(2,000 PPM)
failures
0.04% Probability
4
(400 PPM)

0.0067% Probability
3
Low: (67 PPM)
Relatively few
Failures 0.0007% Probability
2
(7 PPM)
Remote:
> 0.0007% Probability
1 Failure is
(> 7 PPM)
unlikely

RPN design = S x O x D
Design FMEA Rankings
DETECTION

Rankin Description Criteria

g
Non- No Known Controls available
10
Detectable Failure Mode or its Cause is undetectable
Very Remote Likelihood
9 Very Remote that Current Controls will detect Failure Mode or
its Cause
Remote Likelihood
8 Remote That the Current Controls will detect Failure Mode
or its Cause
Very Low Likelihood
7 Very Low That the Current Controls will detect Failure Mode
or its Cause
Low Likelihood
6 Low That the Current Controls will detect Failure Mode
or its Cause
Moderate Likelihood
5 Moderate That the Current Controls will detect Failure Mode
or its Cause
Moderately High Likelihood
Moderately
4 That the Current Controls will detect Failure Mode
High
or its Cause
High Likelihood
3 High That the Current Controls will detect Failure Mode
or its Cause
Very High Likelihood
2 Very High That the Current Controls will detect Failure Mode
or its Cause
Highly The Design Control will almost certainly detect the
1
Detectable potential Cause of the failure or its Failure Mode

RPN design = S x O x D

RETURN